AMD EPYC 4364P EPYC 4564P Intel Xeon E-2488 Benchmarks

Benchmarks preparing for a future article by Michael Larabel.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2405241-NE-EPYC4004428
Jump To Table - Results

View

Do Not Show Noisy Results
Do Not Show Results With Incomplete Data
Do Not Show Results With Little Change/Spread
List Notable Results
Show Result Confidence Charts

Limit displaying results to tests within:

AV1 2 Tests
BLAS (Basic Linear Algebra Sub-Routine) Tests 3 Tests
C++ Boost Tests 5 Tests
Chess Test Suite 2 Tests
Timed Code Compilation 10 Tests
C/C++ Compiler Tests 19 Tests
Compression Tests 3 Tests
CPU Massive 39 Tests
Creator Workloads 29 Tests
Cryptocurrency Benchmarks, CPU Mining Tests 2 Tests
Cryptography 6 Tests
Database Test Suite 10 Tests
Encoding 9 Tests
Fortran Tests 9 Tests
Game Development 6 Tests
HPC - High Performance Computing 29 Tests
Imaging 5 Tests
Java 4 Tests
Common Kernel Benchmarks 5 Tests
Linear Algebra 2 Tests
Large Language Models 2 Tests
Machine Learning 9 Tests
Molecular Dynamics 10 Tests
MPI Benchmarks 10 Tests
Multi-Core 51 Tests
NVIDIA GPU Compute 4 Tests
Intel oneAPI 7 Tests
OpenMPI Tests 17 Tests
Programmer / Developer System Benchmarks 15 Tests
Python 2 Tests
Raytracing 2 Tests
Renderers 6 Tests
Scientific Computing 15 Tests
Software Defined Radio 2 Tests
Server 18 Tests
Server CPU Tests 26 Tests
Single-Threaded 5 Tests
Texture Compression 2 Tests
Video Encoding 9 Tests
Common Workstation Benchmarks 3 Tests

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Geometric Means Per-Suite/Category
Show Wins / Losses Counts (Pie Chart)
Normalize Results
Remove Outliers Before Calculating Averages

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Disable Color Branding
Prefer Vertical Bar Graphs

Additional Graphs

Show Perf Per Core/Thread Calculation Graphs Where Applicable
Show Perf Per Clock Calculation Graphs Where Applicable

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
Xeon E-2488 @ DDR5-4800
April 01
  2 Days, 4 Hours, 1 Minute
EPYC 4364P @ DDR5-4800
March 26
  2 Days, 29 Minutes
EPYC 4364P @ DDR5-5200
April 28
  2 Days, 33 Minutes
EPYC 4564P @ DDR5-4800
April 21
  1 Day, 15 Hours, 46 Minutes
EPYC 4564P @ DDR5-5200
April 27
  1 Day, 14 Hours, 12 Minutes
Invert Hiding All Results Option
  1 Day, 21 Hours, 24 Minutes

Only show results where is faster than
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):


AMD EPYC 4364P EPYC 4564P Intel Xeon E-2488 BenchmarksProcessorMotherboardChipsetMemoryDiskGraphicsNetworkAudioOSKernelDesktopDisplay ServerCompilerFile-SystemScreen ResolutionXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200Intel Xeon E E-2488 @ 3.20GHz (8 Cores / 16 Threads)Supermicro Super Server X13SCL-F v0123456789 (1.1 BIOS)Intel Raptor Lake-S PCH2 x 32GB DRAM-4800MT/s Micron MTC20C2085S1EC48BA13201GB Micron_7450_MTFDKCC3T2TFS + 960GB Micron_7450_MTFDKBA960TFRASPEED2 x Intel I210Ubuntu 24.046.8.0-11-generic (x86_64)GNOME Shell 45.3X Server 1.21.1.11GCC 13.2.0ext41920x1080AMD EPYC 4364P 8-Core @ 5.57GHz (8 Cores / 16 Threads)Supermicro AS-3015A-I H13SAE-MF v1.00 (1.1 BIOS)AMD Device 14d8AMD Rembrandt Radeon HD Audio2 x 32GB DRAM-5200MT/sAMD EPYC 4564P 16-Core @ 5.88GHz (16 Cores / 32 Threads)2 x 32GB DRAM-4800MT/s Micron MTC20C2085S1EC48BA12 x 32GB DRAM-5200MT/sOpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler Details- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-fxIygj/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-fxIygj/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Disk Details- Xeon E-2488 @ DDR5-4800, EPYC 4364P @ DDR5-4800: NONE / errors=remount-ro,relatime,rw,stripe=64 / Block Size: 4096Processor Details- Xeon E-2488 @ DDR5-4800: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0x122 - Thermald 2.5.6- EPYC 4364P @ DDR5-4800: Scaling Governor: amd-pstate-epp performance (EPP: performance) - CPU Microcode: 0xa601203- EPYC 4364P @ DDR5-5200: Scaling Governor: amd-pstate-epp performance (EPP: performance) - CPU Microcode: 0xa601203- EPYC 4564P @ DDR5-4800: Scaling Governor: amd-pstate-epp performance (EPP: performance) - CPU Microcode: 0xa601203- EPYC 4564P @ DDR5-5200: Scaling Governor: amd-pstate-epp performance (EPP: performance) - CPU Microcode: 0xa601203Java Details- OpenJDK Runtime Environment (build 21.0.2+13-Ubuntu-2)Python Details- Python 3.12.2Security Details- Xeon E-2488 @ DDR5-4800: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected - EPYC 4364P @ DDR5-4800: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Vulnerable: Safe RET no microcode + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 4364P @ DDR5-5200: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Vulnerable: Safe RET no microcode + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 4564P @ DDR5-4800: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Vulnerable: Safe RET no microcode + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 4564P @ DDR5-5200: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Vulnerable: Safe RET no microcode + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200Result OverviewPhoronix Test Suite100%175%250%325%400%miniBUDENAMDJohn The RipperOpenSSLPrimesieveOpenVINOPETScStockfishm-queensAircrack-ngCpuminer-OptBlenderLuxCoreRenderACES DGEMM7-Zip CompressionCoremarkOSPRay StudioChaos Group V-RAYTimed Node.js CompilationHelsingOpenVKLStress-NGEmbreeBRL-CADJava JMHuvg266PostgreSQLIntel Open Image DenoiseParallel BZIP2 CompressionTimed Linux Kernel CompilationTimed FFmpeg CompilationOSPRaySVT-HEVConeDNNAppleseedTimed Godot Game Engine CompilationXmrigASTC EncoderKvazaarTimed LLVM CompilationTimed Mesa CompilationLAMMPS Molecular Dynamics SimulatorsrsRAN ProjectY-CruncherMemcachedGROMACSJPEG-XL libjxlTimed Gem5 CompilationHeFFTe - Highly Efficient FFT for ExascalenginxSPECFEM3DSVT-VP9ASKAPx265ScyllaDBTensorFloweasyWavelibavif avifencSVT-AV1CockroachDBGPAWGraphicsMagickDragonflydbPennantOpenRadiossApache HTTP ServerVVenCNAS Parallel BenchmarksQuantLibRocksDBOpenFOAMONNX RuntimeTimed PHP CompilationminiFECloverLeafSpeedbTimed Wasmer CompilationAlgebraic Multi-Grid BenchmarkXcompact3d Incompact3dPyBenchLlama.cppLULESHApache IoTDBPyTorchLlamafileDuckDBNumpy BenchmarkQMCPACKNode.js V8 Web Tooling BenchmarkRNNoiseGNU Octave BenchmarkZstd CompressionApache HadoopRedisFFmpegMariaDBHigh Performance Conjugate GradientRenaissanceTimed CPython CompilationClickHouseRawTherapeeZ3 Theorem ProverJava SciMarkMariaDB mariadb-slapDolfynWebP Image EncodeDaCapo Benchmark

AMD EPYC 4364P EPYC 4564P Intel Xeon E-2488 Benchmarkslammps: 20k Atomstensorflow: CPU - 512 - ResNet-50openradioss: Chrysler Neon 1Mopenradioss: INIVOL and Fluid Structure Interaction Drop Containerbuild-linux-kernel: allmodconfigopenfoam: drivaerFastback, Medium Mesh Size - Execution Timeopenfoam: drivaerFastback, Medium Mesh Size - Mesh Timetensorflow: CPU - 256 - ResNet-50blender: Barbershop - CPU-Onlyopenradioss: Bird Strike on Windshieldbuild-nodejs: Time To Compileopenvkl: vklBenchmarkCPU ISPCopenradioss: Bumper Beamtensorflow: CPU - 512 - GoogLeNetbuild-llvm: Unix Makefilesgpaw: Carbon Nanotubebuild-llvm: Ninjaxmrig: GhostRider - 1Mopenradioss: Rubber O-Ring Seal Installationllama-cpp: llama-2-70b-chat.Q5_0.ggufopenradioss: Cell Phone Drop Testbuild-gem5: Time To Compilehpcg: 104 104 104 - 60helsing: 14 digitstockfish: Chess Benchmarkapache-iotdb: 800 - 100 - 800 - 400apache-iotdb: 800 - 100 - 800 - 400openssl: RSA4096openssl: RSA4096apache-iotdb: 800 - 100 - 800 - 100apache-iotdb: 800 - 100 - 800 - 100tensorflow: CPU - 256 - GoogLeNetmariadb: oltp_read_write - 32securemark: SecureMark-TLSmariadb: oltp_read_write - 64mariadb: oltp_write_only - 64mariadb: oltp_update_non_index - 64mariadb: oltp_update_index - 32mariadb: oltp_update_index - 64mariadb: oltp_write_only - 32mariadb: oltp_update_non_index - 32build-godot: Time To Compileblender: Pabellon Barcelona - CPU-Onlyclickhouse: 100M Rows Hits Dataset, Third Runclickhouse: 100M Rows Hits Dataset, Second Runclickhouse: 100M Rows Hits Dataset, First Run / Cold Cachenpb: LU.Ctensorflow: CPU - 64 - ResNet-50blender: Classroom - CPU-Onlytensorflow: CPU - 512 - AlexNetonnx: fcn-resnet101-11 - CPU - Parallelonnx: fcn-resnet101-11 - CPU - Parallelonnx: ResNet50 v1-12-int8 - CPU - Standardonnx: ResNet50 v1-12-int8 - CPU - Standardminibude: OpenMP - BM2minibude: OpenMP - BM2brl-cad: VGR Performance Metricapache-iotdb: 800 - 100 - 500 - 400apache-iotdb: 800 - 100 - 500 - 400apache-iotdb: 800 - 100 - 500 - 100apache-iotdb: 800 - 100 - 500 - 100apache-iotdb: 500 - 100 - 800 - 400apache-iotdb: 500 - 100 - 800 - 400apache-iotdb: 500 - 100 - 800 - 100apache-iotdb: 500 - 100 - 800 - 100openssl: ChaCha20openssl: AES-128-GCMopenssl: ChaCha20-Poly1305openssl: AES-256-GCMopenssl: SHA512openssl: SHA256incompact3d: input.i3d 193 Cells Per Directionsvt-hevc: 1 - Bosphorus 4Kmysqlslap: 128pennant: sedovbigcloverleaf: clover_bm64_shortjava-jmh: Throughputospray: particle_volume/scivis/real_timeonnx: yolov4 - CPU - Parallelonnx: yolov4 - CPU - Parallelonnx: GPT-2 - CPU - Standardonnx: GPT-2 - CPU - Standardscylladb: Writesonnx: yolov4 - CPU - Standardonnx: yolov4 - CPU - Standardappleseed: Emilypgbench: 100 - 500 - Read Write - Average Latencypgbench: 100 - 500 - Read Writeapache-iotdb: 500 - 100 - 500 - 400apache-iotdb: 500 - 100 - 500 - 400luxcorerender: Danish Mood - CPUospray: particle_volume/pathtracer/real_timeeasywave: e2Asean Grid + BengkuluSept2007 Source - 1200npb: SP.Cmt-dgemm: Sustained Floating-Point Rateprimesieve: 1e13apache-iotdb: 500 - 100 - 500 - 100apache-iotdb: 500 - 100 - 500 - 100astcenc: Very Thoroughastcenc: Exhaustivepgbench: 100 - 800 - Read Only - Average Latencypgbench: 100 - 800 - Read Onlypgbench: 100 - 500 - Read Only - Average Latencypgbench: 100 - 500 - Read Onlypgbench: 100 - 800 - Read Write - Average Latencypgbench: 100 - 800 - Read Writepennant: leblancbigduckdb: IMDBffmpeg: libx265 - Platformffmpeg: libx265 - Video On Demandcompress-zstd: 19, Long Mode - Decompression Speedcompress-zstd: 19, Long Mode - Compression Speedllamafile: mistral-7b-instruct-v0.2.Q8_0 - CPUpytorch: CPU - 64 - ResNet-50ffmpeg: libx265 - Uploadospray: particle_volume/ao/real_timeblender: Junkshop - CPU-Onlycompress-zstd: 19 - Decompression Speedcompress-zstd: 19 - Compression Speedspecfem3d: Layered Halfspacenpb: IS.Dspecfem3d: Water-layered Halfspacetensorflow: CPU - 32 - ResNet-50specfem3d: Mount St. Helensllamafile: wizardcoder-python-34b-v1.0.Q6_K - CPUcompress-zstd: 8 - Decompression Speedcompress-zstd: 8 - Compression Speedoidn: RTLightmap.hdr.4096x4096 - CPU-Onlyblender: Fishy Cat - CPU-Onlynamd: STMV with 1,066,628 Atomsonnx: Faster R-CNN R-50-FPN-int8 - CPU - Standardonnx: Faster R-CNN R-50-FPN-int8 - CPU - Standardonnx: bertsquad-12 - CPU - Standardonnx: bertsquad-12 - CPU - Standardonnx: T5 Encoder - CPU - Standardonnx: T5 Encoder - CPU - Standardcockroach: KV, 60% Reads - 128tensorflow: CPU - 256 - AlexNetcockroach: KV, 50% Reads - 128cockroach: KV, 95% Reads - 128dragonflydb: 10 - 1:10askap: tConvolve MPI - Griddingaskap: tConvolve MPI - Degriddinggromacs: MPI CPU - water_GMX50_barecompress-zstd: 12 - Decompression Speedcompress-zstd: 12 - Compression Speedavifenc: 0numpy: npb: BT.Cvvenc: Bosphorus 4K - Fastonnx: ArcFace ResNet-100 - CPU - Parallelonnx: ArcFace ResNet-100 - CPU - Parallelredis: SET - 500petsc: Streamsopenfoam: drivaerFastback, Small Mesh Size - Execution Timeopenfoam: drivaerFastback, Small Mesh Size - Mesh Timedragonflydb: 10 - 1:100nginx: 1000apache: 1000nginx: 500apache: 500nginx: 200appleseed: Material Testermysqlslap: 64appleseed: Disney Materialblender: BMW27 - CPU-Onlybuild-linux-kernel: defconfigrenaissance: Genetic Algorithm Using Jenetics + Futuresonednn: Recurrent Neural Network Training - CPUcloverleaf: clover_bmospray-studio: 3 - 1080p - 32 - Path Tracer - CPUonednn: Recurrent Neural Network Inference - CPUpytorch: CPU - 1 - ResNet-50tensorflow: CPU - 64 - GoogLeNetv-ray: CPUastcenc: Thoroughospray-studio: 3 - 1080p - 1 - Path Tracer - CPUnpb: CG.Cqmcpack: simple-H2Oaskap: tConvolve MT - Degriddingaskap: tConvolve MT - Griddingonnx: Faster R-CNN R-50-FPN-int8 - CPU - Parallelonnx: Faster R-CNN R-50-FPN-int8 - CPU - Parallelspecfem3d: Tomographic Modelospray-studio: 2 - 1080p - 1 - Path Tracer - CPUospray-studio: 1 - 1080p - 1 - Path Tracer - CPUopenvino: Person Detection FP16 - CPUopenvino: Person Detection FP16 - CPUstress-ng: CPU Cacheospray-studio: 2 - 1080p - 32 - Path Tracer - CPUllama-cpp: llama-2-13b.Q4_0.ggufospray-studio: 1 - 1080p - 32 - Path Tracer - CPUospray: gravity_spheres_volume/dim_512/pathtracer/real_timememcached: 1:10memcached: 1:100build-python: Released Build, PGO + LTO Optimizedospray: gravity_spheres_volume/dim_512/ao/real_timeospray: gravity_spheres_volume/dim_512/scivis/real_timeluxcorerender: Orange Juice - CPUopenvino: Face Detection FP16 - CPUopenvino: Face Detection FP16 - CPUcompress-zstd: 8, Long Mode - Decompression Speedcompress-zstd: 8, Long Mode - Compression Speedluxcorerender: LuxCore Benchmark - CPUopenvino: Face Detection FP16-INT8 - CPUopenvino: Face Detection FP16-INT8 - CPUuvg266: Bosphorus 4K - Mediumonnx: GPT-2 - CPU - Parallelonnx: GPT-2 - CPU - Parallelincompact3d: input.i3d 129 Cells Per Directiononnx: bertsquad-12 - CPU - Parallelonnx: bertsquad-12 - CPU - Parallelluxcorerender: DLSC - CPUgraphics-magick: Sharpenopenvino: Machine Translation EN To DE FP16 - CPUopenvino: Machine Translation EN To DE FP16 - CPUonnx: T5 Encoder - CPU - Parallelonnx: T5 Encoder - CPU - Parallelopenvino: Person Detection FP32 - CPUopenvino: Person Detection FP32 - CPUopenvino: Noise Suppression Poconet-Like FP16 - CPUopenvino: Noise Suppression Poconet-Like FP16 - CPUopenvino: Road Segmentation ADAS FP16-INT8 - CPUopenvino: Road Segmentation ADAS FP16-INT8 - CPUgraphics-magick: Enhancedgraphics-magick: Noise-Gaussianopenvino: Person Vehicle Bike Detection FP16 - CPUopenvino: Person Vehicle Bike Detection FP16 - CPUopenvino: Road Segmentation ADAS FP16 - CPUopenvino: Road Segmentation ADAS FP16 - CPUopenvino: Person Re-Identification Retail FP16 - CPUopenvino: Person Re-Identification Retail FP16 - CPUospray-studio: 3 - 1080p - 16 - Path Tracer - CPUopenvino: Handwritten English Recognition FP16-INT8 - CPUopenvino: Handwritten English Recognition FP16-INT8 - CPUspeedb: Rand Fillopenvino: Vehicle Detection FP16-INT8 - CPUopenvino: Vehicle Detection FP16-INT8 - CPUopenvino: Handwritten English Recognition FP16 - CPUopenvino: Handwritten English Recognition FP16 - CPUopenvino: Face Detection Retail FP16-INT8 - CPUopenvino: Face Detection Retail FP16-INT8 - CPUspeedb: Rand Fill Syncopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUopenvino: Vehicle Detection FP16 - CPUopenvino: Vehicle Detection FP16 - CPUopenvino: Weld Porosity Detection FP16 - CPUopenvino: Weld Porosity Detection FP16 - CPUspeedb: Update Randopenvino: Age Gender Recognition Retail 0013 FP16 - CPUopenvino: Age Gender Recognition Retail 0013 FP16 - CPUopenvino: Weld Porosity Detection FP16-INT8 - CPUopenvino: Weld Porosity Detection FP16-INT8 - CPUonnx: CaffeNet 12-int8 - CPU - Parallelonnx: CaffeNet 12-int8 - CPU - Parallelonnx: CaffeNet 12-int8 - CPU - Standardonnx: CaffeNet 12-int8 - CPU - Standardopenvino: Face Detection Retail FP16 - CPUopenvino: Face Detection Retail FP16 - CPUjohn-the-ripper: MD5graphics-magick: Swirlgraphics-magick: Rotatespeedb: Read Rand Write Randgraphics-magick: HWB Color Spacespeedb: Read While Writinggraphics-magick: Resizingrocksdb: Rand Fillonnx: ResNet50 v1-12-int8 - CPU - Parallelonnx: ResNet50 v1-12-int8 - CPU - Parallelrocksdb: Rand Fill Syncrocksdb: Overwriterocksdb: Update Randspeedb: Rand Readrocksdb: Read While Writingrocksdb: Read Rand Write Randonnx: super-resolution-10 - CPU - Parallelonnx: super-resolution-10 - CPU - Paralleljohn-the-ripper: HMAC-SHA512rocksdb: Rand Readx265: Bosphorus 4Kspecfem3d: Homogeneous Halfspaceliquid-dsp: 1 - 256 - 512oidn: RT.hdr_alb_nrm.3840x2160 - CPU-Onlyoidn: RT.ldr_alb_nrm.3840x2160 - CPU-Onlyospray-studio: 2 - 1080p - 16 - Path Tracer - CPUospray-studio: 1 - 1080p - 16 - Path Tracer - CPUhadoop: Create - 100 - 100000node-web-tooling: z3: 2.smt2build-php: Time To Compileliquid-dsp: 2 - 256 - 512cpuminer-opt: Myriad-Groestlavifenc: 2m-queens: Time To Solveembree: Pathtracer ISPC - Asian Dragon Objcoremark: CoreMark Size 666 - Iterations Per Secondrawtherapee: Total Benchmark Timesvt-hevc: 1 - Bosphorus 1080pkvazaar: Bosphorus 4K - Mediumrenaissance: Savina Reactors.IOvvenc: Bosphorus 4K - Fasterjpegxl: JPEG - 90liquid-dsp: 8 - 256 - 512ffmpeg: libx265 - Livellama-cpp: llama-2-7b.Q4_0.ggufjava-scimark2: Compositerenaissance: Finagle HTTP Requestsembree: Pathtracer ISPC - Crownquantlib: Multi-Threadednamd: ATPase with 327,506 Atomsbuild-ffmpeg: Time To Compiletensorflow: CPU - 32 - GoogLeNetminife: Smallembree: Pathtracer ISPC - Asian Dragonhadoop: Create - 500 - 100000build-wasmer: Time To Compilestress-ng: System V Message Passingcpuminer-opt: Ringcoindacapobench: Eclipseminibude: OpenMP - BM1minibude: OpenMP - BM1svt-vp9: Visual Quality Optimized - Bosphorus 4Ktensorflow: CPU - 64 - AlexNetcpuminer-opt: Magijpegxl: PNG - 90vvenc: Bosphorus 1080p - Fastsrsran: PUSCH Processor Benchmark, Throughput Threadaircrack-ng: john-the-ripper: WPA PSKjohn-the-ripper: bcryptjohn-the-ripper: Blowfishliquid-dsp: 16 - 256 - 512cpuminer-opt: Garlicoinstress-ng: Vector Floating Pointstress-ng: Matrix 3D Mathliquid-dsp: 4 - 256 - 512stress-ng: Floating Pointstress-ng: Vector Shufflestress-ng: Fused Multiply-Addliquid-dsp: 2 - 256 - 32cpuminer-opt: scryptstress-ng: Wide Vector Mathliquid-dsp: 4 - 256 - 32liquid-dsp: 16 - 256 - 32liquid-dsp: 8 - 256 - 32liquid-dsp: 1 - 256 - 32cpuminer-opt: Deepcoinstress-ng: AVX-512 VNNIstress-ng: x86_64 RdRandcpuminer-opt: LBC, LBRY Creditscpuminer-opt: Skeincoincpuminer-opt: Triple SHA-256, Onecoincpuminer-opt: Quad SHA-256, Pyritecpuminer-opt: Blake-2 Ssrsran: PUSCH Processor Benchmark, Throughput Totalwebp: Quality 100, Lossless, Highest Compressionsvt-av1: Preset 4 - Bosphorus 4Kllamafile: llava-v1.5-7b-q4 - CPUnpb: SP.Bsvt-av1: Preset 13 - Bosphorus 4Ky-cruncher: 1Bjpegxl: JPEG - 100jpegxl: PNG - 100dacapobench: Tradesoapbuild-mesa: Time To Compileuvg266: Bosphorus 4K - Very Fastdacapobench: H2 Database Enginedacapobench: Apache Lucene Search Enginekvazaar: Bosphorus 4K - Ultra Fastnpb: EP.Castcenc: Fasttensorflow: CPU - 32 - AlexNetuvg266: Bosphorus 4K - Super Fastlulesh: dacapobench: jMonkeyEngineonednn: Deconvolution Batch shapes_1d - CPUcompress-7zip: Decompression Ratingcompress-7zip: Compression Ratingsvt-vp9: VMAF Optimized - Bosphorus 4Kkvazaar: Bosphorus 4K - Super Fastdacapobench: Spring Bootonednn: IP Shapes 1D - CPUz3: 1.smt2quantlib: Single-Threadeduvg266: Bosphorus 4K - Ultra Fastdacapobench: Apache Tomcatkvazaar: Bosphorus 4K - Very Fastamg: dacapobench: Tradebeanscompress-pbzip2: FreeBSD-13.0-RELEASE-amd64-memstick.img Compressiondacapobench: PMD Source Code Analyzerluxcorerender: Rainbow Colors and Prism - CPUdacapobench: BioJava Biological Data Frameworky-cruncher: 500Mnpb: MG.Cvvenc: Bosphorus 1080p - Fasterdolfyn: Computational Fluid Dynamicsdacapobench: Jythonuvg266: Bosphorus 1080p - Mediumastcenc: Mediumdacapobench: Apache Lucene Search Indexwebp: Quality 100, Losslessx265: Bosphorus 1080pdacapobench: Apache Kafkasvt-av1: Preset 8 - Bosphorus 4Keasywave: e2Asean Grid + BengkuluSept2007 Source - 240pybench: Total For Average Test Timessvt-hevc: 10 - Bosphorus 4Ksvt-vp9: PSNR/SSIM Optimized - Bosphorus 4Ksvt-hevc: 7 - Bosphorus 4Kaskap: tConvolve OpenMP - Degriddingaskap: tConvolve OpenMP - Griddingtensorflow: CPU - 1 - ResNet-50dacapobench: GraphChiprimesieve: 1e12svt-av1: Preset 12 - Bosphorus 4Kuvg266: Bosphorus 1080p - Very Fastonednn: IP Shapes 3D - CPUdacapobench: Avrora AVR Simulation Frameworkavifenc: 6avifenc: 6, Losslessuvg266: Bosphorus 1080p - Super Fastsvt-av1: Preset 4 - Bosphorus 1080pkvazaar: Bosphorus 1080p - Mediumtensorflow: CPU - 1 - AlexNetrnnoise: 26 Minute Long Talking Samplesvt-av1: Preset 8 - Bosphorus 1080psrsran: PDSCH Processor Benchmark, Throughput Totalonednn: Convolution Batch Shapes Auto - CPUdraco: Church Facadeuvg266: Bosphorus 1080p - Ultra Fastoctave-benchmark: dacapobench: FOP Print Formatterwebp: Quality 100, Highest Compressionavifenc: 10, Losslessdacapobench: Zxing 1D/2D Barcode Image Processingdraco: Lionkvazaar: Bosphorus 1080p - Super Fastkvazaar: Bosphorus 1080p - Very Fastdacapobench: Batik SVG Toolkitsvt-hevc: 7 - Bosphorus 1080phelsing: 12 digittensorflow: CPU - 1 - GoogLeNetsvt-vp9: Visual Quality Optimized - Bosphorus 1080ponednn: Deconvolution Batch shapes_3d - CPUlammps: Rhodopsin Proteinsrsran: PDSCH Processor Benchmark, Throughput Threadkvazaar: Bosphorus 1080p - Ultra Fastsvt-vp9: VMAF Optimized - Bosphorus 1080psvt-vp9: PSNR/SSIM Optimized - Bosphorus 1080pdacapobench: Apache Xalan XSLTsvt-hevc: 10 - Bosphorus 1080pwebp: Quality 100svt-av1: Preset 13 - Bosphorus 1080psvt-av1: Preset 12 - Bosphorus 1080pbuild-python: Defaultwebp: Defaultheffte: r2c - FFTW - double - 128heffte: r2c - FFTW - float - 128java-scimark2: Dense LU Matrix Factorizationjava-scimark2: Monte Carlojava-scimark2: Jacobi Successive Over-Relaxationjava-scimark2: Fast Fourier TransformXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52007.57922.98983.54538.421508.0422260.9881225.9549123.071281.54259.88915.937288153.9076.13719.392228.899697.9371742.9177.741.3693.26468.3787.58350437.41416898291524.1356494499195618.83000.2137.925603527876.84180020426706186418296034220068210974187089304169253110364.422411.15249.90249.50234.3940491.5623.10336.94192.28960.8561.0428124.46170224.10612.550313.737194127290.146265554977.2461842388382.4864886364119.49625580983465185580064812725680238002630205619324919044293971601233655726373.72036132.2657455.89647138.3236765943472.2685.88640123.5938.094217.89080126.674104832127.6677.83432323.27059311.79142412275.14577601971.45157.969145.99211387.054.204110245.67480.56570283981.46390.87411.8524319620.89955587323.1983448632.8676599.57558.7558.751906.18.917.9132.2528.895.93161173.171972.116.3144.5395371751604.85135.41673524223.1953.2871622922.202375.6485.60.21165.000.2056523.610642.350591.518810.92655.90786169.62063021.4189.1459866.683987.54522560.586199.095146.331.4342377.0170.6133.567675.4828326.364.97543.396723.04363327348.6569652.7320242.9447935.186084604951.3687132.22113097.8592582.29125745.6198692.95173.4085411128174.948967123.51108.338886.62931.9564.58954471548.7161.4978.172006012.230728849750.3424.0573033.562260.8224.385841.009352.89378166324742441133.9929.834023464.53824038.38812363.795393072373.873048853.56197.9212.734962.638513.581506.632.652375.4504.11.69384.0910.416.628.10864123.28419.752149086.687311.53722.1336118.4533.746.19876161.313134.5429.7115.21523.0236.35219.8649709.94401.4239.48101.2118.22437.964892945.70174.9287613015.95500.8653.07150.645.071559.062256070.418507.3116.04249.0230.05265.838281841.007499.477.781022.101.79347557.4081.70107587.7329.76816.179197911303172120915315254883724416479884.66040214.5472015941653489908300578244261942760197653313.179275.8803407210004751226019.4867.114929508202550000.420.4242275415148899223.3063.07654.449408213334843.7261.84372.85214.5499430017.46058342.6599.169.283695.411.76832.506130002500178.1516.074311.801668.713.388337214.40.6597750.05178.799869.7416.57577695638.13417858353.801437.47694312.452311.31059.56173.78278.0931.81415.463153.635441.7072956018787188411592700001356.3128846.062778.89795536674740.289138.2929762395.57127376667119.86576480.15253450000711050000439150000632366673294.491125583.0181538.346037.26139333884726527561601068.90.884.34813.3212750.97104.61433.89113.48413.650191432.51518.972128311339.591470.81205.9321160.5220.178126.321168276.85318615279499079.8528.6412202.9140818.6605019.224.69668723.0545370388034106.61248611229.35370814.92624688.5840.06710.421243336.5989.140022962.1784.33498436.1406.278415101.8383.1947.984282.182556.2213.71267521.095105.28089.317.5969514516.2608.96996.3315.91850.4718.506.400120.5332908.56.380535607115.406.4624264.734.16210533830139.72103.04897168.394.08051.33221.334.947927.236809.6197.51289.26295.69408383.6415.37649.426548.04216.8523.9326.478785.006211976.061426.112667.901023.238.03130.411060.53582.321312.3512926.9791276.7086430.33928.10276.24737.286327140.2691.07670.269217.495643.2231690.9175.461.09100.19436.0486.67485372.10322631087460.2164354682193565.97692.3123.746242003590.98159448437475168388265093200246198305177842310088224636329.179316.02228.79226.35211.5937048.8030.22259.85240.74889.1701.125973.05542327.52526.494662.344229365267.996889760269.0568672756359.2669856949106.4469450850683623544171328905125904863208206711384362553058135552301796566088390.69223283.0458761.27547177.8338717919571.1144.64920107.50819.323077.81820128.00894899100.027110.46448264.55047312.45440163248.62632968642.06168.642159.0439707.085.160364152.67972.28636279451.41090.86332.0773854110.94053201524.7613231643.80306101.93965.4565.271836.410.36.3445.2032.034.58353139.942026.017.9130.5280090501228.58129.55163219630.5249.5430423971.842313.4513.90.23127.120.4900719.102552.344985.115011.74957.03702142.06167444.0237.7664458.484052.34759359.713975.693349.851.5262394.8168.4107.789812.1622551.095.68136.248127.71253656582.3539918.9683311.3610441.6771484818585.2584772.3187686.6391505.2496630.23100527.89143.3799591132156.370961101.5297.918819.52496.6589.36846111329.6868.3091.422305610.324125397857.5321.4491791.991550.6419.334551.723646.5765081032171214373.8754.115854131.56729066.70720674.653052217988.242124192.99202.7723.878273.823214.42556.147.192324.5518.52.41283.7114.078.628.26788120.90922.534623681.567612.25972.743148.5582.327.00331142.81673.7154.238.03980.8414.59273.5050764.78834.1818.02221.523.851036.334401318.90422.6210117884.52882.7520.34392.933.132522.561288550.2826465.338.57465.1510.97728.185560400.418754.595.591427.321.46770680.9511.48191674.5062.21791.0622303331381692144575244321227625111547163.07441325.209980541155121609510720879472476441213789210.737993.15641170656676936663117.7357.793286206201066670.480.4838010376779047621.8948.12253.091396930006023.3851.37753.08215.5714505633.41942546.73212.3112.483036.513.08336.320151906667179.9812.735519.551925.016.326444058.21.7345442.47492.827278.8718.35447637136.54812527831.361868.89648326.365659.13161.45205.32332.4036.03518.902183.548802.9968571223271231742129600001691.2930715.141953.52778110004283.488606.9430389834.82107140000163.88416118.50212860000740166667411290000536613334405.851376354.004400145.018437.93180605137032010748831517.40.845.71110.5710661.66121.43725.24216.57816.689196428.19525.322000240452.251101.23204.8423171.1126.736454.672668043.876658780410184970.2938.8511591.9877320.5014518.432.09646731.0535065675027756.176853114211.89434711.55519566.6744.58410.634379045.7580.526024072.1689.23503146.8376.166532107.6574.9356.953429.452703.3616.77253512.416123.638112.054.6121124305.3018.140120.3019.96264.8014.537.028147.3127792.68.232205163143.285.6684235.304.20210163922168.47129.55927178.683.75362.02201.533.798906.8481082.7226.69240.66249.86448358.5416.59621.633539.87216.49926.7822.474174.131718030.922570.602487.93470.918.21530.701003.59585.751310.9622827.8789296.2109430.63925.38266.65739.343331145.3291.90662.522205.874636.8321699.3162.031.0695.48432.2216.73761375.91523143858462.1464182563194210.37697.3122.556295145891.77158499430421168593265119201401198779177579306926224118326.162315.13232.47236.47222.4737149.8830.53259.46240.23864.5611.160543.09150323.38126.357658.931226198268.426914416468.1669561449347.9771338740105.5870305285683661088101331340746734854273966311394125887757944053071787635833787.77034253.0458061.96916172.3638776376503.4654.61871107.7939.291798.06389123.9529421797.123510.70547263.51109812.35540473243.61646555022.10170.390152.78610027.115.330500153.59870.70647812851.39790.85572.0513900180.96152050224.4983265742.93941100.62765.2665.351926.910.26.2845.1131.954.63828139.482066.718.0133.3090441851269.37129.33205161230.7951.2869770351.822322.3520.30.23127.010.4934519.260951.912970.826315.21747.16751139.47767506.7237.5764052.783155.84729191.294256.283635.691.5462404.2171.0108.654808.6222935.275.69337.002527.02473684418.9826545.7101296.6009838.1730044762998.4285990.1287708.3994947.7596508.30101214.88142.9369581124156.069532101.7297.405821.72491.1486.82842091337.4767.4892.372307810.208925267898.4123.1191806.711619.8219.791050.531049.5764706882156213372.5255.115997921.20722506.57715304.667572195688.492104490.55200.9783.910703.821384.42560.147.142352.5527.92.41285.9713.978.618.31107120.28021.874643179.703712.54832.753048.2682.807.06097141.62972.7054.987.95991.3214.48275.7250764.78833.7317.79224.433.871029.884383019.13417.6410059734.53879.5020.37392.233.162495.551279550.2826230.858.44472.5711.05723.135537910.418615.125.631417.331.47531677.5831.47738676.6522.211781.6222343331391682138776255319497125511598653.09905322.610997461152820618187716486832519594213112210.918291.58551167703336940825417.6163.865203042196116670.480.4837900374589093822.1448.67653.755390943335972.0951.94253.50115.6889499103.50290546.65912.2812.493042.013.14137.180150576667179.9912.545476.911963.816.462543490.71.7376442.19593.717325.1018.45117758536.90212445240.761851.25641026.294657.34561.90206.40330.6236.02418.773183.648527.8158553223235232432116700001674.9930936.502029.45775086674251.398556.7229927440.69107146667162.88413926.82209743333735566667409386667536590004377.741371242.644374980.168389.15179835108731903745131470.50.845.74210.4811152.31122.90825.45516.51916.661197928.26325.251977240952.301169.68203.1039172.8926.656744.856668023.910898894510357171.4838.8211621.9998620.4324457.831.86648231.0435236177528156.154460111311.93432711.58920265.5245.14410.727376245.6379.604023902.1789.12502747.2855.990527109.0376.0657.233426.032739.3417.47256112.506125.277111.784.6784024155.3388.147119.8319.89164.6815.017.184148.3367530.08.168005083142.975.5874145.284.20510193891168.08129.02930178.203.82862.99203.163.834126.8261061.5226.13242.58251.44450361.0516.54633.291546.70616.57626.8123.717977.519617934.592541.202489.61450.1014.47139.73915.82349.91717.7112445.681245.1043139.37484.46177.58427.637596107.17129.06395.186166.513360.4633239.4101.621.1668.68277.7256.94846199.82542791247415.0471135797372775.014222.8106.7271922519128.54184015435859201579349473234051206240200156352931239542191.082166.82241.90240.37222.8046710.3439.07136.92422.88567.0211.763842.49599401.34949.9231248.077394575238.347589304762.0975706552322.247760997595.727678078212932099131325049722507091743101110214525919597110048439673392488142079.60404725.7461444.42709178.8175186159786.8238.8172285.634811.67777.60699131.43815184996.994310.3104153.3727318.69657520235.20677946814.14219.737132.12812630.559.53671182.11667.07681662482.66081.62620.70911288930.360138815716.1024968330.2360383.00569.2469.521890.511.56.5345.7634.068.8107873.731998.319.585.9959815411223.2287.85601157838.9634.7950202931.942279.9778.50.4067.340.7407216.784960.060465.480215.27186.36984157.01093864.1405.6490528.3119084.96406922.867317.336613.592.4682308.6207.572.424800.7843371.787.74627.349336.56243827317.9639471.0714199.8800729.7393797046043.06144355.05130637.70156935.63136714.05162256.48102.744509118285.93450553.5657.927887.01354.1275.5544561703.75066.63129.594356719.678112758784.8325.6192291.851917.6719.316651.784430.4876436061090107892.3886.562537425.46384546.98379848.921493775389.093850014.66193.5857.457827.319217.70592.5913.462278.0778.84.62301.8326.4415.708.40671118.91017.106541363.107215.84844.965759.77133.656.57601152.04491.3487.5010.181544.5517.22462.96931165.131551.7526.38302.454.111938.682046120.11794.338698164.831648.3221.99726.613.324743.731703650.2949580.0711.72680.5611.751359.296247430.4134850.475.962669.641.19855833.5201.11712894.6602.373283.9840666672461612770899269514618733212292562.58704386.4091707351229747703294135603493436839729127477.55840132.29317828333313121863934.1339.915238393192993330.850.8517491172477494620.0648.99740.8473860566711228.1336.33527.82326.8839954988.72897942.79122.8621.803270.617.32146.773147935000186.7113.125321.502288.130.790883133.52.7499925.541131.797330.0731.58357108029.89016207347.123541.50662749.8121245.30497.12316.11644.7447.23122.997183.692030.34916080543512429242869600003196.4259053.011924.70764156678242.9616537.8357429956.15103326667310.84781906.97206306667787633333404010000524930008421.152572055.588497821.55163173475397753613471434801611.50.827.83110.9018360.82160.65418.80326.95827.555243917.73042.201872208380.502117.29388.6309236.7145.878006.354768082.9865416895716718699.8462.2315541.1320320.7784425.054.07336747.6736257443333593.325543107416.9444138.92020223.6050.41410.927379563.89153.701625172.08117.25503071.1403.164536153.78107.3797.705023.74886.9414.7825216.669164.852165.554.0705623933.4475.932180.5423.94487.5114.377.291187.53712647.56.725365237201.925.7544035.153.8045974049245.56184.98969307.892.01157.21355.842.4616412.4491027.8299.12398.58408.36508545.3115.84711.526649.20414.14925.7835.0651125.53617230.852456.472464.14453.0513.69141.02891.84322.67714.3842351.5135226.0110440.68481.15183.64421.430603108.64132.62387.011154.989354.7643266.9110.051.2167.71273.4937.22104200.76844270858399.5673716025373632.314289.7103.3774075244132.13181519429779201182348099233313207212198538346114239192188.525165.35251.80248.89227.3947556.6940.24136.11425.47567.0651.763492.38299422.41749.9851249.619397138235.167765891460.6677605627321.297845431692.087931531912985352647325122440251392129622150215336847677110415799233398965227373.88367845.8061242.22565171.3075134161573.0428.9819586.840311.51997.51744133.16815172997.935810.2115152.3017838.85056517230.41697311664.18220.907127.02413374.359.57775382.47365.63698575192.68411.63910.72211082430.361138416616.1574952229.5529782.37769.8069.811908.310.76.7943.3834.218.9839972.752010.818.883.9703858861278.4485.25437026039.9933.1267765182.012245.7793.80.4166.780.7564618.663353.575766.028215.14436.16145162.32995084.8409.9692195.2121061.76266297.877477.196642.942.5432314.7220.972.106803.0444844.137.81826.828437.27143419833.5041077.6339183.2345130.7652716289782.08136360.72131437.32154835.06133534.11162348.33102.152381118385.31792253.2557.472882.71348.3772.8844247706.64565.69133.214387919.903812659036.6525.2522387.891977.0018.966452.721430.1322315801080106989.9888.842446936.66381927.32379238.978133761626.613844031.28193.4657.466497.369097.72590.8013.522299.7797.04.65301.7926.4615.928.15155122.62716.588468564.282715.55605.005858.43136.736.37806156.77390.4988.329.901587.7416.73476.63941195.121556.0725.15317.254.111938.322028320.09795.058667594.821653.6921.89729.843.334734.271716230.2949485.3511.36701.8411.721362.006233660.4134815.395.962672.511.20504829.1941.11577895.7732.373269.1241180002501672770195278511989133812259212.62524380.8181711991225361700416134971626429463129157587.50080133.32218133100013069306334.6539.561071375189644000.860.8617285171167538120.8148.87740.110387126671133736.33127.91427.3873951674.50646142.55023.1022.023265.217.43548.588149903333186.4213.825441.202300.731.365883365.12.7884625.301135.307673.6632.22097043629.44316150369.043500.13658749.9531248.810100.00322.53633.4448.52523.190183.592259.95316066443695437952923900003179.3259037.142011.17768090008199.6516452.4756410032.25103696667308.21778238.33205890000795966667404976667530186678356.742571394.828469095.71161203464097340613601427531611.50.827.83811.3718660.90165.28518.21127.22327.748247417.46442.601826208081.232123.37392.2377242.4946.418106.154568063.00636170373170888103.5262.5615541.1300020.3254425.254.54337048.1137869700033293.299879105316.9844258.65220961.3550.91410.963387164.16154.720224712.09118.06505572.7443.037533158.94111.0199.615172.445073.2314.9324726.691170.214166.813.8868523593.4215.890181.8324.17188.3014.967.414189.41312852.66.407195224203.395.7803995.103.8415893878248.39186.56936311.952.01657.34364.902.4615912.4811049.1301.57408.42419.26506560.8116.11728.577660.33614.13426.2537.2403127.93117774.362529.182518.58443.49OpenBenchmarking.org

LAMMPS Molecular Dynamics Simulator

LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 23Jun2022Model: 20k AtomsXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.177, N = 9SE +/- 0.120, N = 9SE +/- 0.174, N = 9SE +/- 0.342, N = 7SE +/- 0.297, N = 97.5798.0318.21514.47113.6911. (CXX) g++ options: -O3 -lm -ldl

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: ResNet-50Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200918273645SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 322.9830.4130.7039.7341.02

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Chrysler Neon 1MXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 17.37, N = 7SE +/- 2.88, N = 3SE +/- 10.68, N = 9SE +/- 5.48, N = 3SE +/- 4.49, N = 3983.541060.531003.59915.82891.84

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: INIVOL and Fluid Structure Interaction Drop ContainerXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200130260390520650SE +/- 8.95, N = 9SE +/- 10.02, N = 7SE +/- 5.87, N = 9SE +/- 4.48, N = 9SE +/- 2.52, N = 3538.42582.32585.75349.91322.67

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: allmodconfigXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 0.63, N = 3SE +/- 0.19, N = 3SE +/- 0.32, N = 3SE +/- 0.17, N = 3SE +/- 0.38, N = 31508.041312.351310.96717.71714.38

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Medium Mesh Size - Execution TimeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520060012001800240030002260.992926.982827.882445.682351.511. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Medium Mesh Size - Mesh TimeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520060120180240300225.95276.71296.21245.10226.011. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: ResNet-50Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200918273645SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 323.0730.3330.6339.3740.68

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Barbershop - Compute: CPU-OnlyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 4.06, N = 3SE +/- 0.87, N = 3SE +/- 0.87, N = 3SE +/- 0.08, N = 3SE +/- 0.46, N = 31281.54928.10925.38484.46481.15

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Bird Strike on WindshieldXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520060120180240300SE +/- 3.59, N = 9SE +/- 6.27, N = 9SE +/- 7.66, N = 9SE +/- 2.98, N = 12SE +/- 2.93, N = 12259.88276.24266.65177.58183.64

Timed Node.js Compilation

This test profile times how long it takes to build/compile Node.js itself from source. Node.js is a JavaScript run-time built from the Chrome V8 JavaScript engine while itself is written in C/C++. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Node.js Compilation 21.7.2Time To CompileXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 1.08, N = 3SE +/- 0.37, N = 3SE +/- 0.38, N = 3SE +/- 0.94, N = 3SE +/- 0.68, N = 3915.94737.29739.34427.64421.43

OpenVKL

OpenVKL is the Intel Open Volume Kernel Library that offers high-performance volume computation kernels and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 2.0.0Benchmark: vklBenchmarkCPU ISPCXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200130260390520650SE +/- 0.33, N = 3SE +/- 2.52, N = 3SE +/- 0.33, N = 3SE +/- 0.58, N = 3SE +/- 0.67, N = 3288327331596603MIN: 20 / MAX: 4338MIN: 25 / MAX: 4775MIN: 25 / MAX: 4820MIN: 47 / MAX: 8430MIN: 48 / MAX: 8450

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Bumper BeamXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 2.49, N = 12SE +/- 2.19, N = 12SE +/- 3.91, N = 12SE +/- 1.80, N = 15SE +/- 1.49, N = 15153.90140.26145.32107.17108.64

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: GoogLeNetXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 376.1391.0791.90129.06132.62

Timed LLVM Compilation

This test times how long it takes to compile/build the LLVM compiler stack. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: Unix MakefilesXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200160320480640800SE +/- 0.50, N = 3SE +/- 0.98, N = 3SE +/- 0.38, N = 3SE +/- 1.22, N = 3SE +/- 1.00, N = 3719.39670.27662.52395.19387.01

GPAW

GPAW is a density-functional theory (DFT) Python code based on the projector-augmented wave (PAW) method and the atomic simulation environment (ASE). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGPAW 23.6Input: Carbon NanotubeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050100150200250SE +/- 8.24, N = 12SE +/- 4.83, N = 12SE +/- 2.92, N = 3SE +/- 2.33, N = 12SE +/- 1.88, N = 3228.90217.50205.87166.51154.991. (CC) gcc options: -shared -fwrapv -O2 -lxc -lblas -lmpi

Timed LLVM Compilation

This test times how long it takes to compile/build the LLVM compiler stack. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: NinjaXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200150300450600750SE +/- 0.51, N = 3SE +/- 0.50, N = 3SE +/- 0.77, N = 3SE +/- 0.22, N = 3SE +/- 0.24, N = 3697.94643.22636.83360.46354.76

Xmrig

Xmrig is an open-source cross-platform CPU/GPU miner for RandomX, KawPow, CryptoNight and AstroBWT. This test profile is setup to measure the Xmrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgH/s, More Is BetterXmrig 6.21Variant: GhostRider - Hash Count: 1MXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52007001400210028003500SE +/- 3.35, N = 3SE +/- 5.04, N = 3SE +/- 10.83, N = 3SE +/- 4.40, N = 3SE +/- 22.46, N = 31742.91690.91699.33239.43266.91. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Rubber O-Ring Seal InstallationXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 5.45, N = 12SE +/- 9.50, N = 9SE +/- 8.68, N = 9SE +/- 3.52, N = 15SE +/- 1.00, N = 3177.74175.46162.03101.62110.05

Llama.cpp

Llama.cpp is a port of Facebook's LLaMA model in C/C++ developed by Georgi Gerganov. Llama.cpp allows the inference of LLaMA and other supported models in C/C++. For CPU inference Llama.cpp supports AVX2/AVX-512, ARM NEON, and other modern ISAs along with features like OpenBLAS usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b1808Model: llama-2-70b-chat.Q5_0.ggufXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.3060.6120.9181.2241.53SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 31.361.091.061.161.211. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Cell Phone Drop TestXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 3.30, N = 15SE +/- 3.05, N = 12SE +/- 3.06, N = 15SE +/- 1.27, N = 15SE +/- 1.68, N = 1293.26100.1995.4868.6867.71

Timed Gem5 Compilation

This test times how long it takes to compile Gem5. Gem5 is a simulator for computer system architecture research. Gem5 is widely used for computer architecture research within the industry, academia, and more. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Gem5 Compilation 23.0.1Time To CompileXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200100200300400500SE +/- 0.89, N = 3SE +/- 0.18, N = 3SE +/- 0.34, N = 3SE +/- 0.46, N = 3SE +/- 0.80, N = 3468.38436.05432.22277.73273.49

High Performance Conjugate Gradient

HPCG is the High Performance Conjugate Gradient and is a new scientific benchmark from Sandia National Lans focused for super-computer testing with modern real-world workloads compared to HPCC. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHigh Performance Conjugate Gradient 3.1X Y Z: 104 104 104 - RT: 60Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.12782, N = 12SE +/- 0.06329, N = 12SE +/- 0.04118, N = 3SE +/- 0.01035, N = 3SE +/- 0.05325, N = 37.583506.674856.737616.948467.221041. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi

Helsing

Helsing is an open-source POSIX vampire number generator. This test profile measures the time it takes to generate vampire numbers between varying numbers of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterHelsing 1.0-betaDigit Range: 14 digitXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520090180270360450SE +/- 0.23, N = 3SE +/- 0.35, N = 3SE +/- 0.18, N = 3SE +/- 0.45, N = 3SE +/- 0.37, N = 3437.41372.10375.92199.83200.771. (CC) gcc options: -O2 -pthread

Stockfish

This is a test of Stockfish, an advanced open-source C++11 chess benchmark that can scale up to 1024 CPU threads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 16.1Chess BenchmarkXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52009M18M27M36M45MSE +/- 121681.53, N = 15SE +/- 150504.97, N = 15SE +/- 160295.89, N = 3SE +/- 511472.18, N = 4SE +/- 574948.55, N = 151689829122631087231438584279124744270858-mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl-mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl-mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl-mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl1. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fno-peel-loops -fno-tracer -pedantic -O3 -funroll-loops -msse -msse3 -mpopcnt -mavx2 -mbmi -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto-partition=one -flto=jobserver

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200110220330440550SE +/- 1.92, N = 3SE +/- 4.57, N = 3SE +/- 4.56, N = 3SE +/- 7.45, N = 3SE +/- 4.31, N = 3524.13460.21462.14415.04399.56MAX: 26588.63MAX: 27836.55MAX: 27920.53MAX: 27057.75MAX: 27980.82

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520016M32M48M64M80MSE +/- 207001.08, N = 3SE +/- 766998.72, N = 3SE +/- 441409.11, N = 3SE +/- 669845.56, N = 3SE +/- 597457.28, N = 35649449964354682641825637113579773716025

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgverify/s, More Is BetterOpenSSL 3.3Algorithm: RSA4096Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520080K160K240K320K400KSE +/- 1626.68, N = 3SE +/- 1102.39, N = 3SE +/- 20.61, N = 3SE +/- 28.66, N = 3SE +/- 143.01, N = 3195618.8193565.9194210.3372775.0373632.31. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgsign/s, More Is BetterOpenSSL 3.3Algorithm: RSA4096Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003K6K9K12K15KSE +/- 28.09, N = 3SE +/- 6.93, N = 3SE +/- 2.36, N = 3SE +/- 10.46, N = 3SE +/- 5.90, N = 33000.27692.37697.314222.814289.71. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.51, N = 3SE +/- 0.94, N = 3SE +/- 0.48, N = 3SE +/- 0.16, N = 3SE +/- 0.21, N = 3137.92123.74122.55106.72103.37MAX: 23917.2MAX: 23948.92MAX: 23938.16MAX: 23939.91MAX: 23852.59

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520016M32M48M64M80MSE +/- 209117.79, N = 3SE +/- 503428.59, N = 3SE +/- 268932.10, N = 3SE +/- 109473.22, N = 3SE +/- 205604.00, N = 35603527862420035629514587192251974075244

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: GoogLeNetXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.05, N = 3SE +/- 0.00, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.08, N = 376.8490.9891.77128.54132.13

MariaDB

This is a MariaDB MySQL database server benchmark making use of sysbench rather than the existing pts/mysqlslap test profile that uses MariaDB with mysqlslap/mariadb-slap as the benchmark driver. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_read_write - Threads: 32Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520040K80K120K160K200KSE +/- 554.89, N = 3SE +/- 199.70, N = 3SE +/- 73.28, N = 3SE +/- 135.64, N = 3SE +/- 158.42, N = 3180020159448158499184015181519-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lz -lm -lpthread -ldl

SecureMark

SecureMark is an objective, standardized benchmarking framework for measuring the efficiency of cryptographic processing solutions developed by EEMBC. SecureMark-TLS is benchmarking Transport Layer Security performance with a focus on IoT/edge computing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmarks, More Is BetterSecureMark 1.0.4Benchmark: SecureMark-TLSXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520090K180K270K360K450KSE +/- 739.32, N = 3SE +/- 1137.99, N = 3SE +/- 1282.82, N = 3SE +/- 1244.71, N = 3SE +/- 1954.24, N = 34267064374754304214358594297791. (CC) gcc options: -pedantic -O3

MariaDB

This is a MariaDB MySQL database server benchmark making use of sysbench rather than the existing pts/mysqlslap test profile that uses MariaDB with mysqlslap/mariadb-slap as the benchmark driver. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_read_write - Threads: 64Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520040K80K120K160K200KSE +/- 562.40, N = 3SE +/- 60.79, N = 3SE +/- 211.83, N = 3SE +/- 108.10, N = 3SE +/- 389.87, N = 3186418168388168593201579201182-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lz -lm -lpthread -ldl

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_write_only - Threads: 64Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520070K140K210K280K350KSE +/- 798.38, N = 3SE +/- 526.67, N = 3SE +/- 419.19, N = 3SE +/- 559.43, N = 3SE +/- 56.85, N = 3296034265093265119349473348099-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lz -lm -lpthread -ldl

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_update_non_index - Threads: 64Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050K100K150K200K250KSE +/- 516.32, N = 3SE +/- 410.68, N = 3SE +/- 295.88, N = 3SE +/- 437.08, N = 3SE +/- 290.99, N = 3220068200246201401234051233313-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lz -lm -lpthread -ldl

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_update_index - Threads: 32Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050K100K150K200K250KSE +/- 290.25, N = 3SE +/- 249.27, N = 3SE +/- 331.57, N = 3SE +/- 113.22, N = 3SE +/- 307.26, N = 3210974198305198779206240207212-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lz -lm -lpthread -ldl

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_update_index - Threads: 64Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520040K80K120K160K200KSE +/- 250.58, N = 3SE +/- 125.38, N = 3SE +/- 308.54, N = 3SE +/- 130.01, N = 3SE +/- 282.52, N = 3187089177842177579200156198538-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lz -lm -lpthread -ldl

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_write_only - Threads: 32Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520080K160K240K320K400KSE +/- 574.53, N = 3SE +/- 133.55, N = 3SE +/- 471.57, N = 3SE +/- 101.24, N = 3SE +/- 441.31, N = 3304169310088306926352931346114-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lz -lm -lpthread -ldl

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_update_non_index - Threads: 32Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050K100K150K200K250KSE +/- 223.46, N = 3SE +/- 181.91, N = 3SE +/- 399.51, N = 3SE +/- 466.56, N = 3SE +/- 861.02, N = 3253110224636224118239542239192-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lz -lm -lpthread -ldl

Timed Godot Game Engine Compilation

This test times how long it takes to compile the Godot Game Engine. Godot is a popular, open-source, cross-platform 2D/3D game engine and is built using the SCons build system and targeting the X11 platform. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Godot Game Engine Compilation 4.0Time To CompileXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520080160240320400SE +/- 0.34, N = 3SE +/- 0.20, N = 3SE +/- 0.15, N = 3SE +/- 0.56, N = 3SE +/- 0.40, N = 3364.42329.18326.16191.08188.53

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Pabellon Barcelona - Compute: CPU-OnlyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520090180270360450SE +/- 1.05, N = 3SE +/- 0.09, N = 3SE +/- 0.40, N = 3SE +/- 0.09, N = 3SE +/- 0.18, N = 3411.15316.02315.13166.82165.35

ClickHouse

ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, Third RunXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520060120180240300SE +/- 1.74, N = 3SE +/- 1.72, N = 3SE +/- 4.93, N = 3SE +/- 1.20, N = 3SE +/- 2.29, N = 5249.90228.79232.47241.90251.80MIN: 9.25 / MAX: 12000MIN: 10.54 / MAX: 10000MIN: 10.67 / MAX: 12000MIN: 15.08 / MAX: 7500MIN: 15.75 / MAX: 10000

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, Second RunXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050100150200250SE +/- 3.87, N = 3SE +/- 1.92, N = 3SE +/- 1.00, N = 3SE +/- 1.60, N = 3SE +/- 1.51, N = 5249.50226.35236.47240.37248.89MIN: 9.27 / MAX: 12000MIN: 10.69 / MAX: 10000MIN: 10.66 / MAX: 10000MIN: 15.16 / MAX: 10000MIN: 15.92 / MAX: 8571.43

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, First Run / Cold CacheXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050100150200250SE +/- 1.98, N = 3SE +/- 2.84, N = 3SE +/- 1.84, N = 3SE +/- 2.08, N = 3SE +/- 2.37, N = 5234.39211.59222.47222.80227.39MIN: 9.14 / MAX: 12000MIN: 10.43 / MAX: 10000MIN: 10.49 / MAX: 12000MIN: 12.9 / MAX: 7500MIN: 13.31 / MAX: 10000

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: LU.CXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520010K20K30K40K50KSE +/- 714.33, N = 15SE +/- 662.65, N = 12SE +/- 613.23, N = 15SE +/- 545.16, N = 15SE +/- 521.71, N = 1540491.5637048.8037149.8846710.3447556.691. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: ResNet-50Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200918273645SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 323.1030.2230.5339.0740.24

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Classroom - Compute: CPU-OnlyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520070140210280350SE +/- 1.57, N = 3SE +/- 0.38, N = 3SE +/- 0.07, N = 3SE +/- 0.14, N = 3SE +/- 0.07, N = 3336.94259.85259.46136.92136.11

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: AlexNetXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520090180270360450SE +/- 0.02, N = 3SE +/- 0.11, N = 3SE +/- 0.55, N = 3SE +/- 0.05, N = 3SE +/- 1.74, N = 3192.28240.74240.23422.88425.47

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: fcn-resnet101-11 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 11.74, N = 14SE +/- 8.12, N = 15SE +/- 12.91, N = 15SE +/- 4.73, N = 3SE +/- 1.90, N = 3960.86889.17864.56567.02567.071. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: fcn-resnet101-11 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.39690.79381.19071.58761.9845SE +/- 0.013081, N = 14SE +/- 0.010436, N = 15SE +/- 0.018640, N = 15SE +/- 0.014643, N = 3SE +/- 0.005924, N = 31.0428121.1259701.1605401.7638401.7634901. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: ResNet50 v1-12-int8 - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001.00392.00783.01174.01565.0195SE +/- 0.03583, N = 3SE +/- 0.02531, N = 15SE +/- 0.00310, N = 3SE +/- 0.03044, N = 15SE +/- 0.05300, N = 154.461703.055423.091502.495992.382991. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: ResNet50 v1-12-int8 - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520090180270360450SE +/- 1.81, N = 3SE +/- 2.89, N = 15SE +/- 0.32, N = 3SE +/- 4.78, N = 15SE +/- 9.29, N = 15224.11327.53323.38401.35422.421. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

miniBUDE

MiniBUDE is a mini application for the the core computation of the Bristol University Docking Engine (BUDE). This test profile currently makes use of the OpenMP implementation of miniBUDE for CPU benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001122334455SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.06, N = 312.5526.4926.3649.9249.991. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

OpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 0.15, N = 3SE +/- 0.43, N = 3SE +/- 0.13, N = 3SE +/- 0.90, N = 3SE +/- 1.54, N = 3313.74662.34658.931248.081249.621. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

BRL-CAD

BRL-CAD is a cross-platform, open-source solid modeling system with built-in benchmark mode. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgVGR Performance Metric, More Is BetterBRL-CAD 7.38.2VGR Performance MetricXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520090K180K270K360K450K1941272293652261983945753971381. (CXX) g++ options: -std=c++17 -pipe -fvisibility=hidden -fno-strict-aliasing -fno-common -fexceptions -ftemplate-depth-128 -m64 -ggdb3 -O3 -fipa-pta -fstrength-reduce -finline-functions -flto -ltcl8.6 -lnetpbm -lregex_brl -lz_brl -lassimp -ldl -lm -ltk8.6

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 400Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520060120180240300SE +/- 4.53, N = 3SE +/- 1.83, N = 3SE +/- 1.82, N = 3SE +/- 2.23, N = 3SE +/- 2.35, N = 3290.14267.99268.42238.34235.16MAX: 26535.45MAX: 27513.85MAX: 27440.82MAX: 26718.64MAX: 26882.16

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 400Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520017M34M51M68M85MSE +/- 787042.09, N = 3SE +/- 239601.11, N = 3SE +/- 261270.67, N = 3SE +/- 362879.93, N = 3SE +/- 360981.97, N = 36265554968897602691441647589304777658914

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.20, N = 3SE +/- 0.06, N = 3SE +/- 0.29, N = 3SE +/- 0.11, N = 3SE +/- 0.29, N = 377.2469.0568.1662.0960.66MAX: 23891.04MAX: 23877.52MAX: 23889.9MAX: 23874.44MAX: 23861.25

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520017M34M51M68M85MSE +/- 232749.13, N = 3SE +/- 75302.81, N = 3SE +/- 296062.71, N = 3SE +/- 113056.05, N = 3SE +/- 390613.00, N = 36184238868672756695614497570655277605627

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520080160240320400SE +/- 10.20, N = 3SE +/- 1.80, N = 3SE +/- 6.12, N = 3SE +/- 8.45, N = 3SE +/- 5.40, N = 3382.48359.26347.97322.24321.29MAX: 27376.22MAX: 27346.14MAX: 27085.8MAX: 27055.68MAX: 27126.33

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020M40M60M80M100MSE +/- 922667.78, N = 3SE +/- 357039.39, N = 3SE +/- 255843.12, N = 3SE +/- 919852.08, N = 3SE +/- 818556.17, N = 36488636469856949713387407760997578454316

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.54, N = 3SE +/- 0.47, N = 3SE +/- 0.24, N = 3SE +/- 0.76, N = 3SE +/- 0.82, N = 3119.49106.44105.5895.7292.08MAX: 13845.12MAX: 13870.13MAX: 15103.61MAX: 11335.36MAX: 11356.99

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020M40M60M80M100MSE +/- 262085.32, N = 3SE +/- 351738.51, N = 3SE +/- 282897.65, N = 3SE +/- 224737.25, N = 3SE +/- 368630.91, N = 36255809869450850703052857678078279315319

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: ChaCha20Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030000M60000M90000M120000M150000MSE +/- 7700475.42, N = 3SE +/- 85172787.78, N = 3SE +/- 32786296.71, N = 3SE +/- 23437434.86, N = 3SE +/- 13732325.68, N = 33465185580068362354417683661088101293209913131298535264731. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: AES-128-GCMXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050000M100000M150000M200000M250000MSE +/- 45687236.99, N = 3SE +/- 53677368.01, N = 3SE +/- 48206635.82, N = 3SE +/- 105284888.52, N = 3SE +/- 197196146.33, N = 3648127256801328905125901331340746732504972250702512244025131. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: ChaCha20-Poly1305Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020000M40000M60000M80000M100000MSE +/- 5622113.45, N = 3SE +/- 17581304.10, N = 3SE +/- 11858939.41, N = 3SE +/- 14328313.13, N = 3SE +/- 63328398.60, N = 323800263020486320820674854273966391743101110921296221501. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: AES-256-GCMXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050000M100000M150000M200000M250000MSE +/- 5611535.27, N = 3SE +/- 93170961.59, N = 3SE +/- 101608115.02, N = 3SE +/- 47970633.04, N = 3SE +/- 27189836.88, N = 3561932491901138436255301139412588772145259195972153368476771. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: SHA512Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002000M4000M6000M8000M10000MSE +/- 2787035.99, N = 3SE +/- 1105217.67, N = 3SE +/- 749962.39, N = 3SE +/- 1453158.01, N = 3SE +/- 2951460.39, N = 344293971605813555230579440530711004843967110415799231. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: SHA256Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52007000M14000M21000M28000M35000MSE +/- 27097230.48, N = 3SE +/- 18056187.30, N = 3SE +/- 8102567.82, N = 3SE +/- 8523911.43, N = 3SE +/- 7892274.19, N = 312336557263179656608831787635833733924881420339896522731. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

Xcompact3d Incompact3d

Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 193 Cells Per DirectionXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.87, N = 15SE +/- 0.76, N = 3SE +/- 0.27, N = 3SE +/- 1.01, N = 3SE +/- 0.58, N = 1073.7290.6987.7779.6073.881. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 1 - Input: Bosphorus 4KXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001.3052.613.9155.226.525SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 32.263.043.045.745.801. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

MariaDB mariadb-slap

This is a MariaDB MySQL database server benchmark making use of mysqlslap (mariadb-slap). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB mariadb-slap 11.5Clients: 128Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200130260390520650SE +/- 1.05, N = 3SE +/- 0.57, N = 3SE +/- 0.74, N = 3SE +/- 0.55, N = 3SE +/- 0.54, N = 3574587580614612-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lz -lm -lpthread -ldl

Pennant

Pennant is an application focused on hydrodynamics on general unstructured meshes in 2D. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: sedovbigXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001428425670SE +/- 0.67, N = 4SE +/- 0.86, N = 15SE +/- 0.59, N = 6SE +/- 0.31, N = 15SE +/- 0.35, N = 855.9061.2861.9744.4342.231. (CXX) g++ options: -fopenmp -lmpi_cxx -lmpi

CloverLeaf

CloverLeaf is a Lagrangian-Eulerian hydrodynamics benchmark. This test profile currently makes use of CloverLeaf's OpenMP version. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCloverLeaf 1.3Input: clover_bm64_shortXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3138.32177.83172.36178.81171.301. (F9X) gfortran options: -O3 -march=native -funroll-loops -fopenmp

Java JMH

This very basic test profile runs the stock benchmark of the Java JMH benchmark via Maven. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/s, More Is BetterJava JMHThroughputXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520016000M32000M48000M64000M80000M36765943472.2738717919571.1138776376503.4775186159786.8275134161573.04

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: particle_volume/scivis/real_timeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.00749, N = 3SE +/- 0.00354, N = 3SE +/- 0.00347, N = 3SE +/- 0.00801, N = 3SE +/- 0.00296, N = 35.886404.649204.618718.817228.98195

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: yolov4 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 1.78, N = 3SE +/- 1.34, N = 15SE +/- 1.14, N = 15SE +/- 0.47, N = 3SE +/- 1.04, N = 4123.59107.51107.7985.6386.841. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: yolov4 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.11600, N = 3SE +/- 0.12476, N = 15SE +/- 0.10193, N = 15SE +/- 0.06318, N = 3SE +/- 0.13812, N = 48.094219.323079.2917911.6777011.519901. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: GPT-2 - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.02440, N = 3SE +/- 0.07551, N = 15SE +/- 0.01376, N = 3SE +/- 0.09699, N = 3SE +/- 0.08068, N = 157.890807.818208.063897.606997.517441. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: GPT-2 - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.39, N = 3SE +/- 1.28, N = 15SE +/- 0.21, N = 3SE +/- 1.70, N = 3SE +/- 1.39, N = 15126.67128.01123.95131.44133.171. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

ScyllaDB

This is a benchmark of ScyllaDB and is making use of Apache Cassandra's cassandra-stress for conducting the benchmark. ScyllaDB is an open-source distributed NoSQL data store that is compatible with Apache Cassandra while focusing on higher throughput and lower latency. ScyllaDB uses a sharded design on each node. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterScyllaDB 5.2.9Test: WritesXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030K60K90K120K150KSE +/- 395.91, N = 3SE +/- 446.24, N = 3SE +/- 60.93, N = 3SE +/- 78.75, N = 3SE +/- 255.51, N = 31048329489994217151849151729

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: yolov4 - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 1.27, N = 3SE +/- 5.32, N = 15SE +/- 5.00, N = 15SE +/- 0.60, N = 3SE +/- 0.70, N = 3127.67100.0397.1296.9997.941. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: yolov4 - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.07886, N = 3SE +/- 0.63014, N = 15SE +/- 0.57269, N = 15SE +/- 0.06449, N = 3SE +/- 0.07376, N = 37.8343210.4644810.7054710.3104010.211501. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: EmilyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520070140210280350323.27264.55263.51153.37152.30

PostgreSQL

This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 500 - Mode: Read Write - Average LatencyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.101, N = 3SE +/- 0.141, N = 4SE +/- 0.078, N = 3SE +/- 0.094, N = 4SE +/- 0.109, N = 311.79112.45412.3558.6968.8501. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 500 - Mode: Read WriteXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520012K24K36K48K60KSE +/- 362.37, N = 3SE +/- 453.73, N = 4SE +/- 256.18, N = 3SE +/- 617.95, N = 4SE +/- 699.92, N = 342412401634047357520565171. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 400Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520060120180240300SE +/- 1.91, N = 3SE +/- 3.81, N = 3SE +/- 2.74, N = 3SE +/- 3.26, N = 3SE +/- 3.67, N = 3275.14248.62243.61235.20230.41MAX: 27185.57MAX: 26892.13MAX: 26975.13MAX: 26852.2MAX: 26735.87

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 400Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520015M30M45M60M75MSE +/- 245819.57, N = 3SE +/- 382550.66, N = 3SE +/- 418677.10, N = 3SE +/- 438509.29, N = 3SE +/- 489754.87, N = 35776019763296864646555026779468169731166

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Danish Mood - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.94051.8812.82153.7624.7025SE +/- 0.01, N = 15SE +/- 0.02, N = 3SE +/- 0.02, N = 11SE +/- 0.06, N = 3SE +/- 0.04, N = 31.452.062.104.144.18MIN: 0.39 / MAX: 1.82MIN: 0.66 / MAX: 2.48MIN: 0.64 / MAX: 2.54MIN: 1.71 / MAX: 4.75MIN: 1.81 / MAX: 4.75

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: particle_volume/pathtracer/real_timeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050100150200250SE +/- 0.42, N = 3SE +/- 0.86, N = 3SE +/- 0.61, N = 3SE +/- 0.83, N = 3SE +/- 0.68, N = 3157.97168.64170.39219.74220.91

easyWave

The easyWave software allows simulating tsunami generation and propagation in the context of early warning systems. EasyWave supports making use of OpenMP for CPU multi-threading and there are also GPU ports available but not currently incorporated as part of this test profile. The easyWave tsunami generation software is run with one of the example/reference input files for measuring the CPU execution time. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BettereasyWave r34Input: e2Asean Grid + BengkuluSept2007 Source - Time: 1200Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 0.20, N = 3SE +/- 0.14, N = 3SE +/- 0.21, N = 3SE +/- 0.20, N = 3SE +/- 0.25, N = 3145.99159.04152.79132.13127.021. (CXX) g++ options: -O3 -fopenmp

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.CXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003K6K9K12K15KSE +/- 4.20, N = 3SE +/- 2.24, N = 3SE +/- 1.44, N = 3SE +/- 175.17, N = 3SE +/- 146.76, N = 511387.059707.0810027.1112630.5513374.351. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

ACES DGEMM

This is a multi-threaded DGEMM benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterACES DGEMM 1.0Sustained Floating-Point RateXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.054700, N = 3SE +/- 0.133403, N = 15SE +/- 0.144416, N = 15SE +/- 0.166167, N = 15SE +/- 0.258971, N = 124.2041105.1603645.3305009.5367119.5777531. (CC) gcc options: -O3 -march=native -fopenmp

Primesieve

Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 12.1Length: 1e13Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050100150200250SE +/- 0.18, N = 3SE +/- 0.09, N = 3SE +/- 0.15, N = 3SE +/- 0.09, N = 3SE +/- 0.20, N = 3245.67152.68153.6082.1282.471. (CXX) g++ options: -O3

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.51, N = 3SE +/- 0.41, N = 3SE +/- 0.31, N = 3SE +/- 0.42, N = 3SE +/- 0.10, N = 380.5672.2870.7067.0765.63MAX: 15084.51MAX: 13851.96MAX: 13822.83MAX: 12570.5MAX: 12576.01

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520015M30M45M60M75MSE +/- 205982.51, N = 3SE +/- 161254.73, N = 3SE +/- 42980.57, N = 3SE +/- 313014.56, N = 3SE +/- 41672.18, N = 35702839863627945647812856816624869857519

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: Very ThoroughXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.60391.20781.81172.41563.0195SE +/- 0.0116, N = 3SE +/- 0.0023, N = 3SE +/- 0.0010, N = 3SE +/- 0.0070, N = 3SE +/- 0.0093, N = 31.46391.41091.39792.66082.68411. (CXX) g++ options: -O3 -flto -pthread

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: ExhaustiveXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.36880.73761.10641.47521.844SE +/- 0.0051, N = 3SE +/- 0.0007, N = 3SE +/- 0.0000, N = 3SE +/- 0.0018, N = 3SE +/- 0.0020, N = 30.87410.86330.85571.62621.63911. (CXX) g++ options: -O3 -flto -pthread

PostgreSQL

This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 800 - Mode: Read Only - Average LatencyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.46730.93461.40191.86922.3365SE +/- 0.005, N = 3SE +/- 0.030, N = 3SE +/- 0.023, N = 3SE +/- 0.009, N = 3SE +/- 0.006, N = 31.8522.0772.0510.7090.7221. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 800 - Mode: Read OnlyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200200K400K600K800K1000KSE +/- 1154.98, N = 3SE +/- 5557.21, N = 3SE +/- 4300.20, N = 3SE +/- 14172.11, N = 3SE +/- 8757.92, N = 3431962385411390018112889311082431. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 500 - Mode: Read Only - Average LatencyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.21620.43240.64860.86481.081SE +/- 0.003, N = 3SE +/- 0.010, N = 3SE +/- 0.012, N = 3SE +/- 0.001, N = 3SE +/- 0.001, N = 30.8990.9400.9610.3600.3611. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 500 - Mode: Read OnlyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200300K600K900K1200K1500KSE +/- 1617.01, N = 3SE +/- 5499.26, N = 3SE +/- 6448.13, N = 3SE +/- 3056.15, N = 3SE +/- 3948.07, N = 3555873532015520502138815713841661. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 800 - Mode: Read Write - Average LatencyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200612182430SE +/- 0.06, N = 3SE +/- 0.26, N = 3SE +/- 0.09, N = 3SE +/- 0.05, N = 3SE +/- 0.16, N = 323.2024.7624.5016.1016.161. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 800 - Mode: Read WriteXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520011K22K33K44K55KSE +/- 85.58, N = 3SE +/- 337.25, N = 3SE +/- 123.75, N = 3SE +/- 155.76, N = 3SE +/- 481.81, N = 334486323163265749683495221. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

Pennant

Pennant is an application focused on hydrodynamics on general unstructured meshes in 2D. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: leblancbigXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001020304050SE +/- 1.09, N = 12SE +/- 0.82, N = 15SE +/- 0.88, N = 15SE +/- 0.33, N = 5SE +/- 0.27, N = 332.8743.8042.9430.2429.551. (CXX) g++ options: -fopenmp -lmpi_cxx -lmpi

DuckDB

DuckDB is an in-progress SQL OLAP database management system optimized for analytics and features a vectorized and parallel engine. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDuckDB 0.9.1Benchmark: IMDBXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.14, N = 3SE +/- 0.10, N = 3SE +/- 0.06, N = 3SE +/- 0.07, N = 3SE +/- 0.05, N = 399.58101.94100.6383.0182.381. (CXX) g++ options: -O3 -rdynamic -lssl -lcrypto -ldl

FFmpeg

This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 6.1Encoder: libx265 - Scenario: PlatformXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001632486480SE +/- 0.22, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.13, N = 3SE +/- 0.10, N = 358.7565.4565.2669.2469.801. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 6.1Encoder: libx265 - Scenario: Video On DemandXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001632486480SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.09, N = 3SE +/- 0.13, N = 3SE +/- 0.13, N = 358.7565.2765.3569.5269.811. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19, Long Mode - Decompression SpeedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200400800120016002000SE +/- 9.70, N = 3SE +/- 97.47, N = 3SE +/- 9.44, N = 3SE +/- 6.90, N = 15SE +/- 20.05, N = 31906.11836.41926.91890.51908.31. (CC) gcc options: -O3 -pthread -lz -llzma

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19, Long Mode - Compression SpeedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.00, N = 3SE +/- 0.18, N = 15SE +/- 0.09, N = 38.9110.3010.2011.5010.701. (CC) gcc options: -O3 -pthread -lz -llzma

Llamafile

Mozilla's Llamafile allows distributing and running large language models (LLMs) as a single file. Llamafile aims to make open-source LLMs more accessible to developers and users. Llamafile supports a variety of models, CPUs and GPUs, and other options. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.6Test: mistral-7b-instruct-v0.2.Q8_0 - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 37.916.346.286.536.79

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: ResNet-50Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001020304050SE +/- 1.58, N = 15SE +/- 0.35, N = 3SE +/- 0.16, N = 3SE +/- 0.27, N = 3SE +/- 0.94, N = 1532.2545.2045.1145.7643.38MIN: 23.14 / MAX: 40.93MIN: 29.05 / MAX: 46.24MIN: 43.61 / MAX: 45.86MIN: 34.75 / MAX: 46.5MIN: 34.95 / MAX: 46.71

FFmpeg

This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 6.1Encoder: libx265 - Scenario: UploadXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200816243240SE +/- 0.08, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.13, N = 3SE +/- 0.07, N = 328.8932.0331.9534.0634.211. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: particle_volume/ao/real_timeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.00839, N = 3SE +/- 0.05544, N = 4SE +/- 0.00129, N = 3SE +/- 0.00268, N = 3SE +/- 0.00600, N = 35.931614.583534.638288.810788.98399

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Junkshop - Compute: CPU-OnlyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 0.02, N = 3SE +/- 0.19, N = 3SE +/- 0.43, N = 3SE +/- 0.01, N = 3SE +/- 0.17, N = 3173.17139.94139.4873.7372.75

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19 - Decompression SpeedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200400800120016002000SE +/- 7.69, N = 3SE +/- 20.68, N = 3SE +/- 18.77, N = 3SE +/- 7.95, N = 15SE +/- 9.65, N = 31972.12026.02066.71998.32010.81. (CC) gcc options: -O3 -pthread -lz -llzma

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19 - Compression SpeedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200510152025SE +/- 0.03, N = 3SE +/- 0.15, N = 3SE +/- 0.03, N = 3SE +/- 0.23, N = 15SE +/- 0.03, N = 316.317.918.019.518.81. (CC) gcc options: -O3 -pthread -lz -llzma

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Layered HalfspaceXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 1.26, N = 3SE +/- 0.89, N = 3SE +/- 0.49, N = 3SE +/- 0.17, N = 3SE +/- 0.14, N = 3144.54130.53133.3186.0083.971. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: IS.DXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 16.64, N = 3SE +/- 18.32, N = 12SE +/- 16.77, N = 15SE +/- 10.32, N = 15SE +/- 4.80, N = 31604.851228.581269.371223.221278.441. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Water-layered HalfspaceXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.21, N = 3SE +/- 0.40, N = 3SE +/- 0.42, N = 3SE +/- 0.26, N = 3SE +/- 0.06, N = 3135.42129.55129.3387.8685.251. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: ResNet-50Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200918273645SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.04, N = 323.1930.5230.7938.9639.99

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Mount St. HelensXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001224364860SE +/- 0.34, N = 3SE +/- 0.33, N = 15SE +/- 0.25, N = 3SE +/- 0.33, N = 3SE +/- 0.29, N = 1553.2949.5451.2934.8033.131. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

Llamafile

Mozilla's Llamafile allows distributing and running large language models (LLMs) as a single file. Llamafile aims to make open-source LLMs more accessible to developers and users. Llamafile supports a variety of models, CPUs and GPUs, and other options. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.6Test: wizardcoder-python-34b-v1.0.Q6_K - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.4950.991.4851.982.475SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 32.201.841.821.942.01

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 8 - Decompression SpeedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005001000150020002500SE +/- 2.75, N = 3SE +/- 3.55, N = 3SE +/- 4.31, N = 3SE +/- 5.41, N = 9SE +/- 24.36, N = 92375.62313.42322.32279.92245.71. (CC) gcc options: -O3 -pthread -lz -llzma

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 8 - Compression SpeedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 3.15, N = 3SE +/- 1.37, N = 3SE +/- 1.88, N = 3SE +/- 6.44, N = 9SE +/- 6.27, N = 9485.6513.9520.3778.5793.81. (CC) gcc options: -O3 -pthread -lz -llzma

Intel Open Image Denoise

Open Image Denoise is a denoising library for ray-tracing and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.2Run: RTLightmap.hdr.4096x4096 - Device: CPU-OnlyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.09230.18460.27690.36920.4615SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.210.230.230.400.41

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Fishy Cat - Compute: CPU-OnlyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 0.24, N = 3SE +/- 0.38, N = 3SE +/- 0.39, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 3165.00127.12127.0167.3466.78

NAMD

NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterNAMD 3.0b6Input: STMV with 1,066,628 AtomsXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.17020.34040.51060.68080.851SE +/- 0.00016, N = 3SE +/- 0.00121, N = 3SE +/- 0.00059, N = 3SE +/- 0.00230, N = 3SE +/- 0.00230, N = 30.205650.490070.493450.740720.75646

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200612182430SE +/- 0.02, N = 3SE +/- 0.07, N = 3SE +/- 0.05, N = 3SE +/- 0.41, N = 15SE +/- 0.03, N = 323.6119.1019.2616.7818.661. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001326395265SE +/- 0.04, N = 3SE +/- 0.20, N = 3SE +/- 0.12, N = 3SE +/- 1.44, N = 15SE +/- 0.09, N = 342.3552.3451.9160.0653.581. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: bertsquad-12 - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.14, N = 3SE +/- 0.63, N = 3SE +/- 4.77, N = 15SE +/- 0.32, N = 3SE +/- 0.03, N = 391.5285.1270.8365.4866.031. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: bertsquad-12 - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.02, N = 3SE +/- 0.09, N = 3SE +/- 1.17, N = 15SE +/- 0.08, N = 3SE +/- 0.01, N = 310.9311.7515.2215.2715.141. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: T5 Encoder - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.07186, N = 15SE +/- 0.04051, N = 3SE +/- 0.03666, N = 3SE +/- 0.08219, N = 3SE +/- 0.08220, N = 35.907867.037027.167516.369846.161451. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: T5 Encoder - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 2.25, N = 15SE +/- 0.82, N = 3SE +/- 0.71, N = 3SE +/- 2.01, N = 3SE +/- 2.20, N = 3169.62142.06139.48157.01162.331. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

CockroachDB

CockroachDB is a cloud-native, distributed SQL database for data intensive applications. This test profile uses a server-less CockroachDB configuration to test various Coackroach workloads on the local host with a single node. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgops/s, More Is BetterCockroachDB 22.2Workload: KV, 60% Reads - Concurrency: 128Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020K40K60K80K100KSE +/- 198.29, N = 3SE +/- 204.39, N = 3SE +/- 47.09, N = 3SE +/- 137.27, N = 3SE +/- 156.58, N = 363021.467444.067506.793864.195084.8

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: AlexNetXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520090180270360450SE +/- 0.06, N = 3SE +/- 0.15, N = 3SE +/- 0.09, N = 3SE +/- 0.13, N = 3SE +/- 0.09, N = 3189.14237.76237.57405.64409.96

CockroachDB

CockroachDB is a cloud-native, distributed SQL database for data intensive applications. This test profile uses a server-less CockroachDB configuration to test various Coackroach workloads on the local host with a single node. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgops/s, More Is BetterCockroachDB 22.2Workload: KV, 50% Reads - Concurrency: 128Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020K40K60K80K100KSE +/- 189.35, N = 3SE +/- 103.24, N = 3SE +/- 264.18, N = 3SE +/- 764.38, N = 3SE +/- 831.25, N = 359866.664458.464052.790528.392195.2

OpenBenchmarking.orgops/s, More Is BetterCockroachDB 22.2Workload: KV, 95% Reads - Concurrency: 128Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030K60K90K120K150KSE +/- 237.27, N = 3SE +/- 258.97, N = 3SE +/- 302.28, N = 3SE +/- 716.09, N = 3SE +/- 765.52, N = 383987.584052.383155.8119084.9121061.7

Dragonflydb

Dragonfly is an open-source database server that is a "modern Redis replacement" that aims to be the fastest memory store while being compliant with the Redis and Memcached protocols. For benchmarking Dragonfly, Memtier_benchmark is used as a NoSQL Redis/Memcache traffic generation plus benchmarking tool developed by Redis Labs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterDragonflydb 1.6.2Clients Per Thread: 10 - Set To Get Ratio: 1:10Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001.4M2.8M4.2M5.6M7MSE +/- 44122.15, N = 3SE +/- 43642.69, N = 3SE +/- 38431.66, N = 3SE +/- 45878.41, N = 3SE +/- 90168.82, N = 124522560.584759359.714729191.296406922.866266297.871. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

ASKAP

ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMpix/sec, More Is BetterASKAP 1.0Test: tConvolve MPI - GriddingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520016003200480064008000SE +/- 48.43, N = 3SE +/- 0.00, N = 3SE +/- 180.65, N = 12SE +/- 298.76, N = 15SE +/- 298.02, N = 156199.093975.694256.287317.337477.191. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

OpenBenchmarking.orgMpix/sec, More Is BetterASKAP 1.0Test: tConvolve MPI - DegriddingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520014002800420056007000SE +/- 58.27, N = 3SE +/- 14.20, N = 3SE +/- 192.19, N = 12SE +/- 328.86, N = 15SE +/- 321.53, N = 155146.333349.853635.696613.596642.941. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

GROMACS

The GROMACS (GROningen MAchine for Chemical Simulations) molecular dynamics package testing with the water_GMX50 data. This test profile allows selecting between CPU and GPU-based GROMACS builds. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2024Implementation: MPI CPU - Input: water_GMX50_bareXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.57221.14441.71662.28882.861SE +/- 0.000, N = 3SE +/- 0.001, N = 3SE +/- 0.002, N = 3SE +/- 0.004, N = 3SE +/- 0.001, N = 31.4341.5261.5462.4682.5431. (CXX) g++ options: -O3 -lm

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 12 - Decompression SpeedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005001000150020002500SE +/- 2.01, N = 3SE +/- 7.37, N = 3SE +/- 10.81, N = 3SE +/- 11.41, N = 12SE +/- 25.03, N = 32377.02394.82404.22308.62314.71. (CC) gcc options: -O3 -pthread -lz -llzma

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 12 - Compression SpeedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050100150200250SE +/- 2.27, N = 3SE +/- 1.44, N = 3SE +/- 0.92, N = 3SE +/- 4.91, N = 12SE +/- 2.97, N = 3170.6168.4171.0207.5220.91. (CC) gcc options: -O3 -pthread -lz -llzma

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 0Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.49, N = 3SE +/- 0.12, N = 3SE +/- 0.18, N = 3SE +/- 0.08, N = 3SE +/- 0.33, N = 3133.57107.79108.6572.4272.111. (CXX) g++ options: -O3 -fPIC -lm

Numpy Benchmark

This is a test to obtain the general Numpy performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterNumpy BenchmarkXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 1.43, N = 3SE +/- 2.58, N = 3SE +/- 3.59, N = 3SE +/- 3.29, N = 3SE +/- 2.05, N = 3675.48812.16808.62800.78803.04

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: BT.CXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520010K20K30K40K50KSE +/- 96.46, N = 3SE +/- 21.68, N = 3SE +/- 14.25, N = 3SE +/- 553.44, N = 3SE +/- 110.88, N = 328326.3622551.0922935.2743371.7844844.131. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.11Video Input: Bosphorus 4K - Video Preset: FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.018, N = 3SE +/- 0.024, N = 3SE +/- 0.021, N = 3SE +/- 0.061, N = 3SE +/- 0.058, N = 34.9755.6815.6937.7467.8181. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: ArcFace ResNet-100 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001020304050SE +/- 0.21, N = 3SE +/- 0.69, N = 12SE +/- 0.15, N = 3SE +/- 0.07, N = 3SE +/- 0.04, N = 343.4036.2537.0027.3526.831. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: ArcFace ResNet-100 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200918273645SE +/- 0.11, N = 3SE +/- 0.60, N = 12SE +/- 0.11, N = 3SE +/- 0.09, N = 3SE +/- 0.05, N = 323.0427.7127.0236.5637.271. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Redis

Redis is an open-source in-memory data structure store, used as a database, cache, and message broker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 7.0.4Test: SET - Parallel Connections: 500Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200800K1600K2400K3200K4000KSE +/- 128040.11, N = 15SE +/- 181123.85, N = 15SE +/- 201571.91, N = 12SE +/- 72840.48, N = 12SE +/- 164667.68, N = 153327348.653656582.353684418.983827317.963419833.501. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3

PETSc

PETSc, the Portable, Extensible Toolkit for Scientific Computation, is for the scalable (parallel) solution of scientific applications modeled by partial differential equations. This test profile runs the PETSc "make streams" benchmark and records the throughput rate when all available cores are utilized for the MPI Streams build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterPETSc 3.19Test: StreamsXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520015K30K45K60K75KSE +/- 12.95, N = 3SE +/- 40.69, N = 3SE +/- 249.54, N = 3SE +/- 31.25, N = 3SE +/- 23.59, N = 369652.7339918.9726545.7139471.0741077.631. (CC) gcc options: -fPIC -O3 -O2 -lpthread -ludev -lpciaccess -lm

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Small Mesh Size - Execution TimeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520070140210280350242.94311.36296.60199.88183.231. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Small Mesh Size - Mesh TimeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200102030405035.1941.6838.1729.7430.771. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm

Dragonflydb

Dragonfly is an open-source database server that is a "modern Redis replacement" that aims to be the fastest memory store while being compliant with the Redis and Memcached protocols. For benchmarking Dragonfly, Memtier_benchmark is used as a NoSQL Redis/Memcache traffic generation plus benchmarking tool developed by Redis Labs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterDragonflydb 1.6.2Clients Per Thread: 10 - Set To Get Ratio: 1:100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001.5M3M4.5M6M7.5MSE +/- 43112.16, N = 3SE +/- 16892.77, N = 3SE +/- 6458.55, N = 3SE +/- 57648.71, N = 9SE +/- 34452.25, N = 34604951.364818585.254762998.427046043.066289782.081. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients/connections. HTTPS with a self-signed OpenSSL certificate is used by this test for local benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 1000Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030K60K90K120K150KSE +/- 46.60, N = 3SE +/- 52.64, N = 3SE +/- 98.21, N = 3SE +/- 536.64, N = 3SE +/- 301.50, N = 387132.2284772.3185990.12144355.05136360.721. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

Apache HTTP Server

This is a test of the Apache HTTPD web server. This Apache HTTPD web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 1000Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030K60K90K120K150KSE +/- 873.03, N = 3SE +/- 59.58, N = 3SE +/- 62.36, N = 3SE +/- 802.85, N = 3SE +/- 797.39, N = 3113097.8587686.6387708.39130637.70131437.321. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients/connections. HTTPS with a self-signed OpenSSL certificate is used by this test for local benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 500Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030K60K90K120K150KSE +/- 173.47, N = 3SE +/- 239.98, N = 3SE +/- 191.32, N = 3SE +/- 126.64, N = 3SE +/- 288.60, N = 392582.2991505.2494947.75156935.63154835.061. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

Apache HTTP Server

This is a test of the Apache HTTPD web server. This Apache HTTPD web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 500Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030K60K90K120K150KSE +/- 774.43, N = 3SE +/- 141.36, N = 3SE +/- 152.51, N = 3SE +/- 667.30, N = 3SE +/- 521.55, N = 3125745.6196630.2396508.30136714.05133534.111. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients/connections. HTTPS with a self-signed OpenSSL certificate is used by this test for local benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 200Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030K60K90K120K150KSE +/- 231.66, N = 3SE +/- 204.87, N = 3SE +/- 127.61, N = 3SE +/- 223.05, N = 3SE +/- 150.85, N = 398692.95100527.89101214.88162256.48162348.331. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Material TesterXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200173.41143.38142.94102.74102.15

MariaDB mariadb-slap

This is a MariaDB MySQL database server benchmark making use of mysqlslap (mariadb-slap). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB mariadb-slap 11.5Clients: 64Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 4.64, N = 3SE +/- 5.02, N = 3SE +/- 5.66, N = 3SE +/- 5.34, N = 3SE +/- 4.85, N = 311281132112411821183-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-lnuma -lpcre2-8 -lcrypt -laio -lssl -lcrypto-shared -lrt -lstdc++-shared -lrt -lstdc++-shared -lrt -lstdc++1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lz -lm -lpthread -ldl

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Disney MaterialXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200174.95156.37156.0785.9385.32

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: BMW27 - Compute: CPU-OnlyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.18, N = 3SE +/- 0.24, N = 3SE +/- 0.33, N = 3SE +/- 0.16, N = 3SE +/- 0.05, N = 3123.51101.52101.7253.5653.25

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: defconfigXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.45, N = 3SE +/- 0.32, N = 3SE +/- 0.32, N = 3SE +/- 0.42, N = 3SE +/- 0.36, N = 3108.3497.9297.4157.9357.47

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Genetic Algorithm Using Jenetics + FuturesXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 9.79, N = 3SE +/- 9.08, N = 3SE +/- 5.89, N = 3SE +/- 8.75, N = 3SE +/- 6.12, N = 15886.6819.5821.7887.0882.7MIN: 844.84 / MAX: 942.25MIN: 773.51 / MAX: 850.94MIN: 782.09 / MAX: 843.31MIN: 771.69 / MAX: 950.63MIN: 756.16 / MAX: 981

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Recurrent Neural Network Training - Engine: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52006001200180024003000SE +/- 9.27, N = 3SE +/- 3.90, N = 3SE +/- 1.64, N = 3SE +/- 3.18, N = 3SE +/- 1.43, N = 32931.952496.652491.141354.121348.37MIN: 2882.15MIN: 2449.87MIN: 2459.9MIN: 1317.27MIN: 1314.051. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

CloverLeaf

CloverLeaf is a Lagrangian-Eulerian hydrodynamics benchmark. This test profile currently makes use of CloverLeaf's OpenMP version. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCloverLeaf 1.3Input: clover_bmXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.27, N = 3SE +/- 0.09, N = 3SE +/- 0.21, N = 3SE +/- 0.09, N = 3SE +/- 0.45, N = 364.5889.3686.8275.5572.881. (F9X) gfortran options: -O3 -march=native -funroll-loops -fopenmp

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020K40K60K80K100KSE +/- 76.92, N = 3SE +/- 59.04, N = 3SE +/- 273.73, N = 3SE +/- 146.12, N = 3SE +/- 101.69, N = 39544784611842094456144247

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Recurrent Neural Network Inference - Engine: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 6.38, N = 3SE +/- 2.00, N = 3SE +/- 7.18, N = 3SE +/- 2.19, N = 3SE +/- 1.91, N = 31548.711329.681337.47703.75706.65MIN: 1512.56MIN: 1297.27MIN: 1299.81MIN: 675.7MIN: 679.481. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 1 - Model: ResNet-50Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001530456075SE +/- 1.37, N = 12SE +/- 1.71, N = 12SE +/- 1.84, N = 15SE +/- 0.29, N = 3SE +/- 0.90, N = 1561.4968.3067.4866.6365.69MIN: 40.06 / MAX: 72.22MIN: 48 / MAX: 73.52MIN: 46.76 / MAX: 74.15MIN: 57.04 / MAX: 69.34MIN: 53.12 / MAX: 69.76

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: GoogLeNetXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.03, N = 3SE +/- 0.07, N = 3SE +/- 0.04, N = 3SE +/- 0.17, N = 3SE +/- 0.07, N = 378.1791.4292.37129.59133.21

Chaos Group V-RAY

This is a test of Chaos Group's V-RAY benchmark. V-RAY is a commercial renderer that can integrate with various creator software products like SketchUp and 3ds Max. The V-RAY benchmark is standalone and supports CPU and NVIDIA CUDA/RTX based rendering. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgvsamples, More Is BetterChaos Group V-RAY 6.0Mode: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52009K18K27K36K45KSE +/- 16.84, N = 3SE +/- 57.13, N = 3SE +/- 11.00, N = 3SE +/- 38.48, N = 3SE +/- 204.38, N = 32006023056230784356743879

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: ThoroughXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200510152025SE +/- 0.32, N = 12SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 312.2310.3210.2119.6819.901. (CXX) g++ options: -O3 -flto -pthread

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52006001200180024003000SE +/- 0.88, N = 3SE +/- 1.20, N = 3SE +/- 4.04, N = 3SE +/- 1.53, N = 3SE +/- 0.33, N = 328842539252612751265

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: CG.CXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002K4K6K8K10KSE +/- 210.57, N = 15SE +/- 81.68, N = 15SE +/- 92.49, N = 15SE +/- 68.61, N = 9SE +/- 89.79, N = 69750.347857.537898.418784.839036.651. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

QMCPACK

QMCPACK is a modern high-performance open-source Quantum Monte Carlo (QMC) simulation code making use of MPI for this benchmark of the H20 example code. QMCPACK is an open-source production level many-body ab initio Quantum Monte Carlo code for computing the electronic structure of atoms, molecules, and solids. QMCPACK is supported by the U.S. Department of Energy. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Execution Time - Seconds, Fewer Is BetterQMCPACK 3.17.1Input: simple-H2OXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200612182430SE +/- 1.38, N = 12SE +/- 0.82, N = 15SE +/- 0.73, N = 12SE +/- 0.19, N = 3SE +/- 0.05, N = 324.0621.4523.1225.6225.251. (CXX) g++ options: -fopenmp -foffload=disable -finline-limit=1000 -fstrict-aliasing -funroll-all-loops -ffast-math -march=native -O3 -lm -ldl

ASKAP

ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 1.0Test: tConvolve MT - DegriddingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52007001400210028003500SE +/- 6.39, N = 3SE +/- 4.57, N = 3SE +/- 6.80, N = 3SE +/- 6.45, N = 3SE +/- 14.04, N = 33033.561791.991806.712291.852387.891. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

OpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 1.0Test: tConvolve MT - GriddingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005001000150020002500SE +/- 4.91, N = 3SE +/- 2.12, N = 3SE +/- 3.42, N = 3SE +/- 1.09, N = 3SE +/- 1.46, N = 32260.821550.641619.821917.671977.001. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200612182430SE +/- 0.18, N = 3SE +/- 0.16, N = 3SE +/- 0.17, N = 3SE +/- 0.19, N = 5SE +/- 0.07, N = 324.3919.3319.7919.3218.971. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001224364860SE +/- 0.30, N = 3SE +/- 0.43, N = 3SE +/- 0.43, N = 3SE +/- 0.52, N = 5SE +/- 0.21, N = 341.0151.7250.5351.7852.721. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Tomographic ModelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001224364860SE +/- 0.06, N = 3SE +/- 0.38, N = 3SE +/- 0.48, N = 3SE +/- 0.37, N = 3SE +/- 0.20, N = 1552.8946.5849.5830.4930.131. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 2 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005001000150020002500SE +/- 2.33, N = 3SE +/- 2.73, N = 3SE +/- 0.33, N = 3SE +/- 1.00, N = 3SE +/- 2.08, N = 324742171215610901080

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005001000150020002500SE +/- 0.88, N = 3SE +/- 2.19, N = 3SE +/- 1.76, N = 3SE +/- 1.15, N = 3SE +/- 1.00, N = 324412143213310781069

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Detection FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.73, N = 3SE +/- 0.11, N = 3SE +/- 0.15, N = 3SE +/- 1.00, N = 5SE +/- 1.01, N = 3133.9973.8772.5292.3889.98MIN: 103.21 / MAX: 233.72MIN: 36.14 / MAX: 99.53MIN: 28.47 / MAX: 87.6MIN: 31.8 / MAX: 128.67MIN: 63.39 / MAX: 124.531. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Detection FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.16, N = 3SE +/- 0.08, N = 3SE +/- 0.11, N = 3SE +/- 0.92, N = 5SE +/- 0.98, N = 329.8354.1155.1186.5688.841. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Stress-NG

Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: CPU CacheXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001.3M2.6M3.9M5.2M6.5MSE +/- 59300.10, N = 15SE +/- 44904.06, N = 10SE +/- 27236.78, N = 3SE +/- 29190.39, N = 3SE +/- 18366.37, N = 34023464.535854131.565997921.202537425.462446936.661. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 2 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020K40K60K80K100KSE +/- 34.91, N = 3SE +/- 128.55, N = 3SE +/- 152.10, N = 3SE +/- 126.32, N = 3SE +/- 61.02, N = 38240372906722503845438192

Llama.cpp

Llama.cpp is a port of Facebook's LLaMA model in C/C++ developed by Georgi Gerganov. Llama.cpp allows the inference of LLaMA and other supported models in C/C++. For CPU inference Llama.cpp supports AVX2/AVX-512, ARM NEON, and other modern ISAs along with features like OpenBLAS usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b1808Model: llama-2-13b.Q4_0.ggufXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 38.386.706.576.987.321. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020K40K60K80K100KSE +/- 135.20, N = 3SE +/- 135.51, N = 3SE +/- 145.56, N = 3SE +/- 192.07, N = 3SE +/- 168.15, N = 38123672067715303798437923

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_timeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.00822, N = 3SE +/- 0.00132, N = 3SE +/- 0.00236, N = 3SE +/- 0.00685, N = 3SE +/- 0.00353, N = 33.795394.653054.667578.921498.97813

Memcached

Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:10Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200800K1600K2400K3200K4000KSE +/- 7388.22, N = 3SE +/- 1626.82, N = 3SE +/- 2761.61, N = 3SE +/- 5916.60, N = 3SE +/- 5303.54, N = 33072373.872217988.242195688.493775389.093761626.611. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200800K1600K2400K3200K4000KSE +/- 11523.53, N = 3SE +/- 1526.11, N = 3SE +/- 1471.26, N = 3SE +/- 3096.61, N = 3SE +/- 5665.65, N = 33048853.562124192.992104490.553850014.663844031.281. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

Timed CPython Compilation

This test times how long it takes to build the reference Python implementation, CPython, with optimizations and LTO enabled for a release build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed CPython Compilation 3.10.6Build Configuration: Released Build, PGO + LTO OptimizedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200197.92202.77200.98193.59193.47

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: gravity_spheres_volume/dim_512/ao/real_timeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.00479, N = 3SE +/- 0.01340, N = 3SE +/- 0.00682, N = 3SE +/- 0.01707, N = 3SE +/- 0.00259, N = 32.734963.878273.910707.457827.46649

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: gravity_spheres_volume/dim_512/scivis/real_timeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.00622, N = 3SE +/- 0.01059, N = 3SE +/- 0.00260, N = 3SE +/- 0.00203, N = 3SE +/- 0.02548, N = 32.638513.823213.821387.319217.36909

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Orange Juice - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.04, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 33.584.424.427.707.72MIN: 3.44 / MAX: 4.42MIN: 4.35 / MAX: 4.49MIN: 4.32 / MAX: 4.49MIN: 6.73 / MAX: 8.14MIN: 6.87 / MAX: 8.15

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Face Detection FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 8.08, N = 3SE +/- 0.34, N = 3SE +/- 0.52, N = 3SE +/- 1.15, N = 3SE +/- 1.73, N = 31506.63556.14560.14592.59590.80MIN: 1180.01 / MAX: 1642.12MIN: 444.24 / MAX: 578.41MIN: 538 / MAX: 581.72MIN: 531.26 / MAX: 631.27MIN: 555.5 / MAX: 623.221. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Face Detection FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 32.657.197.1413.4613.521. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 8, Long Mode - Decompression SpeedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005001000150020002500SE +/- 5.21, N = 3SE +/- 7.92, N = 3SE +/- 7.19, N = 3SE +/- 3.89, N = 3SE +/- 6.24, N = 32375.42324.52352.52278.02299.71. (CC) gcc options: -O3 -pthread -lz -llzma

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 8, Long Mode - Compression SpeedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 1.12, N = 3SE +/- 0.67, N = 3SE +/- 0.95, N = 3SE +/- 3.67, N = 3SE +/- 3.49, N = 3504.1518.5527.9778.8797.01. (CC) gcc options: -O3 -pthread -lz -llzma

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: LuxCore Benchmark - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001.04632.09263.13894.18525.2315SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 31.692.412.414.624.65MIN: 0.51 / MAX: 2.04MIN: 0.83 / MAX: 2.84MIN: 0.82 / MAX: 2.85MIN: 2 / MAX: 5.24MIN: 2.03 / MAX: 5.27

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Face Detection FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520080160240320400SE +/- 0.59, N = 3SE +/- 0.44, N = 3SE +/- 0.15, N = 3SE +/- 0.18, N = 3SE +/- 0.24, N = 3384.09283.71285.97301.83301.79MIN: 285.45 / MAX: 543.49MIN: 240.12 / MAX: 303.33MIN: 150.08 / MAX: 303.22MIN: 264.12 / MAX: 322.86MIN: 278.87 / MAX: 327.81. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Face Detection FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200612182430SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 310.4114.0713.9726.4426.461. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: MediumXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 36.628.628.6115.7015.92

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: GPT-2 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.03060, N = 3SE +/- 0.00562, N = 3SE +/- 0.02822, N = 3SE +/- 0.01596, N = 3SE +/- 0.00809, N = 38.108648.267888.311078.406718.151551. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: GPT-2 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.46, N = 3SE +/- 0.08, N = 3SE +/- 0.41, N = 3SE +/- 0.22, N = 3SE +/- 0.12, N = 3123.28120.91120.28118.91122.631. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Xcompact3d Incompact3d

Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 129 Cells Per DirectionXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200510152025SE +/- 0.22, N = 3SE +/- 0.17, N = 15SE +/- 0.18, N = 9SE +/- 0.21, N = 3SE +/- 0.12, N = 1519.7522.5321.8717.1116.591. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: bertsquad-12 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.78, N = 3SE +/- 0.35, N = 3SE +/- 0.76, N = 3SE +/- 0.63, N = 3SE +/- 0.22, N = 386.6981.5779.7063.1164.281. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: bertsquad-12 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.10, N = 3SE +/- 0.05, N = 3SE +/- 0.12, N = 3SE +/- 0.16, N = 3SE +/- 0.05, N = 311.5412.2612.5515.8515.561. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: DLSC - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001.1252.253.3754.55.625SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 32.132.742.754.965.00MIN: 2.03 / MAX: 2.65MIN: 2.66 / MAX: 2.85MIN: 2.66 / MAX: 2.86MIN: 4.84 / MAX: 5.29MIN: 4.85 / MAX: 5.35

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: SharpenXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001326395265SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 336313057581. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Machine Translation EN To DE FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.08, N = 3SE +/- 0.15, N = 3SE +/- 0.10, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3118.4548.5548.2659.7758.43MIN: 68.57 / MAX: 244.71MIN: 29.96 / MAX: 68.41MIN: 29.16 / MAX: 69.56MIN: 40 / MAX: 88.32MIN: 31.85 / MAX: 94.551. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Machine Translation EN To DE FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.02, N = 3SE +/- 0.25, N = 3SE +/- 0.17, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 333.7482.3282.80133.65136.731. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: T5 Encoder - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.03755, N = 3SE +/- 0.09318, N = 3SE +/- 0.06732, N = 3SE +/- 0.01065, N = 3SE +/- 0.02831, N = 36.198767.003317.060976.576016.378061. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: T5 Encoder - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 0.97, N = 3SE +/- 1.92, N = 3SE +/- 1.35, N = 3SE +/- 0.25, N = 3SE +/- 0.69, N = 3161.31142.82141.63152.04156.771. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Detection FP32 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.77, N = 3SE +/- 0.12, N = 3SE +/- 0.13, N = 3SE +/- 0.63, N = 3SE +/- 0.48, N = 3134.5473.7172.7091.3490.49MIN: 98.92 / MAX: 233.45MIN: 31.23 / MAX: 89.7MIN: 40.38 / MAX: 84.07MIN: 68.51 / MAX: 134.9MIN: 55.81 / MAX: 127.951. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Detection FP32 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.17, N = 3SE +/- 0.09, N = 3SE +/- 0.09, N = 3SE +/- 0.61, N = 3SE +/- 0.47, N = 329.7154.2354.9887.5088.321. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Noise Suppression Poconet-Like FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 315.218.037.9510.189.90MIN: 10.08 / MAX: 64.69MIN: 5.57 / MAX: 23.51MIN: 5 / MAX: 23.85MIN: 5.82 / MAX: 38.39MIN: 6.53 / MAX: 37.691. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Noise Suppression Poconet-Like FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 1.17, N = 3SE +/- 0.72, N = 3SE +/- 0.86, N = 3SE +/- 1.94, N = 3SE +/- 1.23, N = 3523.02980.84991.321544.551587.741. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200816243240SE +/- 0.05, N = 3SE +/- 0.06, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 3SE +/- 0.07, N = 336.3514.5914.4817.2216.73MIN: 24.38 / MAX: 96.4MIN: 9.42 / MAX: 26.08MIN: 8.91 / MAX: 26.44MIN: 10.01 / MAX: 28.19MIN: 9.66 / MAX: 30.961. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200100200300400500SE +/- 0.28, N = 3SE +/- 1.13, N = 3SE +/- 0.50, N = 3SE +/- 1.58, N = 3SE +/- 1.94, N = 3219.86273.50275.72462.96476.631. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: EnhancedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 349505093941. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: Noise-GaussianXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 37076761161191. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Vehicle Bike Detection FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 39.944.784.785.135.12MIN: 6.2 / MAX: 41.65MIN: 3.25 / MAX: 16.12MIN: 3.18 / MAX: 16.31MIN: 3.4 / MAX: 14.3MIN: 3.61 / MAX: 15.431. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Vehicle Bike Detection FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 0.87, N = 3SE +/- 2.43, N = 3SE +/- 0.80, N = 3SE +/- 4.59, N = 3SE +/- 2.52, N = 3401.42834.18833.731551.751556.071. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200918273645SE +/- 0.09, N = 3SE +/- 0.10, N = 3SE +/- 0.09, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 339.4818.0217.7926.3825.15MIN: 22.02 / MAX: 90.7MIN: 11.1 / MAX: 33.13MIN: 9.38 / MAX: 30.36MIN: 13.08 / MAX: 46.13MIN: 11.63 / MAX: 50.51. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520070140210280350SE +/- 0.22, N = 3SE +/- 1.23, N = 3SE +/- 1.10, N = 3SE +/- 0.60, N = 3SE +/- 0.46, N = 3101.21221.52224.43302.45317.251. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Re-Identification Retail FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.11, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 318.223.853.874.114.11MIN: 10.36 / MAX: 177.58MIN: 2.35 / MAX: 15.38MIN: 2.38 / MAX: 13.89MIN: 2.52 / MAX: 16.5MIN: 2.45 / MAX: 41.271. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Re-Identification Retail FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200400800120016002000SE +/- 2.58, N = 3SE +/- 0.54, N = 3SE +/- 0.99, N = 3SE +/- 0.55, N = 3SE +/- 0.22, N = 3437.961036.331029.881938.681938.321. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520010K20K30K40K50KSE +/- 115.84, N = 3SE +/- 61.68, N = 3SE +/- 62.68, N = 3SE +/- 13.61, N = 3SE +/- 29.87, N = 34892944013438302046120283

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Handwritten English Recognition FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001020304050SE +/- 0.12, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 345.7018.9019.1320.1120.09MIN: 32.92 / MAX: 117.92MIN: 13.12 / MAX: 33.09MIN: 12.57 / MAX: 34MIN: 14.53 / MAX: 37.6MIN: 13.66 / MAX: 38.021. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Handwritten English Recognition FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 0.45, N = 3SE +/- 0.13, N = 3SE +/- 0.59, N = 3SE +/- 0.92, N = 3SE +/- 1.22, N = 3174.92422.62417.64794.33795.051. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random FillXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200200K400K600K800K1000KSE +/- 5918.01, N = 3SE +/- 738.85, N = 3SE +/- 1877.58, N = 3SE +/- 826.71, N = 3SE +/- 1278.43, N = 3876130101178810059738698168667591. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.05, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 315.954.524.534.834.82MIN: 9.85 / MAX: 70.28MIN: 2.64 / MAX: 15.69MIN: 2.65 / MAX: 17.47MIN: 2.79 / MAX: 19.86MIN: 2.78 / MAX: 17.831. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200400800120016002000SE +/- 1.52, N = 3SE +/- 0.30, N = 3SE +/- 0.43, N = 3SE +/- 0.97, N = 3SE +/- 1.19, N = 3500.86882.75879.501648.321653.691. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Handwritten English Recognition FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001224364860SE +/- 0.12, N = 3SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.06, N = 3SE +/- 0.09, N = 353.0720.3420.3721.9921.89MIN: 35.4 / MAX: 130.32MIN: 13.65 / MAX: 36.45MIN: 14.74 / MAX: 35.54MIN: 14.99 / MAX: 41.87MIN: 14.64 / MAX: 40.371. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Handwritten English Recognition FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200160320480640800SE +/- 0.33, N = 3SE +/- 1.21, N = 3SE +/- 1.00, N = 3SE +/- 1.88, N = 3SE +/- 3.18, N = 3150.64392.93392.23726.61729.841. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Face Detection Retail FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001.14082.28163.42244.56325.704SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 35.073.133.163.323.33MIN: 2.36 / MAX: 27.99MIN: 1.79 / MAX: 14.78MIN: 1.77 / MAX: 13.91MIN: 1.82 / MAX: 16.13MIN: 1.88 / MAX: 15.671. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Face Detection Retail FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520010002000300040005000SE +/- 3.53, N = 3SE +/- 5.39, N = 3SE +/- 1.63, N = 3SE +/- 5.85, N = 3SE +/- 1.99, N = 31559.062522.562495.554743.734734.271. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random Fill SyncXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050K100K150K200K250KSE +/- 573.89, N = 3SE +/- 257.69, N = 3SE +/- 468.76, N = 3SE +/- 490.37, N = 3SE +/- 238.71, N = 32256071288551279551703651716231. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.090.180.270.360.45SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.400.280.280.290.29MIN: 0.18 / MAX: 8.35MIN: 0.15 / MAX: 7.24MIN: 0.15 / MAX: 7.71MIN: 0.16 / MAX: 9.68MIN: 0.16 / MAX: 8.631. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520011K22K33K44K55KSE +/- 38.88, N = 3SE +/- 67.36, N = 3SE +/- 19.90, N = 3SE +/- 49.45, N = 3SE +/- 47.80, N = 318507.3126465.3326230.8549580.0749485.351. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.06, N = 3SE +/- 0.04, N = 316.048.578.4411.7211.36MIN: 9.49 / MAX: 56.98MIN: 3.77 / MAX: 23.35MIN: 4.55 / MAX: 19.23MIN: 4.13 / MAX: 31.49MIN: 4.12 / MAX: 36.811. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200150300450600750SE +/- 0.14, N = 3SE +/- 0.22, N = 3SE +/- 1.08, N = 3SE +/- 3.49, N = 3SE +/- 2.53, N = 3249.02465.15472.57680.56701.841. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Weld Porosity Detection FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200714212835SE +/- 0.06, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 330.0510.9711.0511.7511.72MIN: 17.38 / MAX: 87.99MIN: 5.79 / MAX: 25.09MIN: 5.83 / MAX: 23.56MIN: 5.98 / MAX: 26.57MIN: 6 / MAX: 26.831. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Weld Porosity Detection FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 0.54, N = 3SE +/- 0.11, N = 3SE +/- 0.13, N = 3SE +/- 1.41, N = 3SE +/- 1.37, N = 3265.83728.18723.131359.291362.001. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Update RandomXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200200K400K600K800K1000KSE +/- 1460.51, N = 3SE +/- 819.19, N = 3SE +/- 1312.04, N = 3SE +/- 726.71, N = 3SE +/- 1173.38, N = 38281845560405537916247436233661. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Age Gender Recognition Retail 0013 FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.2250.450.6750.91.125SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 31.000.400.400.410.41MIN: 0.4 / MAX: 10.73MIN: 0.2 / MAX: 8.52MIN: 0.2 / MAX: 7.69MIN: 0.21 / MAX: 10.47MIN: 0.21 / MAX: 10.021. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Age Gender Recognition Retail 0013 FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52007K14K21K28K35KSE +/- 31.71, N = 3SE +/- 3.66, N = 3SE +/- 8.32, N = 3SE +/- 30.10, N = 3SE +/- 16.51, N = 37499.4718754.5918615.1234850.4734815.391. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Weld Porosity Detection FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 37.785.595.635.965.96MIN: 3.64 / MAX: 44.07MIN: 2.93 / MAX: 16.72MIN: 2.98 / MAX: 16.09MIN: 3.09 / MAX: 18.13MIN: 3.08 / MAX: 18.521. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Weld Porosity Detection FP16-INT8 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52006001200180024003000SE +/- 2.31, N = 3SE +/- 0.44, N = 3SE +/- 0.60, N = 3SE +/- 3.06, N = 3SE +/- 4.11, N = 31022.101427.321417.332669.642672.511. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: CaffeNet 12-int8 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.40350.8071.21051.6142.0175SE +/- 0.01234, N = 3SE +/- 0.00719, N = 3SE +/- 0.01555, N = 3SE +/- 0.00633, N = 3SE +/- 0.01197, N = 31.793471.467701.475311.198551.205041. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: CaffeNet 12-int8 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 3.86, N = 3SE +/- 3.38, N = 3SE +/- 7.12, N = 3SE +/- 4.40, N = 3SE +/- 8.14, N = 3557.41680.95677.58833.52829.191. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: CaffeNet 12-int8 - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.38270.76541.14811.53081.9135SE +/- 0.01304, N = 3SE +/- 0.00355, N = 3SE +/- 0.00807, N = 3SE +/- 0.00557, N = 3SE +/- 0.00549, N = 31.701071.481911.477381.117121.115771. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: CaffeNet 12-int8 - Device: CPU - Executor: StandardXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 4.53, N = 3SE +/- 1.59, N = 3SE +/- 3.70, N = 3SE +/- 4.40, N = 3SE +/- 4.41, N = 3587.73674.51676.65894.66895.771. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Face Detection Retail FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 39.762.202.212.372.37MIN: 4.34 / MAX: 46.89MIN: 1.23 / MAX: 12.5MIN: 1.23 / MAX: 13.12MIN: 1.26 / MAX: 11.94MIN: 1.27 / MAX: 12.751. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Face Detection Retail FP16 - Device: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52007001400210028003500SE +/- 1.96, N = 3SE +/- 1.03, N = 3SE +/- 0.66, N = 3SE +/- 6.99, N = 3SE +/- 6.25, N = 3816.171791.061781.623283.983269.121. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: MD5Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200900K1800K2700K3600K4500KSE +/- 2504.83, N = 3SE +/- 10837.18, N = 3SE +/- 7423.69, N = 3SE +/- 19341.95, N = 3SE +/- 11590.23, N = 391979122303332234333406666741180001. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: SwirlXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050100150200250SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 31301381392462501. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: RotateXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520070140210280350SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 1.53, N = 33171691681611671. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read Random Write RandomXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200600K1200K1800K2400K3000KSE +/- 4914.43, N = 3SE +/- 1481.36, N = 3SE +/- 1406.43, N = 3SE +/- 5448.80, N = 3SE +/- 2327.92, N = 3212091521445752138776277089927701951. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: HWB Color SpaceXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520070140210280350SE +/- 0.67, N = 3SE +/- 0.88, N = 3SE +/- 0.33, N = 3SE +/- 0.88, N = 3SE +/- 0.58, N = 33152442552692781. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read While WritingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001.1M2.2M3.3M4.4M5.5MSE +/- 20209.24, N = 3SE +/- 13925.00, N = 3SE +/- 14207.26, N = 3SE +/- 22040.69, N = 3SE +/- 26941.48, N = 3254883732122763194971514618751198911. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: ResizingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520070140210280350SE +/- 1.86, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.88, N = 3SE +/- 1.00, N = 32442512553323381. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Random FillXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200400K800K1200K1600K2000KSE +/- 10692.91, N = 3SE +/- 1270.91, N = 3SE +/- 4512.41, N = 3SE +/- 2055.67, N = 3SE +/- 3781.08, N = 3164798811547161159865122925612259211. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: ResNet50 v1-12-int8 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001.04862.09723.14584.19445.243SE +/- 0.03176, N = 3SE +/- 0.01507, N = 3SE +/- 0.00856, N = 3SE +/- 0.00459, N = 3SE +/- 0.02511, N = 34.660403.074413.099052.587042.625241. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: ResNet50 v1-12-int8 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520080160240320400SE +/- 1.45, N = 3SE +/- 1.58, N = 3SE +/- 0.89, N = 3SE +/- 0.69, N = 3SE +/- 3.68, N = 3214.55325.21322.61386.41380.821. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Random Fill SyncXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520040K80K120K160K200KSE +/- 710.34, N = 3SE +/- 196.51, N = 3SE +/- 48.85, N = 3SE +/- 414.30, N = 3SE +/- 270.54, N = 320159498054997461707351711991. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: OverwriteXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200400K800K1200K1600K2000KSE +/- 14969.35, N = 3SE +/- 1933.24, N = 3SE +/- 1465.82, N = 3SE +/- 1384.95, N = 3SE +/- 1555.52, N = 3165348911551211152820122974712253611. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Update RandomXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200200K400K600K800K1000KSE +/- 2978.74, N = 3SE +/- 1914.75, N = 3SE +/- 748.73, N = 3SE +/- 818.39, N = 3SE +/- 1578.78, N = 39083006095106181877032947004161. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random ReadXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030M60M90M120M150MSE +/- 170759.02, N = 3SE +/- 103627.09, N = 3SE +/- 34435.41, N = 3SE +/- 185772.57, N = 3SE +/- 286737.26, N = 35782442672087947716486831356034931349716261. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Read While WritingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200900K1800K2700K3600K4500KSE +/- 3646.67, N = 3SE +/- 15255.97, N = 3SE +/- 22460.23, N = 3SE +/- 13356.72, N = 3SE +/- 7045.87, N = 3194276024764412519594436839742946311. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Read Random Write RandomXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200600K1200K1800K2400K3000KSE +/- 17144.87, N = 3SE +/- 2242.91, N = 3SE +/- 5891.62, N = 3SE +/- 24632.94, N = 3SE +/- 24549.22, N = 3197653321378922131122291274729157581. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: super-resolution-10 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.09981, N = 3SE +/- 0.14350, N = 3SE +/- 0.02135, N = 3SE +/- 0.02431, N = 3SE +/- 0.06152, N = 313.1792010.7379010.918207.558407.500801. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: super-resolution-10 - Device: CPU - Executor: ParallelXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.57, N = 3SE +/- 1.26, N = 3SE +/- 0.18, N = 3SE +/- 0.42, N = 3SE +/- 1.10, N = 375.8893.1691.59132.29133.321. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: HMAC-SHA512Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520040M80M120M160M200MSE +/- 518797.01, N = 3SE +/- 719862.56, N = 3SE +/- 411753.30, N = 3SE +/- 877445.98, N = 3SE +/- 580393.26, N = 3407210001170656671167703331782833331813310001. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Random ReadXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030M60M90M120M150MSE +/- 236765.68, N = 3SE +/- 112684.33, N = 3SE +/- 144351.01, N = 3SE +/- 342858.26, N = 3SE +/- 780691.20, N = 34751226069366631694082541312186391306930631. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

x265

This is a simple test of the x265 encoder run on the CPU with 1080p and 4K options for H.265 video encode performance with x265. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 4KXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200816243240SE +/- 0.17, N = 3SE +/- 0.21, N = 4SE +/- 0.15, N = 15SE +/- 0.10, N = 3SE +/- 0.08, N = 319.4817.7317.6134.1334.651. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Homogeneous HalfspaceXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001530456075SE +/- 0.41, N = 3SE +/- 0.66, N = 3SE +/- 0.35, N = 3SE +/- 0.10, N = 3SE +/- 0.28, N = 367.1157.7963.8739.9239.561. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 1 - Buffer Length: 256 - Filter Length: 512Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004M8M12M16M20MSE +/- 13650.40, N = 3SE +/- 17835.67, N = 3SE +/- 250051.55, N = 3SE +/- 154682.83, N = 3SE +/- 410224.76, N = 1520255000201066671961166719299333189644001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Intel Open Image Denoise

Open Image Denoise is a denoising library for ray-tracing and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.2Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-OnlyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.19350.3870.58050.7740.9675SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.420.480.480.850.86

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.2Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-OnlyXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.19350.3870.58050.7740.9675SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.420.480.480.850.86

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 2 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52009K18K27K36K45KSE +/- 172.31, N = 3SE +/- 82.83, N = 3SE +/- 127.68, N = 3SE +/- 24.89, N = 3SE +/- 19.75, N = 34227538010379001749117285

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52009K18K27K36K45KSE +/- 112.38, N = 3SE +/- 115.78, N = 3SE +/- 43.80, N = 3SE +/- 11.15, N = 3SE +/- 33.74, N = 34151437677374581724717116

Apache Hadoop

This is a benchmark of the Apache Hadoop making use of its built-in name-node throughput benchmark (NNThroughputBenchmark). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps per sec, More Is BetterApache Hadoop 3.3.6Operation: Create - Threads: 100 - Files: 100000Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020K40K60K80K100KSE +/- 1027.34, N = 3SE +/- 496.20, N = 3SE +/- 670.20, N = 11SE +/- 292.68, N = 3SE +/- 376.90, N = 38899290476909387494675381

Node.js V8 Web Tooling Benchmark

Running the V8 project's Web-Tooling-Benchmark under Node.js. The Web-Tooling-Benchmark stresses JavaScript-related workloads common to web developers like Babel and TypeScript and Babylon. This test profile can test the system's JavaScript performance with Node.js. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgruns/s, More Is BetterNode.js V8 Web Tooling BenchmarkXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200612182430SE +/- 0.15, N = 3SE +/- 0.04, N = 3SE +/- 0.12, N = 3SE +/- 0.11, N = 3SE +/- 0.12, N = 323.3021.8922.1420.0620.81

Z3 Theorem Prover

The Z3 Theorem Prover / SMT solver is developed by Microsoft Research under the MIT license. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterZ3 Theorem Prover 4.12.1SMT File: 2.smt2Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001428425670SE +/- 0.51, N = 3SE +/- 0.21, N = 3SE +/- 0.27, N = 3SE +/- 0.35, N = 3SE +/- 0.53, N = 363.0848.1248.6849.0048.881. (CXX) g++ options: -lpthread -std=c++17 -fvisibility=hidden -mfpmath=sse -msse -msse2 -O3 -fPIC

Timed PHP Compilation

This test times how long it takes to build PHP. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed PHP Compilation 8.3.4Time To CompileXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001224364860SE +/- 0.20, N = 3SE +/- 0.14, N = 3SE +/- 0.10, N = 3SE +/- 0.25, N = 3SE +/- 0.18, N = 354.4553.0953.7640.8540.11

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 2 - Buffer Length: 256 - Filter Length: 512Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52009M18M27M36M45MSE +/- 14666.67, N = 3SE +/- 267690.74, N = 3SE +/- 368485.05, N = 12SE +/- 184763.21, N = 3SE +/- 254700.04, N = 340821333396930003909433338605667387126671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: Myriad-GroestlXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002K4K6K8K10KSE +/- 30.23, N = 3SE +/- 0.86, N = 3SE +/- 6.11, N = 3SE +/- 195.61, N = 12SE +/- 13.33, N = 34843.726023.385972.0911228.1311337.001. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 2Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001428425670SE +/- 0.18, N = 3SE +/- 0.14, N = 3SE +/- 0.13, N = 3SE +/- 0.11, N = 3SE +/- 0.03, N = 361.8451.3851.9436.3436.331. (CXX) g++ options: -O3 -fPIC -lm

m-queens

A solver for the N-queens problem with multi-threading support via the OpenMP library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterm-queens 1.2Time To SolveXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001632486480SE +/- 0.62, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 372.8553.0853.5027.8227.911. (CXX) g++ options: -fopenmp -O2 -march=native

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer ISPC - Model: Asian Dragon ObjXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200612182430SE +/- 0.10, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 314.5515.5715.6926.8827.39MIN: 14.14 / MAX: 16.68MIN: 15.43 / MAX: 15.99MIN: 15.56 / MAX: 16.03MIN: 26.58 / MAX: 27.83MIN: 27.11 / MAX: 28.25

Coremark

This is a test of EEMBC CoreMark processor benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations/Sec, More Is BetterCoremark 1.0CoreMark Size 666 - Iterations Per SecondXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200200K400K600K800K1000KSE +/- 3256.58, N = 15SE +/- 1288.71, N = 3SE +/- 1076.82, N = 3SE +/- 2950.54, N = 3SE +/- 9073.63, N = 3430017.46505633.42499103.50954988.73951674.511. (CC) gcc options: -O2 -lrt" -lrt

RawTherapee

RawTherapee is a cross-platform, open-source multi-threaded RAW image processing program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRawTherapeeTotal Benchmark TimeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001122334455SE +/- 0.02, N = 3SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.07, N = 3SE +/- 0.03, N = 342.6646.7346.6642.7942.551. RawTherapee, version 5.10, command line.

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 1 - Input: Bosphorus 1080pXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200612182430SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 39.1612.3112.2822.8623.101. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: MediumXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200510152025SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 39.2812.4812.4921.8022.021. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Savina Reactors.IOXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52008001600240032004000SE +/- 16.31, N = 3SE +/- 27.86, N = 3SE +/- 28.17, N = 3SE +/- 33.81, N = 3SE +/- 15.03, N = 33695.43036.53042.03270.63265.2MIN: 3664.8 / MAX: 5363.87MIN: 2985.79 / MAX: 4135.64MIN: 3011.87 / MAX: 4321.7MIN: 3216.47 / MAX: 4638.43MIN: 3236.75 / MAX: 4741.82

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.11Video Input: Bosphorus 4K - Video Preset: FasterXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 311.7713.0813.1417.3217.441. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

JPEG-XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG-XL libjxl 0.10.1Input: JPEG - Quality: 90Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001122334455SE +/- 0.09, N = 3SE +/- 0.20, N = 3SE +/- 0.30, N = 3SE +/- 0.27, N = 3SE +/- 0.37, N = 1032.5136.3237.1846.7748.591. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 8 - Buffer Length: 256 - Filter Length: 512Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030M60M90M120M150MSE +/- 1116779.09, N = 8SE +/- 368435.13, N = 3SE +/- 612490.36, N = 3SE +/- 1579485.68, N = 4SE +/- 292137.71, N = 31300025001519066671505766671479350001499033331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

FFmpeg

This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 6.1Encoder: libx265 - Scenario: LiveXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 1.39, N = 3SE +/- 0.16, N = 3SE +/- 0.17, N = 3SE +/- 0.58, N = 3SE +/- 0.71, N = 3178.15179.98179.99186.71186.421. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

Llama.cpp

Llama.cpp is a port of Facebook's LLaMA model in C/C++ developed by Georgi Gerganov. Llama.cpp allows the inference of LLaMA and other supported models in C/C++. For CPU inference Llama.cpp supports AVX2/AVX-512, ARM NEON, and other modern ISAs along with features like OpenBLAS usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b1808Model: llama-2-7b.Q4_0.ggufXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.18, N = 4SE +/- 0.12, N = 6SE +/- 0.11, N = 3SE +/- 0.02, N = 3SE +/- 0.12, N = 316.0712.7312.5413.1213.821. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas

Java SciMark

This test runs the Java version of SciMark 2, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.2Computational Test: CompositeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520012002400360048006000SE +/- 52.94, N = 3SE +/- 56.78, N = 3SE +/- 42.37, N = 3SE +/- 46.58, N = 8SE +/- 35.75, N = 34311.805519.555476.915321.505441.20

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Finagle HTTP RequestsXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005001000150020002500SE +/- 25.68, N = 12SE +/- 7.85, N = 3SE +/- 22.75, N = 3SE +/- 14.49, N = 3SE +/- 15.93, N = 31668.71925.01963.82288.12300.7MIN: 1155.42 / MAX: 1849.98MIN: 1772.86 / MAX: 2146.59MIN: 1776.88 / MAX: 2143.78MIN: 2126.73 / MAX: 2373.63MIN: 2117.57 / MAX: 2419.92

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer ISPC - Model: CrownXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200714212835SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.09, N = 3SE +/- 0.02, N = 313.3916.3316.4630.7931.37MIN: 13.27 / MAX: 16.49MIN: 16.12 / MAX: 16.74MIN: 16.24 / MAX: 16.97MIN: 30.28 / MAX: 31.99MIN: 30.96 / MAX: 32.3

QuantLib

QuantLib is an open-source library/framework around quantitative finance for modeling, trading and risk management scenarios. QuantLib is written in C++ with Boost and its built-in benchmark used reports the QuantLib Benchmark Index benchmark score. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMFLOPS, More Is BetterQuantLib 1.32Configuration: Multi-ThreadedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020K40K60K80K100KSE +/- 294.52, N = 3SE +/- 56.56, N = 3SE +/- 49.82, N = 3SE +/- 103.02, N = 3SE +/- 169.22, N = 337214.444058.243490.783133.583365.11. (CXX) g++ options: -O3 -march=native -fPIE -pie

NAMD

NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterNAMD 3.0b6Input: ATPase with 327,506 AtomsXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.62741.25481.88222.50963.137SE +/- 0.00164, N = 3SE +/- 0.00243, N = 3SE +/- 0.00212, N = 3SE +/- 0.01282, N = 3SE +/- 0.03311, N = 40.659771.734541.737642.749992.78846

Timed FFmpeg Compilation

This test times how long it takes to build the FFmpeg multimedia library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed FFmpeg Compilation 6.1Time To CompileXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001122334455SE +/- 0.33, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.10, N = 3SE +/- 0.12, N = 350.0542.4742.2025.5425.30

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: GoogLeNetXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.10, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.12, N = 3SE +/- 0.14, N = 378.7992.8293.71131.79135.30

miniFE

MiniFE Finite Element is an application for unstructured implicit finite element codes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgCG Mflops, More Is BetterminiFE 2.2Problem Size: SmallXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002K4K6K8K10KSE +/- 8.52, N = 3SE +/- 7.52, N = 3SE +/- 3.31, N = 3SE +/- 16.40, N = 3SE +/- 8.13, N = 39869.747278.877325.107330.077673.661. (CXX) g++ options: -O3 -fopenmp -lmpi_cxx -lmpi

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer ISPC - Model: Asian DragonXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200714212835SE +/- 0.12, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 316.5818.3518.4531.5832.22MIN: 16.29 / MAX: 19.34MIN: 18.21 / MAX: 18.8MIN: 18.3 / MAX: 18.87MIN: 31.27 / MAX: 32.45MIN: 31.88 / MAX: 33.23

Apache Hadoop

This is a benchmark of the Apache Hadoop making use of its built-in name-node throughput benchmark (NNThroughputBenchmark). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps per sec, More Is BetterApache Hadoop 3.3.6Operation: Create - Threads: 500 - Files: 100000Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520017K34K51K68K85KSE +/- 717.72, N = 3SE +/- 761.92, N = 3SE +/- 465.56, N = 3SE +/- 488.73, N = 3SE +/- 682.82, N = 37695676371775857108070436

Timed Wasmer Compilation

This test times how long it takes to compile Wasmer. Wasmer is written in the Rust programming language and is a WebAssembly runtime implementation that supports WASI and EmScripten. This test profile builds Wasmer with the Cranelift and Singlepast compiler features enabled. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Wasmer Compilation 2.3Time To CompileXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200918273645SE +/- 0.23, N = 3SE +/- 0.18, N = 3SE +/- 0.23, N = 3SE +/- 0.10, N = 3SE +/- 0.13, N = 338.1336.5536.9029.8929.441. (CC) gcc options: -m64 -ldl -lgcc_s -lutil -lrt -lpthread -lm -lc -pie -nodefaultlibs

Stress-NG

Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: System V Message PassingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004M8M12M16M20MSE +/- 185878.60, N = 5SE +/- 6257.32, N = 3SE +/- 7976.77, N = 3SE +/- 20135.83, N = 3SE +/- 18408.29, N = 317858353.8012527831.3612445240.7616207347.1216150369.041. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: RingcoinXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52008001600240032004000SE +/- 3.73, N = 3SE +/- 11.68, N = 3SE +/- 4.79, N = 3SE +/- 37.02, N = 5SE +/- 20.59, N = 31437.471868.891851.253541.503500.131. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: EclipseXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520015003000450060007500SE +/- 61.34, N = 8SE +/- 12.50, N = 3SE +/- 37.34, N = 3SE +/- 28.45, N = 3SE +/- 26.61, N = 369436483641066276587

miniBUDE

MiniBUDE is a mini application for the the core computation of the Bristol University Docking Engine (BUDE). This test profile currently makes use of the OpenMP implementation of miniBUDE for CPU benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001122334455SE +/- 0.06, N = 3SE +/- 0.09, N = 3SE +/- 0.10, N = 3SE +/- 0.17, N = 4SE +/- 0.03, N = 412.4526.3726.2949.8149.951. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

OpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 1.56, N = 3SE +/- 2.35, N = 3SE +/- 2.56, N = 3SE +/- 4.15, N = 4SE +/- 0.84, N = 4311.31659.13657.351245.301248.811. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: Visual Quality Optimized - Input: Bosphorus 4KXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.47, N = 15SE +/- 0.50, N = 5SE +/- 0.56, N = 5SE +/- 1.37, N = 15SE +/- 1.42, N = 1559.5661.4561.9097.12100.001. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: AlexNetXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520070140210280350SE +/- 0.21, N = 3SE +/- 0.21, N = 3SE +/- 0.21, N = 3SE +/- 0.14, N = 3SE +/- 0.34, N = 3173.78205.32206.40316.11322.53

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: MagiXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200140280420560700SE +/- 0.39, N = 3SE +/- 0.10, N = 3SE +/- 0.30, N = 3SE +/- 7.55, N = 4SE +/- 0.26, N = 3278.09332.40330.62644.74633.441. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

JPEG-XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG-XL libjxl 0.10.1Input: PNG - Quality: 90Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001122334455SE +/- 0.19, N = 3SE +/- 0.21, N = 3SE +/- 0.36, N = 3SE +/- 0.68, N = 3SE +/- 0.28, N = 331.8136.0436.0247.2348.531. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.11Video Input: Bosphorus 1080p - Video Preset: FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200612182430SE +/- 0.06, N = 3SE +/- 0.06, N = 3SE +/- 0.08, N = 3SE +/- 0.15, N = 3SE +/- 0.07, N = 315.4618.9018.7723.0023.191. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PUSCH Processor Benchmark, Throughput ThreadXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 7.12, N = 15SE +/- 0.00, N = 4SE +/- 0.00, N = 4SE +/- 0.00, N = 4SE +/- 0.00, N = 4153.6183.5183.6183.6183.5MIN: 105.1 / MAX: 182.1-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 105.9-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 106-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 106-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 105.91. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -fno-trapping-math -fno-math-errno -ldl

Aircrack-ng

Aircrack-ng is a tool for assessing WiFi/WLAN network security. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgk/s, More Is BetterAircrack-ng 1.7Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020K40K60K80K100KSE +/- 17.95, N = 3SE +/- 65.97, N = 3SE +/- 45.75, N = 3SE +/- 194.99, N = 3SE +/- 153.97, N = 335441.7148803.0048527.8292030.3592259.951. (CXX) g++ options: -std=gnu++17 -O3 -fvisibility=hidden -fcommon -rdynamic -lnl-3 -lnl-genl-3 -lpcre -lsqlite3 -lpthread -lz -lssl -lcrypto -lhwloc -ldl -lm -pthread

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: WPA PSKXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030K60K90K120K150KSE +/- 135.70, N = 3SE +/- 245.32, N = 3SE +/- 221.78, N = 3SE +/- 573.94, N = 3SE +/- 425.10, N = 32956085712855321608051606641. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: bcryptXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52009K18K27K36K45KSE +/- 127.68, N = 3SE +/- 45.41, N = 3SE +/- 43.32, N = 3SE +/- 61.43, N = 3SE +/- 104.20, N = 318787232712323543512436951. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: BlowfishXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52009K18K27K36K45KSE +/- 68.54, N = 3SE +/- 121.86, N = 3SE +/- 44.47, N = 3SE +/- 344.23, N = 3SE +/- 39.87, N = 318841231742324342924437951. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 16 - Buffer Length: 256 - Filter Length: 512Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520060M120M180M240M300MSE +/- 1245404.89, N = 3SE +/- 176162.80, N = 3SE +/- 75055.53, N = 3SE +/- 2047567.66, N = 3SE +/- 520000.00, N = 31592700002129600002116700002869600002923900001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: GarlicoinXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52007001400210028003500SE +/- 4.51, N = 3SE +/- 4.36, N = 3SE +/- 1.43, N = 3SE +/- 1.95, N = 3SE +/- 4.61, N = 31356.311691.291674.993196.423179.321. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Stress-NG

Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Vector Floating PointXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520013K26K39K52K65KSE +/- 378.33, N = 3SE +/- 404.29, N = 3SE +/- 415.80, N = 3SE +/- 131.10, N = 3SE +/- 261.75, N = 328846.0630715.1430936.5059053.0159037.141. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Matrix 3D MathXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52006001200180024003000SE +/- 0.40, N = 3SE +/- 3.38, N = 3SE +/- 2.55, N = 3SE +/- 1.97, N = 3SE +/- 0.58, N = 32778.891953.522029.451924.702011.171. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 4 - Buffer Length: 256 - Filter Length: 512Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020M40M60M80M100MSE +/- 493083.61, N = 3SE +/- 444695.78, N = 3SE +/- 342214.52, N = 3SE +/- 302227.47, N = 3SE +/- 190528.21, N = 379553667778110007750866776415667768090001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Stress-NG

Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Floating PointXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002K4K6K8K10KSE +/- 7.41, N = 3SE +/- 2.36, N = 3SE +/- 6.00, N = 3SE +/- 8.60, N = 3SE +/- 15.20, N = 34740.284283.484251.398242.968199.651. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Vector ShuffleXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004K8K12K16K20KSE +/- 104.90, N = 3SE +/- 3.63, N = 3SE +/- 2.63, N = 3SE +/- 12.64, N = 3SE +/- 12.57, N = 39138.298606.948556.7216537.8316452.471. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Fused Multiply-AddXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520012M24M36M48M60MSE +/- 267691.10, N = 3SE +/- 28572.99, N = 3SE +/- 252666.10, N = 3SE +/- 59842.82, N = 3SE +/- 500794.58, N = 329762395.5730389834.8229927440.6957429956.1556410032.251. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 2 - Buffer Length: 256 - Filter Length: 32Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030M60M90M120M150MSE +/- 724829.48, N = 3SE +/- 15275.25, N = 3SE +/- 13333.33, N = 3SE +/- 14529.66, N = 3SE +/- 282921.27, N = 31273766671071400001071466671033266671036966671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: scryptXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520070140210280350SE +/- 0.08, N = 3SE +/- 0.23, N = 3SE +/- 0.10, N = 3SE +/- 1.89, N = 3SE +/- 1.34, N = 3119.86163.88162.88310.84308.211. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Stress-NG

Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Wide Vector MathXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200200K400K600K800K1000KSE +/- 373.64, N = 3SE +/- 313.30, N = 3SE +/- 205.90, N = 3SE +/- 1012.30, N = 3SE +/- 1811.04, N = 3576480.15416118.50413926.82781906.97778238.331. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 4 - Buffer Length: 256 - Filter Length: 32Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050M100M150M200M250MSE +/- 17320.51, N = 3SE +/- 1006495.57, N = 3SE +/- 325695.84, N = 3SE +/- 452413.28, N = 3SE +/- 674487.46, N = 32534500002128600002097433332063066672058900001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 16 - Buffer Length: 256 - Filter Length: 32Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200200M400M600M800M1000MSE +/- 5153390.47, N = 3SE +/- 199192.82, N = 3SE +/- 812656.82, N = 3SE +/- 1556216.92, N = 3SE +/- 67412.49, N = 37110500007401666677355666677876333337959666671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 8 - Buffer Length: 256 - Filter Length: 32Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520090M180M270M360M450MSE +/- 2917967.33, N = 3SE +/- 1727319.70, N = 3SE +/- 388858.73, N = 3SE +/- 1946723.74, N = 3SE +/- 2535286.53, N = 34391500004112900004093866674040100004049766671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 1 - Buffer Length: 256 - Filter Length: 32Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520014M28M42M56M70MSE +/- 637003.49, N = 3SE +/- 881.92, N = 3SE +/- 2645.75, N = 3SE +/- 469067.51, N = 3SE +/- 318365.79, N = 363236667536613335365900052493000530186671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: DeepcoinXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002K4K6K8K10KSE +/- 36.27, N = 3SE +/- 3.81, N = 3SE +/- 3.02, N = 3SE +/- 3.98, N = 3SE +/- 5.77, N = 33294.494405.854377.748421.158356.741. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Stress-NG

Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: AVX-512 VNNIXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200600K1200K1800K2400K3000KSE +/- 5501.16, N = 3SE +/- 1569.35, N = 3SE +/- 1225.91, N = 3SE +/- 3334.21, N = 3SE +/- 2539.18, N = 31125583.011376354.001371242.642572055.582571394.821. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: x86_64 RdRandXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002M4M6M8M10MSE +/- 7.65, N = 3SE +/- 1009.93, N = 3SE +/- 788.85, N = 3SE +/- 980.70, N = 3SE +/- 6736.08, N = 381538.344400145.014374980.168497821.558469095.711. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -ljpeg -lmpfr -lpthread -lrt -lsctp -lz

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: LBC, LBRY CreditsXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003K6K9K12K15KSE +/- 15.18, N = 3SE +/- 4.08, N = 3SE +/- 7.26, N = 3SE +/- 47.02, N = 3SE +/- 5.77, N = 36037.268437.938389.1516317.0016120.001. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: SkeincoinXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52007K14K21K28K35KSE +/- 139.08, N = 3SE +/- 5.77, N = 3SE +/- 39.30, N = 3SE +/- 17.64, N = 3SE +/- 11.55, N = 313933180601798334753346401. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: Triple SHA-256, OnecoinXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020K40K60K80K100KSE +/- 3.33, N = 3SE +/- 5.77, N = 3SE +/- 8.82, N = 3SE +/- 73.33, N = 3SE +/- 146.40, N = 338847513705108797753973401. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: Quad SHA-256, PyriteXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520013K26K39K52K65KSE +/- 3.33, N = 3SE +/- 5.77, N = 3SE +/- 78.81, N = 3SE +/- 297.56, N = 3SE +/- 331.71, N = 326527320103190361347613601. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 23.5Algorithm: Blake-2 SXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030K60K90K120K150KSE +/- 460.90, N = 3SE +/- 74.24, N = 3SE +/- 29.06, N = 3SE +/- 125.03, N = 3SE +/- 99.39, N = 35616074883745131434801427531. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PUSCH Processor Benchmark, Throughput TotalXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 9.26, N = 15SE +/- 17.52, N = 15SE +/- 0.10, N = 6SE +/- 0.04, N = 4SE +/- 0.05, N = 41068.91517.41470.51611.51611.5MIN: 637.2 / MAX: 1111.5-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 849.7 / MAX: 1610.7-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 849.1 / MAX: 1470.7-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 1019.3 / MAX: 1611.6-mavx512f -mavx512cd -mavx512bw -mavx512dq - MIN: 1019.4 / MAX: 1611.61. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -fno-trapping-math -fno-math-errno -ldl

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.2.4Encode Settings: Quality 100, Lossless, Highest CompressionXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.1980.3960.5940.7920.99SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 30.880.840.840.820.821. (CC) gcc options: -fvisibility=hidden -O2 -lm

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 4 - Input: Bosphorus 4KXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.024, N = 3SE +/- 0.014, N = 3SE +/- 0.016, N = 3SE +/- 0.016, N = 3SE +/- 0.009, N = 34.3485.7115.7427.8317.8381. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Llamafile

Mozilla's Llamafile allows distributing and running large language models (LLMs) as a single file. Llamafile aims to make open-source LLMs more accessible to developers and users. Llamafile supports a variety of models, CPUs and GPUs, and other options. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.6Test: llava-v1.5-7b-q4 - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 313.3210.5710.4810.9011.37

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.BXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004K8K12K16K20KSE +/- 44.80, N = 3SE +/- 48.58, N = 3SE +/- 12.84, N = 3SE +/- 211.58, N = 3SE +/- 253.63, N = 312750.9710661.6611152.3118360.8218660.901. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 13 - Input: Bosphorus 4KXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 0.99, N = 15SE +/- 1.96, N = 15SE +/- 2.01, N = 15SE +/- 4.13, N = 15SE +/- 4.27, N = 15104.61121.44122.91160.65165.291. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Y-Cruncher

Y-Cruncher is a multi-threaded Pi benchmark capable of computing Pi to trillions of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 1BXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200816243240SE +/- 0.21, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 333.8925.2425.4618.8018.21

JPEG-XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG-XL libjxl 0.10.1Input: JPEG - Quality: 100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200612182430SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 313.4816.5816.5226.9627.221. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

OpenBenchmarking.orgMP/s, More Is BetterJPEG-XL libjxl 0.10.1Input: PNG - Quality: 100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200714212835SE +/- 0.07, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 313.6516.6916.6627.5627.751. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: TradesoapXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005001000150020002500SE +/- 16.67, N = 4SE +/- 7.41, N = 4SE +/- 20.46, N = 5SE +/- 16.84, N = 13SE +/- 17.19, N = 419141964197924392474

Timed Mesa Compilation

This test profile times how long it takes to compile Mesa with Meson/Ninja. For minimizing build dependencies and avoid versioning conflicts, test this is just the core Mesa build without LLVM or the extra Gallium3D/Mesa drivers enabled. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Mesa Compilation 24.0Time To CompileXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200816243240SE +/- 0.07, N = 3SE +/- 0.02, N = 3SE +/- 0.11, N = 3SE +/- 0.06, N = 3SE +/- 0.21, N = 332.5228.2028.2617.7317.46

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Very FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001020304050SE +/- 0.06, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.07, N = 4SE +/- 0.08, N = 418.9725.3225.2542.2042.60

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: H2 Database EngineXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005001000150020002500SE +/- 23.84, N = 3SE +/- 12.68, N = 3SE +/- 7.51, N = 3SE +/- 9.17, N = 3SE +/- 13.96, N = 321282000197718721826

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache Lucene Search EngineXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52007001400210028003500SE +/- 50.09, N = 15SE +/- 17.06, N = 5SE +/- 18.64, N = 5SE +/- 9.20, N = 5SE +/- 9.53, N = 531132404240920832080

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Ultra FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.30, N = 10SE +/- 0.05, N = 4SE +/- 0.04, N = 4SE +/- 0.08, N = 6SE +/- 0.07, N = 639.5952.2552.3080.5081.231. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: EP.CXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005001000150020002500SE +/- 55.41, N = 15SE +/- 4.34, N = 6SE +/- 41.08, N = 15SE +/- 18.71, N = 8SE +/- 13.01, N = 81470.811101.231169.682117.292123.371. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520090180270360450SE +/- 0.04, N = 3SE +/- 0.35, N = 3SE +/- 0.29, N = 3SE +/- 0.65, N = 4SE +/- 0.54, N = 4205.93204.84203.10388.63392.241. (CXX) g++ options: -O3 -flto -pthread

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: AlexNetXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050100150200250SE +/- 1.37, N = 3SE +/- 0.05, N = 3SE +/- 0.23, N = 3SE +/- 0.14, N = 4SE +/- 0.14, N = 4160.52171.11172.89236.71242.49

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Super FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001122334455SE +/- 0.08, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 4SE +/- 0.07, N = 420.1726.7326.6545.8746.41

LULESH

LULESH is the Livermore Unstructured Lagrangian Explicit Shock Hydrodynamics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgz/s, More Is BetterLULESH 2.0.3Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002K4K6K8K10KSE +/- 210.25, N = 15SE +/- 39.33, N = 5SE +/- 55.23, N = 6SE +/- 15.60, N = 6SE +/- 57.24, N = 128126.326454.676744.868006.358106.151. (CXX) g++ options: -O3 -fopenmp -lm -lmpi_cxx -lmpi

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: jMonkeyEngineXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520015003000450060007500SE +/- 4.67, N = 3SE +/- 0.88, N = 3SE +/- 0.33, N = 3SE +/- 1.20, N = 3SE +/- 0.67, N = 368276804680268086806

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Deconvolution Batch shapes_1d - Engine: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.02381, N = 3SE +/- 0.01491, N = 3SE +/- 0.00421, N = 3SE +/- 0.00488, N = 3SE +/- 0.01295, N = 36.853183.876653.910892.986543.00636MIN: 3.93MIN: 3.48MIN: 3.5MIN: 2.41MIN: 2.41. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

7-Zip Compression

This is a test of 7-Zip compression/decompression with its integrated benchmark feature. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Decompression RatingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520040K80K120K160K200KSE +/- 900.25, N = 3SE +/- 835.95, N = 3SE +/- 299.70, N = 3SE +/- 110.28, N = 3SE +/- 369.09, N = 36152787804889451689571703731. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Compression RatingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520040K80K120K160K200KSE +/- 436.66, N = 3SE +/- 118.34, N = 3SE +/- 195.02, N = 3SE +/- 227.16, N = 3SE +/- 313.05, N = 3949901018491035711671861708881. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 4KXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.46, N = 15SE +/- 0.09, N = 5SE +/- 0.09, N = 5SE +/- 0.05, N = 6SE +/- 0.15, N = 679.8570.2971.4899.84103.521. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Super FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001428425670SE +/- 0.31, N = 4SE +/- 0.02, N = 4SE +/- 0.05, N = 4SE +/- 0.07, N = 5SE +/- 0.07, N = 528.6438.8538.8262.2362.561. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Spring BootXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520030060090012001500SE +/- 6.95, N = 5SE +/- 7.11, N = 5SE +/- 6.93, N = 5SE +/- 10.62, N = 13SE +/- 8.13, N = 512201159116215541554

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: IP Shapes 1D - Engine: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.65571.31141.96712.62283.2785SE +/- 0.01544, N = 4SE +/- 0.00306, N = 4SE +/- 0.00536, N = 4SE +/- 0.00498, N = 4SE +/- 0.00269, N = 42.914081.987731.999861.132031.13000MIN: 2.51MIN: 1.84MIN: 1.85MIN: 1.03MIN: 1.031. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

Z3 Theorem Prover

The Z3 Theorem Prover / SMT solver is developed by Microsoft Research under the MIT license. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterZ3 Theorem Prover 4.12.1SMT File: 1.smt2Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200510152025SE +/- 0.08, N = 3SE +/- 0.05, N = 3SE +/- 0.11, N = 3SE +/- 0.16, N = 3SE +/- 0.15, N = 318.6620.5020.4320.7820.331. (CXX) g++ options: -lpthread -std=c++17 -fvisibility=hidden -mfpmath=sse -msse -msse2 -O3 -fPIC

QuantLib

QuantLib is an open-source library/framework around quantitative finance for modeling, trading and risk management scenarios. QuantLib is written in C++ with Boost and its built-in benchmark used reports the QuantLib Benchmark Index benchmark score. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMFLOPS, More Is BetterQuantLib 1.32Configuration: Single-ThreadedXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520011002200330044005500SE +/- 34.53, N = 4SE +/- 43.08, N = 3SE +/- 17.50, N = 3SE +/- 56.37, N = 3SE +/- 46.95, N = 35019.24518.44457.84425.04425.21. (CXX) g++ options: -O3 -march=native -fPIE -pie

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Ultra FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001224364860SE +/- 0.01, N = 3SE +/- 0.06, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 5SE +/- 0.08, N = 524.6932.0931.8654.0754.54

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache TomcatXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520014002800420056007000SE +/- 14.71, N = 3SE +/- 2.73, N = 3SE +/- 9.21, N = 3SE +/- 4.40, N = 4SE +/- 2.95, N = 466876467648233673370

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Very FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001122334455SE +/- 0.19, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.06, N = 4SE +/- 0.09, N = 423.0531.0531.0447.6748.111. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

Algebraic Multi-Grid Benchmark

AMG is a parallel algebraic multigrid solver for linear systems arising from problems on unstructured grids. The driver provided with AMG builds linear systems for various 3-dimensional problems. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFigure Of Merit, More Is BetterAlgebraic Multi-Grid Benchmark 1.2Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200100M200M300M400M500MSE +/- 269181.94, N = 5SE +/- 333950.91, N = 4SE +/- 445215.93, N = 4SE +/- 201300.38, N = 3SE +/- 143363.91, N = 34537038803506567503523617753625744333786970001. (CC) gcc options: -lparcsr_ls -lparcsr_mv -lseq_mv -lIJ_mv -lkrylov -lHYPRE_utilities -lm -fopenmp -lmpi

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: TradebeansXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52007001400210028003500SE +/- 14.57, N = 4SE +/- 31.95, N = 4SE +/- 22.67, N = 4SE +/- 38.01, N = 4SE +/- 21.46, N = 434102775281533593329

Parallel BZIP2 Compression

This test measures the time needed to compress a file (FreeBSD-13.0-RELEASE-amd64-memstick.img) using Parallel BZIP2 compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterParallel BZIP2 Compression 1.1.13FreeBSD-13.0-RELEASE-amd64-memstick.img CompressionXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.044479, N = 6SE +/- 0.045334, N = 15SE +/- 0.045809, N = 15SE +/- 0.023270, N = 9SE +/- 0.025811, N = 106.6124866.1768536.1544603.3255433.2998791. (CXX) g++ options: -O2 -pthread -lbz2 -lpthread

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: PMD Source Code AnalyzerXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 8.96, N = 15SE +/- 9.22, N = 15SE +/- 9.55, N = 7SE +/- 10.09, N = 7SE +/- 5.68, N = 711221142111310741053

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Rainbow Colors and Prism - Acceleration: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.03, N = 3SE +/- 0.01, N = 4SE +/- 0.04, N = 4SE +/- 0.11, N = 5SE +/- 0.06, N = 59.3511.8911.9316.9416.98MIN: 8.52 / MAX: 10.72MIN: 11.05 / MAX: 12MIN: 11.1 / MAX: 12.12MIN: 15.4 / MAX: 17.31MIN: 15.46 / MAX: 17.3

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: BioJava Biological Data FrameworkXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52009001800270036004500SE +/- 25.22, N = 4SE +/- 19.50, N = 4SE +/- 20.56, N = 4SE +/- 15.83, N = 4SE +/- 33.62, N = 437084347432744134425

Y-Cruncher

Y-Cruncher is a multi-threaded Pi benchmark capable of computing Pi to trillions of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 500MXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.124, N = 4SE +/- 0.040, N = 4SE +/- 0.036, N = 4SE +/- 0.013, N = 5SE +/- 0.010, N = 514.92611.55511.5898.9208.652

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: MG.CXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005K10K15K20K25KSE +/- 532.70, N = 12SE +/- 158.41, N = 5SE +/- 213.35, N = 5SE +/- 94.12, N = 5SE +/- 161.84, N = 524688.5819566.6720265.5220223.6020961.351. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.11Video Input: Bosphorus 1080p - Video Preset: FasterXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001122334455SE +/- 0.25, N = 4SE +/- 0.05, N = 4SE +/- 0.04, N = 4SE +/- 0.09, N = 4SE +/- 0.11, N = 440.0744.5845.1450.4150.911. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

Dolfyn

Dolfyn is a Computational Fluid Dynamics (CFD) code of modern numerical simulation techniques. The Dolfyn test profile measures the execution time of the bundled computational fluid dynamics demos that are bundled with Dolfyn. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDolfyn 0.527Computational Fluid DynamicsXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.03, N = 5SE +/- 0.04, N = 5SE +/- 0.02, N = 5SE +/- 0.06, N = 5SE +/- 0.11, N = 510.4210.6310.7310.9310.96

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: JythonXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52008001600240032004000SE +/- 14.73, N = 5SE +/- 14.78, N = 4SE +/- 16.55, N = 4SE +/- 15.77, N = 4SE +/- 34.16, N = 424333790376237953871

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: MediumXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001428425670SE +/- 0.14, N = 4SE +/- 0.07, N = 4SE +/- 0.03, N = 4SE +/- 0.05, N = 5SE +/- 0.07, N = 536.5945.7545.6363.8964.16

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: MediumXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.00, N = 3SE +/- 0.05, N = 4SE +/- 0.04, N = 4SE +/- 0.09, N = 5SE +/- 0.18, N = 589.1480.5379.60153.70154.721. (CXX) g++ options: -O3 -flto -pthread

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache Lucene Search IndexXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005001000150020002500SE +/- 19.22, N = 3SE +/- 17.23, N = 3SE +/- 18.27, N = 3SE +/- 11.06, N = 3SE +/- 12.68, N = 322962407239025172471

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.2.4Encode Settings: Quality 100, LosslessXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.48830.97661.46491.95322.4415SE +/- 0.01, N = 5SE +/- 0.00, N = 5SE +/- 0.01, N = 5SE +/- 0.01, N = 4SE +/- 0.00, N = 42.172.162.172.082.091. (CC) gcc options: -fvisibility=hidden -O2 -lm

x265

This is a simple test of the x265 encoder run on the CPU with 1080p and 4K options for H.265 video encode performance with x265. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 1080pXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 0.77, N = 15SE +/- 0.38, N = 6SE +/- 0.53, N = 6SE +/- 0.13, N = 7SE +/- 0.21, N = 784.3389.2389.12117.25118.061. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache KafkaXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520011002200330044005500SE +/- 2.00, N = 3SE +/- 29.48, N = 3SE +/- 29.67, N = 3SE +/- 27.06, N = 3SE +/- 32.17, N = 349845031502750305055

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 8 - Input: Bosphorus 4KXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001632486480SE +/- 0.30, N = 3SE +/- 0.02, N = 4SE +/- 0.10, N = 4SE +/- 0.25, N = 5SE +/- 0.15, N = 536.1446.8447.2971.1472.741. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

easyWave

The easyWave software allows simulating tsunami generation and propagation in the context of early warning systems. EasyWave supports making use of OpenMP for CPU multi-threading and there are also GPU ports available but not currently incorporated as part of this test profile. The easyWave tsunami generation software is run with one of the example/reference input files for measuring the CPU execution time. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BettereasyWave r34Input: e2Asean Grid + BengkuluSept2007 Source - Time: 240Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.089, N = 15SE +/- 0.009, N = 6SE +/- 0.030, N = 6SE +/- 0.008, N = 8SE +/- 0.010, N = 96.2786.1665.9903.1643.0371. (CXX) g++ options: -O3 -fopenmp

PyBench

This test profile reports the total time of the different average timed test results from PyBench. PyBench reports average test times for different functions such as BuiltinFunctionCalls and NestedForLoops, with this total result providing a rough estimate as to Python's average performance on a given system. This test profile runs PyBench each time for 20 rounds. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyBench 2018-02-16Total For Average Test TimesXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200120240360480600SE +/- 2.82, N = 5SE +/- 1.38, N = 4SE +/- 2.53, N = 4SE +/- 2.65, N = 4SE +/- 6.41, N = 4415532527536533

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 10 - Input: Bosphorus 4KXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 1.10, N = 15SE +/- 0.10, N = 6SE +/- 0.12, N = 6SE +/- 0.15, N = 8SE +/- 0.14, N = 8101.83107.65109.03153.78158.941. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: PSNR/SSIM Optimized - Input: Bosphorus 4KXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.64, N = 10SE +/- 0.14, N = 5SE +/- 0.08, N = 5SE +/- 0.14, N = 6SE +/- 0.13, N = 683.1974.9376.06107.37111.011. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 7 - Input: Bosphorus 4KXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.08, N = 4SE +/- 0.07, N = 5SE +/- 0.04, N = 5SE +/- 0.08, N = 6SE +/- 0.06, N = 647.9856.9557.2397.7099.611. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

ASKAP

ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 1.0Test: tConvolve OpenMP - DegriddingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520011002200330044005500SE +/- 41.03, N = 5SE +/- 29.29, N = 6SE +/- 27.18, N = 8SE +/- 0.00, N = 6SE +/- 49.85, N = 64282.183429.453426.035023.705172.441. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

OpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 1.0Test: tConvolve OpenMP - GriddingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520011002200330044005500SE +/- 25.17, N = 5SE +/- 11.75, N = 6SE +/- 24.04, N = 8SE +/- 38.41, N = 6SE +/- 41.40, N = 62556.222703.362739.344886.945073.231. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 1 - Model: ResNet-50Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.01, N = 5SE +/- 0.04, N = 5SE +/- 0.01, N = 5SE +/- 0.02, N = 5SE +/- 0.03, N = 513.7116.7717.4714.7814.93

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: GraphChiXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52006001200180024003000SE +/- 18.95, N = 4SE +/- 18.46, N = 4SE +/- 21.22, N = 4SE +/- 10.30, N = 4SE +/- 14.73, N = 426752535256125212472

Primesieve

Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 12.1Length: 1e12Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200510152025SE +/- 0.302, N = 3SE +/- 0.018, N = 4SE +/- 0.007, N = 4SE +/- 0.008, N = 6SE +/- 0.007, N = 621.09512.41612.5066.6696.6911. (CXX) g++ options: -O3

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 12 - Input: Bosphorus 4KXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 1.00, N = 15SE +/- 0.15, N = 7SE +/- 0.14, N = 7SE +/- 0.69, N = 8SE +/- 0.33, N = 8105.28123.64125.28164.85170.211. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Very FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 0.95, N = 15SE +/- 0.05, N = 7SE +/- 0.14, N = 7SE +/- 0.14, N = 9SE +/- 0.11, N = 989.31112.05111.78165.55166.81

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: IP Shapes 3D - Engine: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.01930, N = 5SE +/- 0.02020, N = 5SE +/- 0.00833, N = 5SE +/- 0.00590, N = 5SE +/- 0.00611, N = 57.596954.612114.678404.070563.88685MIN: 7.29MIN: 4.4MIN: 4.46MIN: 3.76MIN: 3.571. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Avrora AVR Simulation FrameworkXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52005001000150020002500SE +/- 2.49, N = 8SE +/- 3.46, N = 6SE +/- 3.26, N = 6SE +/- 21.04, N = 8SE +/- 20.96, N = 614512430241523932359

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 6Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.088, N = 15SE +/- 0.007, N = 7SE +/- 0.014, N = 7SE +/- 0.018, N = 9SE +/- 0.010, N = 96.2605.3015.3383.4473.4211. (CXX) g++ options: -O3 -fPIC -lm

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 6, LosslessXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.014, N = 5SE +/- 0.012, N = 6SE +/- 0.021, N = 6SE +/- 0.018, N = 7SE +/- 0.044, N = 78.9698.1408.1475.9325.8901. (CXX) g++ options: -O3 -fPIC -lm

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Super FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 0.97, N = 15SE +/- 0.07, N = 7SE +/- 0.08, N = 7SE +/- 0.19, N = 9SE +/- 0.16, N = 996.33120.30119.83180.54181.83

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 4 - Input: Bosphorus 1080pXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200612182430SE +/- 0.15, N = 5SE +/- 0.08, N = 5SE +/- 0.07, N = 5SE +/- 0.05, N = 6SE +/- 0.13, N = 615.9219.9619.8923.9424.171. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: MediumXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520020406080100SE +/- 0.37, N = 4SE +/- 0.04, N = 5SE +/- 0.04, N = 5SE +/- 0.15, N = 6SE +/- 0.08, N = 650.4764.8064.6887.5188.301. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 1 - Model: AlexNetXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200510152025SE +/- 0.01, N = 6SE +/- 0.01, N = 5SE +/- 0.00, N = 5SE +/- 0.00, N = 5SE +/- 0.01, N = 518.5014.5315.0114.3714.96

RNNoise

RNNoise is a recurrent neural network for audio noise reduction developed by Mozilla and Xiph.Org. This test profile is a single-threaded test measuring the time to denoise a sample 26 minute long 16-bit RAW audio file using this recurrent neural network noise suppression library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRNNoise 0.2Input: 26 Minute Long Talking SampleXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.059, N = 7SE +/- 0.016, N = 6SE +/- 0.050, N = 6SE +/- 0.063, N = 6SE +/- 0.033, N = 66.4007.0287.1847.2917.4141. (CC) gcc options: -O2 -pedantic -fvisibility=hidden

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 8 - Input: Bosphorus 1080pXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 1.11, N = 15SE +/- 0.12, N = 8SE +/- 0.13, N = 8SE +/- 0.22, N = 9SE +/- 0.29, N = 9120.53147.31148.34187.54189.411. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PDSCH Processor Benchmark, Throughput TotalXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003K6K9K12K15KSE +/- 5.11, N = 5SE +/- 8.66, N = 7SE +/- 38.56, N = 7SE +/- 66.74, N = 6SE +/- 51.53, N = 62908.57792.67530.012647.512852.6-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq1. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -fno-trapping-math -fno-math-errno -ldl

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Convolution Batch Shapes Auto - Engine: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.04485, N = 7SE +/- 0.00879, N = 7SE +/- 0.00717, N = 7SE +/- 0.01136, N = 7SE +/- 0.01415, N = 76.380538.232208.168006.725366.40719MIN: 5.92MIN: 7.77MIN: 7.7MIN: 6.2MIN: 5.931. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

Google Draco

Draco is a library developed by Google for compressing/decompressing 3D geometric meshes and point clouds. This test profile uses some Artec3D PLY models as the sample 3D model input formats for Draco compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterGoogle Draco 1.5.6Model: Church FacadeXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520012002400360048006000SE +/- 17.70, N = 6SE +/- 12.60, N = 6SE +/- 10.00, N = 7SE +/- 31.00, N = 6SE +/- 29.27, N = 6560751635083523752241. (CXX) g++ options: -O3

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Ultra FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 1.21, N = 15SE +/- 0.14, N = 8SE +/- 0.06, N = 8SE +/- 0.15, N = 9SE +/- 0.20, N = 9115.40143.28142.97201.92203.39

GNU Octave Benchmark

This test profile measures how long it takes to complete several reference GNU Octave files via octave-benchmark. GNU Octave is used for numerical computations and is an open-source alternative to MATLAB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGNU Octave Benchmark 8.4.0Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200246810SE +/- 0.011, N = 6SE +/- 0.016, N = 7SE +/- 0.017, N = 7SE +/- 0.043, N = 7SE +/- 0.021, N = 76.4625.6685.5875.7545.780

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: FOP Print FormatterXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520090180270360450SE +/- 6.07, N = 15SE +/- 3.32, N = 15SE +/- 5.35, N = 15SE +/- 3.38, N = 15SE +/- 5.40, N = 15426423414403399

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.2.4Encode Settings: Quality 100, Highest CompressionXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001.19252.3853.57754.775.9625SE +/- 0.02, N = 7SE +/- 0.03, N = 8SE +/- 0.03, N = 8SE +/- 0.03, N = 8SE +/- 0.01, N = 84.735.305.285.155.101. (CC) gcc options: -fvisibility=hidden -O2 -lm

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 10, LosslessXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.94611.89222.83833.78444.7305SE +/- 0.011, N = 8SE +/- 0.009, N = 8SE +/- 0.025, N = 8SE +/- 0.020, N = 8SE +/- 0.038, N = 154.1624.2024.2053.8043.8411. (CXX) g++ options: -O3 -fPIC -lm

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Zxing 1D/2D Barcode Image ProcessingXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 30.56, N = 15SE +/- 3.37, N = 8SE +/- 2.37, N = 8SE +/- 4.29, N = 9SE +/- 4.09, N = 9105310161019597589

Google Draco

Draco is a library developed by Google for compressing/decompressing 3D geometric meshes and point clouds. This test profile uses some Artec3D PLY models as the sample 3D model input formats for Draco compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterGoogle Draco 1.5.6Model: LionXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52009001800270036004500SE +/- 9.84, N = 7SE +/- 5.99, N = 7SE +/- 12.07, N = 7SE +/- 18.70, N = 7SE +/- 17.41, N = 7383039223891404938781. (CXX) g++ options: -O3

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Super FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520050100150200250SE +/- 2.29, N = 15SE +/- 0.11, N = 9SE +/- 0.09, N = 9SE +/- 0.29, N = 10SE +/- 0.26, N = 10139.72168.47168.08245.56248.391. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Very FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004080120160200SE +/- 0.87, N = 7SE +/- 0.08, N = 8SE +/- 0.10, N = 8SE +/- 0.24, N = 9SE +/- 0.25, N = 9103.04129.55129.02184.98186.561. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Batik SVG ToolkitXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 5.39, N = 8SE +/- 6.28, N = 8SE +/- 5.55, N = 8SE +/- 5.97, N = 8SE +/- 3.65, N = 8897927930969936

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 7 - Input: Bosphorus 1080pXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520070140210280350SE +/- 2.05, N = 15SE +/- 0.16, N = 8SE +/- 0.21, N = 8SE +/- 0.34, N = 10SE +/- 0.26, N = 10168.39178.68178.20307.89311.951. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

Helsing

Helsing is an open-source POSIX vampire number generator. This test profile measures the time it takes to generate vampire numbers between varying numbers of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterHelsing 1.0-betaDigit Range: 12 digitXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52000.9181.8362.7543.6724.59SE +/- 0.028, N = 15SE +/- 0.004, N = 8SE +/- 0.004, N = 8SE +/- 0.002, N = 10SE +/- 0.003, N = 114.0803.7533.8282.0112.0161. (CC) gcc options: -O2 -pthread

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 1 - Model: GoogLeNetXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001428425670SE +/- 0.05, N = 8SE +/- 0.11, N = 9SE +/- 0.10, N = 9SE +/- 0.05, N = 8SE +/- 0.18, N = 851.3362.0262.9957.2157.34

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: Visual Quality Optimized - Input: Bosphorus 1080pXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520080160240320400SE +/- 3.49, N = 15SE +/- 0.90, N = 9SE +/- 0.76, N = 9SE +/- 1.84, N = 15SE +/- 1.71, N = 15221.33201.53203.16355.84364.901. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Deconvolution Batch shapes_3d - Engine: CPUXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52001.11332.22663.33994.45325.5665SE +/- 0.10597, N = 15SE +/- 0.00353, N = 9SE +/- 0.00661, N = 9SE +/- 0.00342, N = 9SE +/- 0.00405, N = 94.947923.798903.834122.461642.46159MIN: 4.45MIN: 3.63MIN: 3.58MIN: 2.35MIN: 2.351. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

LAMMPS Molecular Dynamics Simulator

LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 23Jun2022Model: Rhodopsin ProteinXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52003691215SE +/- 0.250, N = 15SE +/- 0.032, N = 9SE +/- 0.055, N = 9SE +/- 0.084, N = 15SE +/- 0.059, N = 117.2366.8486.82612.44912.4811. (CXX) g++ options: -O3 -lm -ldl

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PDSCH Processor Benchmark, Throughput ThreadXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 5.58, N = 9SE +/- 5.66, N = 10SE +/- 3.24, N = 10SE +/- 2.53, N = 9SE +/- 5.92, N = 10809.61082.71061.51027.81049.1-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq-mavx512f -mavx512cd -mavx512bw -mavx512dq1. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -fno-trapping-math -fno-math-errno -ldl

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Ultra FastXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520070140210280350SE +/- 4.30, N = 15SE +/- 0.17, N = 10SE +/- 0.14, N = 10SE +/- 0.21, N = 11SE +/- 0.21, N = 11197.51226.69226.13299.12301.571. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 1080pXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520090180270360450SE +/- 0.18, N = 10SE +/- 0.20, N = 10SE +/- 0.20, N = 10SE +/- 0.22, N = 11SE +/- 0.58, N = 11289.26240.66242.58398.58408.421. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080pXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520090180270360450SE +/- 0.53, N = 10SE +/- 0.36, N = 10SE +/- 0.28, N = 10SE +/- 0.27, N = 11SE +/- 0.42, N = 11295.69249.86251.44408.36419.261. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache Xalan XSLTXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200110220330440550SE +/- 2.76, N = 11SE +/- 2.21, N = 10SE +/- 2.71, N = 10SE +/- 2.09, N = 10SE +/- 3.47, N = 10408448450508506

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 10 - Input: Bosphorus 1080pXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200120240360480600SE +/- 4.53, N = 15SE +/- 0.26, N = 11SE +/- 0.24, N = 11SE +/- 1.05, N = 12SE +/- 1.04, N = 12383.64358.54361.05545.31560.811. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.2.4Encode Settings: Quality 100Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-520048121620SE +/- 0.05, N = 11SE +/- 0.04, N = 12SE +/- 0.06, N = 12SE +/- 0.02, N = 11SE +/- 0.10, N = 1115.3716.5916.5415.8416.111. (CC) gcc options: -fvisibility=hidden -O2 -lm

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 13 - Input: Bosphorus 1080pXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200160320480640800SE +/- 13.14, N = 15SE +/- 14.78, N = 15SE +/- 15.50, N = 15SE +/- 20.45, N = 15SE +/- 20.65, N = 15649.43621.63633.29711.53728.581. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 12 - Input: Bosphorus 1080pXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200140280420560700SE +/- 0.76, N = 12SE +/- 0.68, N = 12SE +/- 0.74, N = 12SE +/- 0.81, N = 12SE +/- 1.66, N = 15548.04539.87546.71649.20660.341. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Timed CPython Compilation

This test times how long it takes to build the reference Python implementation, CPython, with optimizations and LTO enabled for a release build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed CPython Compilation 3.10.6Build Configuration: DefaultXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004812162016.8516.5016.5814.1514.13

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.2.4Encode Settings: DefaultXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200612182430SE +/- 0.07, N = 13SE +/- 0.10, N = 13SE +/- 0.07, N = 13SE +/- 0.08, N = 13SE +/- 0.13, N = 1323.9326.7826.8125.7826.251. (CC) gcc options: -fvisibility=hidden -O2 -lm

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.4Test: r2c - Backend: FFTW - Precision: double - X Y Z: 128Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200918273645SE +/- 0.39, N = 15SE +/- 0.17, N = 15SE +/- 0.15, N = 15SE +/- 0.96, N = 15SE +/- 0.58, N = 1526.4822.4723.7235.0737.241. (CXX) g++ options: -O3

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.4Test: r2c - Backend: FFTW - Precision: float - X Y Z: 128Xeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-5200306090120150SE +/- 3.73, N = 15SE +/- 0.33, N = 15SE +/- 1.52, N = 15SE +/- 1.50, N = 15SE +/- 1.14, N = 1585.0174.1377.52125.54127.931. (CXX) g++ options: -O3

Java SciMark

This test runs the Java version of SciMark 2, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.2Computational Test: Dense LU Matrix FactorizationXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52004K8K12K16K20KSE +/- 8.09, N = 3SE +/- 217.22, N = 3SE +/- 146.88, N = 3SE +/- 171.55, N = 3SE +/- 115.98, N = 311976.0618030.9217934.5917230.8517774.36

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.2Computational Test: Monte CarloXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52006001200180024003000SE +/- 0.88, N = 3SE +/- 32.12, N = 3SE +/- 19.82, N = 3SE +/- 26.32, N = 3SE +/- 17.64, N = 31426.112570.602541.202456.472529.18

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.2Computational Test: Jacobi Successive Over-RelaxationXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52006001200180024003000SE +/- 2.05, N = 3SE +/- 22.92, N = 3SE +/- 22.65, N = 3SE +/- 18.83, N = 3SE +/- 10.01, N = 32667.902487.932489.612464.142518.58

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.2Computational Test: Fast Fourier TransformXeon E-2488 @ DDR5-4800EPYC 4364P @ DDR5-4800EPYC 4364P @ DDR5-5200EPYC 4564P @ DDR5-4800EPYC 4564P @ DDR5-52002004006008001000SE +/- 8.67, N = 3SE +/- 0.60, N = 3SE +/- 0.41, N = 3SE +/- 3.44, N = 3SE +/- 0.73, N = 31023.23470.91450.10453.05443.49

464 Results Shown

LAMMPS Molecular Dynamics Simulator
TensorFlow
OpenRadioss:
  Chrysler Neon 1M
  INIVOL and Fluid Structure Interaction Drop Container
Timed Linux Kernel Compilation
OpenFOAM:
  drivaerFastback, Medium Mesh Size - Execution Time
  drivaerFastback, Medium Mesh Size - Mesh Time
TensorFlow
Blender
OpenRadioss
Timed Node.js Compilation
OpenVKL
OpenRadioss
TensorFlow
Timed LLVM Compilation
GPAW
Timed LLVM Compilation
Xmrig
OpenRadioss
Llama.cpp
OpenRadioss
Timed Gem5 Compilation
High Performance Conjugate Gradient
Helsing
Stockfish
Apache IoTDB:
  800 - 100 - 800 - 400:
    Average Latency
    point/sec
OpenSSL:
  RSA4096:
    verify/s
    sign/s
Apache IoTDB:
  800 - 100 - 800 - 100:
    Average Latency
    point/sec
TensorFlow
MariaDB
SecureMark
MariaDB:
  oltp_read_write - 64
  oltp_write_only - 64
  oltp_update_non_index - 64
  oltp_update_index - 32
  oltp_update_index - 64
  oltp_write_only - 32
  oltp_update_non_index - 32
Timed Godot Game Engine Compilation
Blender
ClickHouse:
  100M Rows Hits Dataset, Third Run
  100M Rows Hits Dataset, Second Run
  100M Rows Hits Dataset, First Run / Cold Cache
NAS Parallel Benchmarks
TensorFlow
Blender
TensorFlow
ONNX Runtime:
  fcn-resnet101-11 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
  ResNet50 v1-12-int8 - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
miniBUDE:
  OpenMP - BM2:
    Billion Interactions/s
    GFInst/s
BRL-CAD
Apache IoTDB:
  800 - 100 - 500 - 400:
    Average Latency
    point/sec
  800 - 100 - 500 - 100:
    Average Latency
    point/sec
  500 - 100 - 800 - 400:
    Average Latency
    point/sec
  500 - 100 - 800 - 100:
    Average Latency
    point/sec
OpenSSL:
  ChaCha20
  AES-128-GCM
  ChaCha20-Poly1305
  AES-256-GCM
  SHA512
  SHA256
Xcompact3d Incompact3d
SVT-HEVC
MariaDB mariadb-slap
Pennant
CloverLeaf
Java JMH
OSPRay
ONNX Runtime:
  yolov4 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
  GPT-2 - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
ScyllaDB
ONNX Runtime:
  yolov4 - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
Appleseed
PostgreSQL:
  100 - 500 - Read Write - Average Latency
  100 - 500 - Read Write
Apache IoTDB:
  500 - 100 - 500 - 400:
    Average Latency
    point/sec
LuxCoreRender
OSPRay
easyWave
NAS Parallel Benchmarks
ACES DGEMM
Primesieve
Apache IoTDB:
  500 - 100 - 500 - 100:
    Average Latency
    point/sec
ASTC Encoder:
  Very Thorough
  Exhaustive
PostgreSQL:
  100 - 800 - Read Only - Average Latency
  100 - 800 - Read Only
  100 - 500 - Read Only - Average Latency
  100 - 500 - Read Only
  100 - 800 - Read Write - Average Latency
  100 - 800 - Read Write
Pennant
DuckDB
FFmpeg:
  libx265 - Platform
  libx265 - Video On Demand
Zstd Compression:
  19, Long Mode - Decompression Speed
  19, Long Mode - Compression Speed
Llamafile
PyTorch
FFmpeg
OSPRay
Blender
Zstd Compression:
  19 - Decompression Speed
  19 - Compression Speed
SPECFEM3D
NAS Parallel Benchmarks
SPECFEM3D
TensorFlow
SPECFEM3D
Llamafile
Zstd Compression:
  8 - Decompression Speed
  8 - Compression Speed
Intel Open Image Denoise
Blender
NAMD
ONNX Runtime:
  Faster R-CNN R-50-FPN-int8 - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
  bertsquad-12 - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
  T5 Encoder - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
CockroachDB
TensorFlow
CockroachDB:
  KV, 50% Reads - 128
  KV, 95% Reads - 128
Dragonflydb
ASKAP:
  tConvolve MPI - Gridding
  tConvolve MPI - Degridding
GROMACS
Zstd Compression:
  12 - Decompression Speed
  12 - Compression Speed
libavif avifenc
Numpy Benchmark
NAS Parallel Benchmarks
VVenC
ONNX Runtime:
  ArcFace ResNet-100 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
Redis
PETSc
OpenFOAM:
  drivaerFastback, Small Mesh Size - Execution Time
  drivaerFastback, Small Mesh Size - Mesh Time
Dragonflydb
nginx
Apache HTTP Server
nginx
Apache HTTP Server
nginx
Appleseed
MariaDB mariadb-slap
Appleseed
Blender
Timed Linux Kernel Compilation
Renaissance
oneDNN
CloverLeaf
OSPRay Studio
oneDNN
PyTorch
TensorFlow
Chaos Group V-RAY
ASTC Encoder
OSPRay Studio
NAS Parallel Benchmarks
QMCPACK
ASKAP:
  tConvolve MT - Degridding
  tConvolve MT - Gridding
ONNX Runtime:
  Faster R-CNN R-50-FPN-int8 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
SPECFEM3D
OSPRay Studio:
  2 - 1080p - 1 - Path Tracer - CPU
  1 - 1080p - 1 - Path Tracer - CPU
OpenVINO:
  Person Detection FP16 - CPU:
    ms
    FPS
Stress-NG
OSPRay Studio
Llama.cpp
OSPRay Studio
OSPRay
Memcached:
  1:10
  1:100
Timed CPython Compilation
OSPRay:
  gravity_spheres_volume/dim_512/ao/real_time
  gravity_spheres_volume/dim_512/scivis/real_time
LuxCoreRender
OpenVINO:
  Face Detection FP16 - CPU:
    ms
    FPS
Zstd Compression:
  8, Long Mode - Decompression Speed
  8, Long Mode - Compression Speed
LuxCoreRender
OpenVINO:
  Face Detection FP16-INT8 - CPU:
    ms
    FPS
uvg266
ONNX Runtime:
  GPT-2 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
Xcompact3d Incompact3d
ONNX Runtime:
  bertsquad-12 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
LuxCoreRender
GraphicsMagick
OpenVINO:
  Machine Translation EN To DE FP16 - CPU:
    ms
    FPS
ONNX Runtime:
  T5 Encoder - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
OpenVINO:
  Person Detection FP32 - CPU:
    ms
    FPS
  Noise Suppression Poconet-Like FP16 - CPU:
    ms
    FPS
  Road Segmentation ADAS FP16-INT8 - CPU:
    ms
    FPS
GraphicsMagick:
  Enhanced
  Noise-Gaussian
OpenVINO:
  Person Vehicle Bike Detection FP16 - CPU:
    ms
    FPS
  Road Segmentation ADAS FP16 - CPU:
    ms
    FPS
  Person Re-Identification Retail FP16 - CPU:
    ms
    FPS
OSPRay Studio
OpenVINO:
  Handwritten English Recognition FP16-INT8 - CPU:
    ms
    FPS
Speedb
OpenVINO:
  Vehicle Detection FP16-INT8 - CPU:
    ms
    FPS
  Handwritten English Recognition FP16 - CPU:
    ms
    FPS
  Face Detection Retail FP16-INT8 - CPU:
    ms
    FPS
Speedb
OpenVINO:
  Age Gender Recognition Retail 0013 FP16-INT8 - CPU:
    ms
    FPS
  Vehicle Detection FP16 - CPU:
    ms
    FPS
  Weld Porosity Detection FP16 - CPU:
    ms
    FPS
Speedb
OpenVINO:
  Age Gender Recognition Retail 0013 FP16 - CPU:
    ms
    FPS
  Weld Porosity Detection FP16-INT8 - CPU:
    ms
    FPS
ONNX Runtime:
  CaffeNet 12-int8 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
  CaffeNet 12-int8 - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
OpenVINO:
  Face Detection Retail FP16 - CPU:
    ms
    FPS
John The Ripper
GraphicsMagick:
  Swirl
  Rotate
Speedb
GraphicsMagick
Speedb
GraphicsMagick
RocksDB
ONNX Runtime:
  ResNet50 v1-12-int8 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
RocksDB:
  Rand Fill Sync
  Overwrite
  Update Rand
Speedb
RocksDB:
  Read While Writing
  Read Rand Write Rand
ONNX Runtime:
  super-resolution-10 - CPU - Parallel:
    Inference Time Cost (ms)
    Inferences Per Second
John The Ripper
RocksDB
x265
SPECFEM3D
Liquid-DSP
Intel Open Image Denoise:
  RT.hdr_alb_nrm.3840x2160 - CPU-Only
  RT.ldr_alb_nrm.3840x2160 - CPU-Only
OSPRay Studio:
  2 - 1080p - 16 - Path Tracer - CPU
  1 - 1080p - 16 - Path Tracer - CPU
Apache Hadoop
Node.js V8 Web Tooling Benchmark
Z3 Theorem Prover
Timed PHP Compilation
Liquid-DSP
Cpuminer-Opt
libavif avifenc
m-queens
Embree
Coremark
RawTherapee
SVT-HEVC
Kvazaar
Renaissance
VVenC
JPEG-XL libjxl
Liquid-DSP
FFmpeg
Llama.cpp
Java SciMark
Renaissance
Embree
QuantLib
NAMD
Timed FFmpeg Compilation
TensorFlow
miniFE
Embree
Apache Hadoop
Timed Wasmer Compilation
Stress-NG
Cpuminer-Opt
DaCapo Benchmark
miniBUDE:
  OpenMP - BM1:
    Billion Interactions/s
    GFInst/s
SVT-VP9
TensorFlow
Cpuminer-Opt
JPEG-XL libjxl
VVenC
srsRAN Project
Aircrack-ng
John The Ripper:
  WPA PSK
  bcrypt
  Blowfish
Liquid-DSP
Cpuminer-Opt
Stress-NG:
  Vector Floating Point
  Matrix 3D Math
Liquid-DSP
Stress-NG:
  Floating Point
  Vector Shuffle
  Fused Multiply-Add
Liquid-DSP
Cpuminer-Opt
Stress-NG
Liquid-DSP:
  4 - 256 - 32
  16 - 256 - 32
  8 - 256 - 32
  1 - 256 - 32
Cpuminer-Opt
Stress-NG:
  AVX-512 VNNI
  x86_64 RdRand
Cpuminer-Opt:
  LBC, LBRY Credits
  Skeincoin
  Triple SHA-256, Onecoin
  Quad SHA-256, Pyrite
  Blake-2 S
srsRAN Project
WebP Image Encode
SVT-AV1
Llamafile
NAS Parallel Benchmarks
SVT-AV1
Y-Cruncher
JPEG-XL libjxl:
  JPEG - 100
  PNG - 100
DaCapo Benchmark
Timed Mesa Compilation
uvg266
DaCapo Benchmark:
  H2 Database Engine
  Apache Lucene Search Engine
Kvazaar
NAS Parallel Benchmarks
ASTC Encoder
TensorFlow
uvg266
LULESH
DaCapo Benchmark
oneDNN
7-Zip Compression:
  Decompression Rating
  Compression Rating
SVT-VP9
Kvazaar
DaCapo Benchmark
oneDNN
Z3 Theorem Prover
QuantLib
uvg266
DaCapo Benchmark
Kvazaar
Algebraic Multi-Grid Benchmark
DaCapo Benchmark
Parallel BZIP2 Compression
DaCapo Benchmark
LuxCoreRender
DaCapo Benchmark
Y-Cruncher
NAS Parallel Benchmarks
VVenC
Dolfyn
DaCapo Benchmark
uvg266
ASTC Encoder
DaCapo Benchmark
WebP Image Encode
x265
DaCapo Benchmark
SVT-AV1
easyWave
PyBench
SVT-HEVC
SVT-VP9
SVT-HEVC
ASKAP:
  tConvolve OpenMP - Degridding
  tConvolve OpenMP - Gridding
TensorFlow
DaCapo Benchmark
Primesieve
SVT-AV1
uvg266
oneDNN
DaCapo Benchmark
libavif avifenc:
  6
  6, Lossless
uvg266
SVT-AV1
Kvazaar
TensorFlow
RNNoise
SVT-AV1
srsRAN Project
oneDNN
Google Draco
uvg266
GNU Octave Benchmark
DaCapo Benchmark
WebP Image Encode
libavif avifenc
DaCapo Benchmark
Google Draco
Kvazaar:
  Bosphorus 1080p - Super Fast
  Bosphorus 1080p - Very Fast
DaCapo Benchmark
SVT-HEVC
Helsing
TensorFlow
SVT-VP9
oneDNN
LAMMPS Molecular Dynamics Simulator
srsRAN Project
Kvazaar
SVT-VP9:
  VMAF Optimized - Bosphorus 1080p
  PSNR/SSIM Optimized - Bosphorus 1080p
DaCapo Benchmark
SVT-HEVC
WebP Image Encode
SVT-AV1:
  Preset 13 - Bosphorus 1080p
  Preset 12 - Bosphorus 1080p
Timed CPython Compilation
WebP Image Encode
HeFFTe - Highly Efficient FFT for Exascale:
  r2c - FFTW - double - 128
  r2c - FFTW - float - 128
Java SciMark:
  Dense LU Matrix Factorization
  Monte Carlo
  Jacobi Successive Over-Relaxation
  Fast Fourier Transform