Intel Optimized Power Mode Xeon Platinum Benchmarks

2 x INTEL XEON PLATINUM 8592 testing by Michael Larabel for a future article.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2312153-NE-XEONEMRPO30
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:

C++ Boost Tests 3 Tests
Timed Code Compilation 3 Tests
C/C++ Compiler Tests 9 Tests
Compression Tests 2 Tests
CPU Massive 12 Tests
Creator Workloads 9 Tests
Database Test Suite 4 Tests
Encoding 6 Tests
Game Development 2 Tests
HPC - High Performance Computing 6 Tests
Java Tests 2 Tests
Common Kernel Benchmarks 2 Tests
Machine Learning 3 Tests
Multi-Core 15 Tests
Intel oneAPI 2 Tests
OpenMPI Tests 3 Tests
Programmer / Developer System Benchmarks 4 Tests
Python Tests 7 Tests
Scientific Computing 2 Tests
Server 8 Tests
Server CPU Tests 9 Tests
Single-Threaded 2 Tests
Video Encoding 6 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
Prefer Vertical Bar Graphs
No Box Plots
On Line Graphs With Missing Data, Connect The Line Gaps

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs
Condense Test Profiles With Multiple Version Results 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
Default
December 14 2023
  16 Hours, 25 Minutes
Optimized Power Mode
December 15 2023
  1 Day, 52 Minutes
Invert Hiding All Results Option
  20 Hours, 38 Minutes
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):

Default

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --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-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/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-build-config=bootstrap-lto-lean --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
Processor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000161
Java Notes: OpenJDK Runtime Environment (build 11.0.21+9-post-Ubuntu-0ubuntu123.10)
Python Notes: Python 3.11.6
Security Notes: 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

Optimized Power Mode

Processor: 2 x INTEL XEON PLATINUM 8592+ @ 3.90GHz (128 Cores / 256 Threads), Motherboard: Quanta Cloud S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 1008GB, Disk: 3201GB Micron_7450_MTFDKCB3T2TFS, Graphics: ASPEED, Network: 2 x Intel X710 for 10GBASE-T

OS: Ubuntu 23.10, Kernel: 6.5.0-13-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1080

Intel Optimized Power Mode Xeon Platinum BenchmarksOpenBenchmarking.orgPhoronix Test Suite2 x INTEL XEON PLATINUM 8592+ @ 3.90GHz (128 Cores / 256 Threads)Quanta Cloud S6Q-MB-MPS (3B05.TEL4P1 BIOS)Intel Device 1bce1008GB3201GB Micron_7450_MTFDKCB3T2TFSASPEED2 x Intel X710 for 10GBASE-TUbuntu 23.106.5.0-13-generic (x86_64)GCC 13.2.0ext41920x1080ProcessorMotherboardChipsetMemoryDiskGraphicsNetworkOSKernelCompilerFile-SystemScreen ResolutionIntel Optimized Power Mode Xeon Platinum Benchmarks PerformanceSystem Logs- Transparent Huge Pages: madvise- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --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-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/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-build-config=bootstrap-lto-lean --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 - Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000161- OpenJDK Runtime Environment (build 11.0.21+9-post-Ubuntu-0ubuntu123.10)- Python 3.11.6- 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

Default vs. Optimized Power Mode ComparisonPhoronix Test SuiteBaseline+13.5%+13.5%+27%+27%+40.5%+40.5%+54%+54%15%10.7%8.4%4.6%4.2%3.7%2.7%2.5%1:1054.1%CPU - 256 - ResNet-5041.2%CPU - 256 - ResNet-15239.3%CPU - 64 - ResNet-15238.5%50035.1%CPU - 64 - ResNet-5033.7%motorBike - Execution Time33.3%100031.7%GET - 5024.5%12 - Compression Speed22.6%CPU - 64 - Efficientnet_v2_l20.2%GhostRider - 1M19.4%Bosphorus 4K - Faster19.1%Bosphorus 4K - Ultra Fast18.3%CPU - 256 - Efficientnet_v2_l18.3%19 - Compression Speed17.9%10 - G.M.O.A.Q16.9%e.G.B.S - 120016.5%19, Long Mode - Compression Speed15.7%Barbershop - CPU-OnlyPreset 12 - Bosphorus 1080p14.4%e.G.B.S - 240014.4%IMDB14.2%Bosphorus 4K - Super Fast14.2%Bosphorus 4K - Very Fast14.2%Create - 100 - 10000014%50012.4%19 - D.S12.3%Preset 8 - Bosphorus 1080p12.3%Preset 8 - Bosphorus 4K12.2%libx265 - Live11.5%Bosphorus 1080p - Faster10.9%12 - D.S10.9%1B32 - 256 - 5710.5%Bosphorus 1080p - Fast10.3%VMAF Optimized - Bosphorus 1080p10.1%64 - 256 - 579.9%Preset 12 - Bosphorus 4K9.9%Preset 4 - Bosphorus 4K9.8%Preset 13 - Bosphorus 4K9.6%Preset 13 - Bosphorus 1080p9.4%R.O.R.S.I9.4%Bosphorus 4K - Ultra Fast9.4%Bosphorus 1080p - Very Fast9.2%Bosphorus 4K - Fast9.2%Bosphorus 1080p - Ultra Fast8.7%Bosphorus 1080p - Super Fast8.6%d.S.M.S - Mesh Time19, Long Mode - D.S8.2%P.S.O - Bosphorus 1080p8%Time To Compile7.5%100 - 1000 - Read Write - Average Latency7.2%100 - 1000 - Read Write7.1%V.Q.O - Bosphorus 1080p6.9%Bosphorus 1080p - Ultra Fast6.9%Bosphorus 4K - Very Fast6.8%Preset 4 - Bosphorus 1080p6.5%10005.9%Unix Makefiles5.9%TPC-H Parquet5.8%Bosphorus 4K - Super Fast5.7%Compression Rating5.4%libx265 - Platform5.2%1:1005.1%Bosphorus 1080p - Super Fast5%128 - 256 - 575%Bosphorus 1080p - Very Fast4.8%SET - 5004.8%R.5.S.I - A.M.S4.7%R.5.S.I - A.M.S4.7%Classroom - CPU-Onlylibx265 - Video On Demand4.2%RT.ldr_alb_nrm.3840x2160 - CPU-Onlylibx265 - Upload3.8%128 - 256 - 32A.G.R.R.0.F.I - CPU3.6%32 - 256 - 5123.4%256 - 256 - 573.2%defconfig3%GET - 5002.8%128 - 256 - 512A.G.R.R.0.F.I - CPU2.5%Bumper BeamNinja2.5%10 - Q2210.5%10 - Q2117.7%10 - Q2019.6%10 - Q1912.6%10 - Q1810.1%10 - Q173.4%10 - Q1613.8%10 - Q158.7%10 - Q1413.3%10 - Q1315.8%10 - Q1226.8%10 - Q1133.9%10 - Q1018.4%10 - Q0923.3%10 - Q0830.5%10 - Q0722.5%10 - Q0626.8%10 - Q0539.4%10 - Q0413.7%10 - Q0319.3%10 - Q0240.2%10 - Q0119%MemcachedPyTorchPyTorchPyTorchnginxPyTorchOpenFOAMnginxRedisZstd CompressionPyTorchXmrigVVenCuvg266PyTorchZstd CompressionApache Spark TPC-HeasyWaveZstd CompressionBlenderSVT-AV1easyWaveDuckDBuvg266uvg266Apache HadoopApache HTTP ServerZstd CompressionSVT-AV1SVT-AV1FFmpegVVenCZstd CompressionY-CruncherLiquid-DSPVVenCSVT-VP9Liquid-DSPSVT-AV1SVT-AV1SVT-AV1SVT-AV1OpenRadiossKvazaaruvg266VVenCuvg266uvg266OpenFOAMZstd CompressionSVT-VP9Timed GCC CompilationPostgreSQLPostgreSQLSVT-VP9KvazaarKvazaarSVT-AV1Apache HTTP ServerTimed LLVM CompilationDuckDBKvazaar7-Zip CompressionFFmpegMemcachedKvazaarLiquid-DSPKvazaarRedisNeural Magic DeepSparseNeural Magic DeepSparseBlenderFFmpegIntel Open Image DenoiseFFmpegLiquid-DSPOpenVINOLiquid-DSPLiquid-DSPTimed Linux Kernel CompilationRedisLiquid-DSPOpenVINOOpenRadiossTimed LLVM CompilationApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HDefaultOptimized Power Mode

Intel Optimized Power Mode Xeon Platinum Benchmarkspytorch: CPU - 256 - ResNet-50pytorch: CPU - 256 - ResNet-152pytorch: CPU - 64 - ResNet-152nginx: 500pytorch: CPU - 64 - ResNet-50openfoam: motorBike - Execution Timenginx: 1000redis: GET - 50compress-zstd: 12 - Compression Speedpytorch: CPU - 64 - Efficientnet_v2_lxmrig: GhostRider - 1Mvvenc: Bosphorus 4K - Fasteruvg266: Bosphorus 4K - Ultra Fastpytorch: CPU - 256 - Efficientnet_v2_lcompress-zstd: 19 - Compression Speedspark-tpch: 10 - Geometric Mean Of All Querieseasywave: e2Asean Grid + BengkuluSept2007 Source - 1200compress-zstd: 19, Long Mode - Compression Speedsvt-av1: Preset 12 - Bosphorus 1080pduckdb: IMDBuvg266: Bosphorus 4K - Super Fastuvg266: Bosphorus 4K - Very Fasthadoop: Create - 100 - 100000apache: 500compress-zstd: 19 - Decompression Speedsvt-av1: Preset 8 - Bosphorus 1080psvt-av1: Preset 8 - Bosphorus 4Kffmpeg: libx265 - Livevvenc: Bosphorus 1080p - Fastercompress-zstd: 12 - Decompression Speedy-cruncher: 1Bliquid-dsp: 32 - 256 - 57vvenc: Bosphorus 1080p - Fastliquid-dsp: 64 - 256 - 57svt-av1: Preset 12 - Bosphorus 4Ksvt-av1: Preset 4 - Bosphorus 4Ksvt-av1: Preset 13 - Bosphorus 4Ksvt-av1: Preset 13 - Bosphorus 1080popenradioss: Rubber O-Ring Seal Installationkvazaar: Bosphorus 4K - Ultra Fastuvg266: Bosphorus 1080p - Very Fastvvenc: Bosphorus 4K - Fastuvg266: Bosphorus 1080p - Ultra Fastuvg266: Bosphorus 1080p - Super Fastopenfoam: drivaerFastback, Small Mesh Size - Mesh Timecompress-zstd: 19, Long Mode - Decompression Speedsvt-vp9: PSNR/SSIM Optimized - Bosphorus 1080pbuild-gcc: Time To Compilepgbench: 100 - 1000 - Read Write - Average Latencypgbench: 100 - 1000 - Read Writesvt-vp9: Visual Quality Optimized - Bosphorus 1080pkvazaar: Bosphorus 1080p - Ultra Fastkvazaar: Bosphorus 4K - Very Fastsvt-av1: Preset 4 - Bosphorus 1080papache: 1000build-llvm: Unix Makefilesduckdb: TPC-H Parquetkvazaar: Bosphorus 4K - Super Fastcompress-7zip: Compression Ratingffmpeg: libx265 - Platformmemcached: 1:100kvazaar: Bosphorus 1080p - Super Fastliquid-dsp: 128 - 256 - 57kvazaar: Bosphorus 1080p - Very Fastredis: SET - 500deepsparse: ResNet-50, Sparse INT8 - Asynchronous Multi-Streamdeepsparse: ResNet-50, Sparse INT8 - Asynchronous Multi-Streamblender: Classroom - CPU-Onlyffmpeg: libx265 - Video On Demandoidn: RT.ldr_alb_nrm.3840x2160 - CPU-Onlyffmpeg: libx265 - Uploadliquid-dsp: 128 - 256 - 32openvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUliquid-dsp: 32 - 256 - 512liquid-dsp: 256 - 256 - 57build-linux-kernel: defconfigredis: GET - 500liquid-dsp: 128 - 256 - 512openvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUopenradioss: Bumper Beambuild-llvm: Ninjauvg266: Bosphorus 1080p - Mediumliquid-dsp: 256 - 256 - 32qmcpack: Li2_STO_aeuvg266: Bosphorus 4K - Slowblender: BMW27 - CPU-Onlyuvg266: Bosphorus 1080p - Slowxmrig: KawPow - 1Mkvazaar: Bosphorus 1080p - Slowopenradioss: Cell Phone Drop Testopenvino: Weld Porosity Detection FP16-INT8 - CPUdeepsparse: BERT-Large, NLP Question Answering - Asynchronous Multi-Streamdeepsparse: BERT-Large, NLP Question Answering - Asynchronous Multi-Streamliquid-dsp: 64 - 256 - 512openvino: Weld Porosity Detection FP16-INT8 - CPUdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streambuild-linux-kernel: allmodconfigopenvino: Handwritten English Recognition FP16-INT8 - CPUopenvino: Handwritten English Recognition FP16-INT8 - CPUxmrig: Wownero - 1Mdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Streamkvazaar: Bosphorus 4K - Slowquantlib: Multi-Threadedkvazaar: Bosphorus 4K - Mediumopenradioss: INIVOL and Fluid Structure Interaction Drop Containerliquid-dsp: 64 - 256 - 32liquid-dsp: 32 - 256 - 32deepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamuvg266: Bosphorus 4K - Mediumopenradioss: Chrysler Neon 1Mdeepsparse: ResNet-50, Baseline - Asynchronous Multi-Streamdeepsparse: ResNet-50, Baseline - Asynchronous Multi-Streamopenfoam: drivaerFastback, Small Mesh Size - Execution Timeopenvino: Face Detection Retail FP16-INT8 - CPUopenvino: Machine Translation EN To DE FP16 - CPUdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamopenvino: Machine Translation EN To DE FP16 - CPUopenradioss: Bird Strike on Windshielddeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamopenvino: Face Detection Retail FP16-INT8 - CPUkvazaar: Bosphorus 1080p - Mediumopenvino: Person Vehicle Bike Detection FP16 - CPUxmrig: CryptoNight-Heavy - 1Mopenvino: Person Vehicle Bike Detection FP16 - CPUredis: SET - 50openvino: Person Detection FP16 - CPUopenvino: Person Detection FP16 - CPUopenvino: Face Detection FP16-INT8 - CPUopenvino: Face Detection FP16-INT8 - CPUdeepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Streamquantlib: Single-Threadedxmrig: CryptoNight-Femto UPX2 - 1Mxmrig: Monero - 1Mdeepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Streamliquid-dsp: 256 - 256 - 512deepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamopenvino: Vehicle Detection FP16-INT8 - CPUdeepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Streamopenvino: Vehicle Detection FP16-INT8 - CPUdeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamcompress-7zip: Decompression Ratingopenvino: Road Segmentation ADAS FP16-INT8 - CPUdeepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Streamopenvino: Road Segmentation ADAS FP16-INT8 - CPUdeepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Streamdeepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Streamblender: Barbershop - CPU-Onlypgbench: 100 - 1000 - Read Only - Average Latencypgbench: 100 - 1000 - Read Onlymemcached: 1:10spark-tpch: 10 - Q22spark-tpch: 10 - Q21spark-tpch: 10 - Q20spark-tpch: 10 - Q19spark-tpch: 10 - Q18spark-tpch: 10 - Q17spark-tpch: 10 - Q16spark-tpch: 10 - Q15spark-tpch: 10 - Q14spark-tpch: 10 - Q13spark-tpch: 10 - Q12spark-tpch: 10 - Q11spark-tpch: 10 - Q10spark-tpch: 10 - Q09spark-tpch: 10 - Q08spark-tpch: 10 - Q07spark-tpch: 10 - Q06spark-tpch: 10 - Q05spark-tpch: 10 - Q04spark-tpch: 10 - Q03spark-tpch: 10 - Q02spark-tpch: 10 - Q01svt-vp9: VMAF Optimized - Bosphorus 1080peasywave: e2Asean Grid + BengkuluSept2007 Source - 2400DefaultOptimized Power Mode44.5817.8517.67273684.7143.914.13288243384.194744838.9333.62.6216522.510.83659.592.5919.18.3591968336.5469.90502.707121.96757.8756.06549090823.061173.5142.26569.466131.8132.0911422.75.199142910000019.9152810166667218.4527.297218.763628.75581.6977.53160.636.921166.47158.2330.042231183.8542.17706.25915.63463962463.27282.0569.3620.95280826.71177.431148.5571.0568951653.983441439.71263.884329866667268.492503177.6711244.79415.673234.9453.634.5327.003611866667121577.55508245000583550000023.7563240001.5814739000000.4084.7896.51888.46618560000094.38427.2012.8980.3070383.2127.2525.7448795.97160.4064397.773110123366672.45350.9008149.79238.413330.1975002.3182.15911791.27371899.222335.669533.638541.45388447.241.9997.8724760333331211766667137.354029.7985.8235.93141778.594523.46982224795.171106.32462.9932137.343928.85111.23463.08545.15131.1410205.6170573.012.513060186.42748.1042.72236.70539.6913.94513650.569677.870469.84580.6318214456666751.70338930.00862.138214.321235.406363731153.5174.07222390.2576.5445834.1767149.491.1248930893360511.124.4956595126.481879558.561719895.0945846212.8941656812.600738214.972441044.245615805.778444925.644020567.933933265.9976617511.8791221017.5612882012.0530392311.474618282.5661711712.944283498.1769722312.590769777.698114088.84185569542.9889.98431.5812.8112.76202507.4232.835.50897184861.993811819.50272.22.1813832.49.09650.362.1916.29.7730412742.5598.56439.477139.32650.6849.10481480817.731044.6126.69761.914118.2428.9261282.74.697129333333318.0532556573333198.7426.645199.549574.59989.3770.87147.086.340153.17145.7527.708731094.2502.05759.21616.75259695433.53263.9664.9719.66976315.65187.837157.10167.2165445951.313274425.32251.334125000000256.192389636.6710738.13525.938533.4051.454.7226.023745233333117369.55491426667565660000024.4763152359.7515134666670.4182.7398.90586.78630148333392.67226.7112.6778.9369196.7125.1825.3749494.06162.5895392.45159996633332.42346.7411151.27638.793297.8175701.7183.71041776.52521883.657335.956133.903241.13391468.441.6998.4924608666671204500000138.149729.6286.3036.12861768.910823.34883324922.401100.76460.6851138.011428.99111.73461.14695.13130.6410168.0670338.612.553050498.00750.4342.59237.36538.2013.90953641.469847.970298.84590.3375214100000051.61768943.88863.347414.301237.025763813653.4573.98962392.3876.5042834.2706129.941.1418784682180648.214.9695701631.1611911210.242673195.7373820014.1998758313.028431495.656485284.614900046.548335146.5348194710.057694168.0304369914.0628064021.6599576115.7232208214.057704933.2531764218.041547789.2947692915.0188730810.7942171110.52531964493.26102.928OpenBenchmarking.org

PyTorch

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: ResNet-50DefaultOptimized Power Mode1020304050SE +/- 0.51, N = 4SE +/- 0.02, N = 344.5831.58MIN: 19.44 / MAX: 46.72MIN: 14.91 / MAX: 34.67
OpenBenchmarking.orgbatches/sec Per Watt, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: ResNet-50DefaultOptimized Power Mode0.01910.03820.05730.07640.09550.0850.079
MinAvgMaxDefault80029513908Optimized Power Mode80029433906OpenBenchmarking.orgMegahertz, More Is BetterPyTorch 2.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207.1523.6583.9Optimized Power Mode181.1397.8477.5OpenBenchmarking.orgWatts, Fewer Is BetterPyTorch 2.1CPU Power Consumption Monitor1603204806408001. Default: Approximate power consumption of 28277 Joules per run.2. Optimized Power Mode: Approximate power consumption of 29038 Joules per run.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: ResNet-152DefaultOptimized Power Mode48121620SE +/- 0.13, N = 11SE +/- 0.17, N = 317.8512.81MIN: 7.02 / MAX: 19.19MIN: 6.6 / MAX: 13.9
OpenBenchmarking.orgbatches/sec Per Watt, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: ResNet-152DefaultOptimized Power Mode0.00720.01440.02160.02880.0360.0320.030
OpenBenchmarking.orgMegahertz, More Is BetterPyTorch 2.1CPU Peak Freq (Highest CPU Core Frequency) MonitorDefaultOptimized Power Mode7001400210028003500Min: 800 / Avg: 2937.97 / Max: 3906Min: 800 / Avg: 2935.94 / Max: 3905
OpenBenchmarking.orgWatts, Fewer Is BetterPyTorch 2.1CPU Power Consumption MonitorDefaultOptimized Power Mode100200300400500Min: 206.82 / Avg: 549.56 / Max: 589.42Min: 181 / Avg: 425.58 / Max: 481.371. Default: Approximate power consumption of 69045 Joules per run.2. Optimized Power Mode: Approximate power consumption of 73200 Joules per run.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 64 - Model: ResNet-152DefaultOptimized Power Mode48121620SE +/- 0.23, N = 3SE +/- 0.12, N = 1217.6712.76MIN: 10.89 / MAX: 18.38MIN: 5.75 / MAX: 15.71
OpenBenchmarking.orgbatches/sec Per Watt, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 64 - Model: ResNet-152DefaultOptimized Power Mode0.00720.01440.02160.02880.0360.0320.030
MinAvgMaxDefault80029133907Optimized Power Mode80029233908OpenBenchmarking.orgMegahertz, More Is BetterPyTorch 2.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.9550.2586.6Optimized Power Mode181.0419.8482.2OpenBenchmarking.orgWatts, Fewer Is BetterPyTorch 2.1CPU Power Consumption Monitor1603204806408001. Default: Approximate power consumption of 69509 Joules per run.2. Optimized Power Mode: Approximate power consumption of 72165 Joules per run.

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: 500DefaultOptimized Power Mode60K120K180K240K300KSE +/- 25.30, N = 3SE +/- 10.52, N = 3273684.71202507.421. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenBenchmarking.orgRequests Per Second Per Watt, More Is Betternginx 1.23.2Connections: 500DefaultOptimized Power Mode100200300400500449.51429.69
MinAvgMaxDefault80029033904Optimized Power Mode80029253905OpenBenchmarking.orgMegahertz, More Is Betternginx 1.23.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault199609655Optimized Power Mode183471554OpenBenchmarking.orgWatts, Fewer Is Betternginx 1.23.2CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 59261 Joules per run.2. Optimized Power Mode: Approximate power consumption of 46030 Joules per run.

PyTorch

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 64 - Model: ResNet-50DefaultOptimized Power Mode1020304050SE +/- 0.55, N = 4SE +/- 0.39, N = 343.9132.83MIN: 17.29 / MAX: 46.23MIN: 15.85 / MAX: 36.48
OpenBenchmarking.orgbatches/sec Per Watt, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 64 - Model: ResNet-50DefaultOptimized Power Mode0.01910.03820.05730.07640.09550.0850.082
MinAvgMaxDefault80029803907Optimized Power Mode80030083905OpenBenchmarking.orgMegahertz, More Is BetterPyTorch 2.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207.1519.2581.2Optimized Power Mode181.0398.9479.9OpenBenchmarking.orgWatts, Fewer Is BetterPyTorch 2.1CPU Power Consumption Monitor1603204806408001. Default: Approximate power consumption of 28427 Joules per run.2. Optimized Power Mode: Approximate power consumption of 28718 Joules per run.

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: motorBike - Execution TimeDefaultOptimized Power Mode1.23952.4793.71854.9586.19754.132885.508971. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm
MinAvgMaxDefault80035753908Optimized Power Mode388939003908OpenBenchmarking.orgMegahertz, More Is BetterOpenFOAM 10CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.9278.4337.3Optimized Power Mode175.6200.2221.0OpenBenchmarking.orgWatts, Fewer Is BetterOpenFOAM 10CPU Power Consumption Monitor80160240320400

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: 1000DefaultOptimized Power Mode50K100K150K200K250KSE +/- 397.04, N = 3SE +/- 72.63, N = 2243384.19184861.991. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenBenchmarking.orgRequests Per Second Per Watt, More Is Betternginx 1.23.2Connections: 1000DefaultOptimized Power Mode90180270360450389.82395.64
MinAvgMaxDefault80028693905Optimized Power Mode80029663907OpenBenchmarking.orgMegahertz, More Is Betternginx 1.23.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207624666Optimized Power Mode183467530OpenBenchmarking.orgWatts, Fewer Is Betternginx 1.23.2CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 60771 Joules per run.2. Optimized Power Mode: Approximate power consumption of 48360 Joules per run.

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: GET - Parallel Connections: 50DefaultOptimized Power Mode1000K2000K3000K4000K5000KSE +/- 8167.04, N = 4SE +/- 1281.58, N = 34744838.903811819.501. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3
OpenBenchmarking.orgRequests Per Second Per Watt, More Is BetterRedis 7.0.4Test: GET - Parallel Connections: 50DefaultOptimized Power Mode4K8K12K16K20K18984.6120690.41
MinAvgMaxDefault80037063905Optimized Power Mode80037563906OpenBenchmarking.orgMegahertz, More Is BetterRedis 7.0.4CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.0249.9290.9Optimized Power Mode94.4184.2188.8OpenBenchmarking.orgWatts, Fewer Is BetterRedis 7.0.4CPU Power Consumption Monitor701402102803501. Default: Approximate power consumption of 5249 Joules per run.2. Optimized Power Mode: Approximate power consumption of 4422 Joules per run.

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 - Compression SpeedDefaultOptimized Power Mode70140210280350SE +/- 3.46, N = 3SE +/- 2.98, N = 5333.6272.21. (CC) gcc options: -O3 -pthread -lz -llzma

PyTorch

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 64 - Model: Efficientnet_v2_lDefaultOptimized Power Mode0.58951.1791.76852.3582.9475SE +/- 0.02, N = 3SE +/- 0.02, N = 32.622.18MIN: 1.32 / MAX: 3.9MIN: 0.98 / MAX: 3.41
OpenBenchmarking.orgbatches/sec Per Watt, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 64 - Model: Efficientnet_v2_lDefaultOptimized Power Mode0.00160.00320.00480.00640.0080.0070.007
OpenBenchmarking.orgMegahertz, More Is BetterPyTorch 2.1CPU Peak Freq (Highest CPU Core Frequency) MonitorDefaultOptimized Power Mode9001800270036004500Min: 800 / Avg: 3771.05 / Max: 4127Min: 800 / Avg: 3770.85 / Max: 5154
OpenBenchmarking.orgWatts, Fewer Is BetterPyTorch 2.1CPU Power Consumption MonitorDefaultOptimized Power Mode80160240320400Min: 191.62 / Avg: 400.65 / Max: 444.68Min: 151.88 / Avg: 323.77 / Max: 364.191. Default: Approximate power consumption of 793148 Joules per run.2. Optimized Power Mode: Approximate power consumption of 699024 Joules per run.

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: 1MDefaultOptimized Power Mode4K8K12K16K20KSE +/- 67.39, N = 3SE +/- 22.03, N = 316522.513832.41. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
OpenBenchmarking.orgH/s Per Watt, More Is BetterXmrig 6.21Variant: GhostRider - Hash Count: 1MDefaultOptimized Power Mode61218243025.2127.32
MinAvgMaxDefault80029383905Optimized Power Mode80029393904OpenBenchmarking.orgMegahertz, More Is BetterXmrig 6.21CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault204655739Optimized Power Mode175506548OpenBenchmarking.orgWatts, Fewer Is BetterXmrig 6.21CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 43914 Joules per run.2. Optimized Power Mode: Approximate power consumption of 40005 Joules per run.

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 1.9Video Input: Bosphorus 4K - Video Preset: FasterDefaultOptimized Power Mode3691215SE +/- 0.081, N = 3SE +/- 0.072, N = 310.8369.0961. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterVVenC 1.9Video Input: Bosphorus 4K - Video Preset: FasterDefaultOptimized Power Mode0.0070.0140.0210.0280.0350.0270.031
MinAvgMaxDefault80038613907Optimized Power Mode80038153907OpenBenchmarking.orgMegahertz, More Is BetterVVenC 1.9CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207.0407.8523.1Optimized Power Mode135.6296.9410.2OpenBenchmarking.orgWatts, Fewer Is BetterVVenC 1.9CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 26505 Joules per run.2. Optimized Power Mode: Approximate power consumption of 22664 Joules per run.

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 FastDefaultOptimized Power Mode1326395265SE +/- 0.37, N = 5SE +/- 0.21, N = 459.5950.36
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Ultra FastDefaultOptimized Power Mode0.03580.07160.10740.14320.1790.1450.159
MinAvgMaxDefault80035093921Optimized Power Mode80037063922OpenBenchmarking.orgMegahertz, More Is Betteruvg266 0.4.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207.2411.1535.8Optimized Power Mode183.0316.8389.0OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor1402804205607001. Default: Approximate power consumption of 6002 Joules per run.2. Optimized Power Mode: Approximate power consumption of 5306 Joules per run.

PyTorch

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: Efficientnet_v2_lDefaultOptimized Power Mode0.58281.16561.74842.33122.914SE +/- 0.00, N = 3SE +/- 0.02, N = 32.592.19MIN: 1.19 / MAX: 4MIN: 0.98 / MAX: 3.3
OpenBenchmarking.orgbatches/sec Per Watt, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: Efficientnet_v2_lDefaultOptimized Power Mode0.00160.00320.00480.00640.0080.0060.007
OpenBenchmarking.orgMegahertz, More Is BetterPyTorch 2.1CPU Peak Freq (Highest CPU Core Frequency) MonitorDefaultOptimized Power Mode8001600240032004000Min: 800 / Avg: 3773.77 / Max: 3934Min: 800 / Avg: 3772.48 / Max: 4480
OpenBenchmarking.orgWatts, Fewer Is BetterPyTorch 2.1CPU Power Consumption MonitorDefaultOptimized Power Mode80160240320400Min: 188.24 / Avg: 400.23 / Max: 457.62Min: 150.79 / Avg: 323.75 / Max: 378.291. Default: Approximate power consumption of 776446 Joules per run.2. Optimized Power Mode: Approximate power consumption of 699740 Joules per run.

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 - Compression SpeedDefaultOptimized Power Mode510152025SE +/- 0.06, N = 3SE +/- 0.19, N = 319.116.21. (CC) gcc options: -O3 -pthread -lz -llzma

Apache Spark TPC-H

This is a benchmark of Apache Spark using TPC-H data-set. Apache Spark is an open-source unified analytics engine for large-scale data processing and dealing with big data. This test profile benchmarks the Apache Spark in a single-system configuration using spark-submit. The test makes use of https://github.com/ssavvides/tpch-spark/ for facilitating the TPC-H benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Geometric Mean Of All QueriesDefaultOptimized Power Mode3691215SE +/- 0.07178690, N = 3SE +/- 0.09145601, N = 78.359196839.77304127MIN: 4.15 / MAX: 27.27MIN: 4.46 / MAX: 32.81
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q01DefaultOptimized Power Mode3691215SE +/- 0.23680200, N = 3SE +/- 0.34672923, N = 78.8418556910.52531964
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q02DefaultOptimized Power Mode3691215SE +/- 0.14674980, N = 3SE +/- 0.18560870, N = 77.6981140810.79421711
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q03DefaultOptimized Power Mode48121620SE +/- 0.05, N = 3SE +/- 0.30, N = 712.5915.02
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q04DefaultOptimized Power Mode3691215SE +/- 0.27780883, N = 3SE +/- 0.18685247, N = 78.176972239.29476929
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q05DefaultOptimized Power Mode48121620SE +/- 0.17, N = 3SE +/- 0.84, N = 712.9418.04
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q06DefaultOptimized Power Mode0.7321.4642.1962.9283.66SE +/- 0.19354811, N = 3SE +/- 0.19375001, N = 72.566171173.25317642
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q07DefaultOptimized Power Mode48121620SE +/- 0.65, N = 3SE +/- 0.30, N = 711.4714.06
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q08DefaultOptimized Power Mode48121620SE +/- 0.29, N = 3SE +/- 0.45, N = 712.0515.72
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q09DefaultOptimized Power Mode510152025SE +/- 1.37, N = 3SE +/- 0.36, N = 717.5621.66
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q10DefaultOptimized Power Mode48121620SE +/- 0.36, N = 3SE +/- 0.34, N = 711.8814.06
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q11DefaultOptimized Power Mode246810SE +/- 0.21358847, N = 3SE +/- 0.21479648, N = 75.997661758.03043699
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q12DefaultOptimized Power Mode3691215SE +/- 0.47598391, N = 3SE +/- 0.34240704, N = 77.9339332610.05769416
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q13DefaultOptimized Power Mode246810SE +/- 0.08149193, N = 3SE +/- 0.30330721, N = 75.644020566.53481947
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q14DefaultOptimized Power Mode246810SE +/- 0.09561638, N = 3SE +/- 0.14967485, N = 75.778444926.54833514
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q15DefaultOptimized Power Mode1.03842.07683.11524.15365.192SE +/- 0.05108329, N = 3SE +/- 0.07270001, N = 74.245615804.61490004
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q16DefaultOptimized Power Mode1.27272.54543.81815.09086.3635SE +/- 0.28343850, N = 3SE +/- 0.10743013, N = 74.972441045.65648528
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q17DefaultOptimized Power Mode3691215SE +/- 0.23, N = 3SE +/- 0.26, N = 712.6013.03
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q18DefaultOptimized Power Mode48121620SE +/- 0.44, N = 3SE +/- 0.53, N = 712.8914.20
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q19DefaultOptimized Power Mode1.29092.58183.87275.16366.4545SE +/- 0.17083033, N = 3SE +/- 0.14465657, N = 75.094584625.73738200
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q20DefaultOptimized Power Mode3691215SE +/- 0.45558852, N = 3SE +/- 0.14303850, N = 78.5617198910.24267319
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q21DefaultOptimized Power Mode714212835SE +/- 0.40, N = 3SE +/- 0.34, N = 726.4831.16
OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Q22DefaultOptimized Power Mode1.11822.23643.35464.47285.591SE +/- 0.16829529, N = 3SE +/- 0.13159559, N = 74.495659514.96957016
MinAvgMaxDefault80038375474Optimized Power Mode80038344663OpenBenchmarking.orgMegahertz, More Is BetterApache Spark TPC-H 3.5CPU Peak Freq (Highest CPU Core Frequency) Monitor14002800420056007000
MinAvgMaxDefault145.9378.9600.0Optimized Power Mode92.8308.2608.0OpenBenchmarking.orgWatts, Fewer Is BetterApache Spark TPC-H 3.5CPU Power Consumption Monitor1603204806408001. Default: Approximate power consumption of 99525 Joules per run.2. Optimized Power Mode: Approximate power consumption of 87523 Joules per run.

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: 1200DefaultOptimized Power Mode1020304050SE +/- 0.28, N = 3SE +/- 0.41, N = 1536.5542.561. (CXX) g++ options: -O3 -fopenmp
MinAvgMaxDefault80034693908Optimized Power Mode80034503907OpenBenchmarking.orgMegahertz, More Is BettereasyWave r34CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault199.5390.6501.4Optimized Power Mode168.2327.1497.9OpenBenchmarking.orgWatts, Fewer Is BettereasyWave r34CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 16795 Joules per run.2. Optimized Power Mode: Approximate power consumption of 15787 Joules per run.

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 - Compression SpeedDefaultOptimized Power Mode3691215SE +/- 0.02, N = 3SE +/- 0.01, N = 39.908.561. (CC) gcc options: -O3 -pthread -lz -llzma

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 12 - Input: Bosphorus 1080pDefaultOptimized Power Mode110220330440550SE +/- 3.53, N = 9SE +/- 4.01, N = 15502.71439.481. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 12 - Input: Bosphorus 1080pDefaultOptimized Power Mode0.48560.97121.45681.94242.4282.0422.158
MinAvgMaxDefault80035223920Optimized Power Mode80035773907OpenBenchmarking.orgMegahertz, More Is BetterSVT-AV1 1.8CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault195.6246.2402.6Optimized Power Mode119.4203.6319.3OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.8CPU Power Consumption Monitor1102203304405501. Default: Approximate power consumption of 1751 Joules per run.2. Optimized Power Mode: Approximate power consumption of 1480 Joules per run.

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: IMDBDefaultOptimized Power Mode306090120150SE +/- 0.49, N = 3SE +/- 0.43, N = 3121.97139.331. (CXX) g++ options: -O3 -rdynamic -lssl -lcrypto -ldl
MinAvgMaxDefault80038733927Optimized Power Mode80038633920OpenBenchmarking.orgMegahertz, More Is BetterDuckDB 0.9.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault160.7312.8391.9Optimized Power Mode101.6224.7314.2OpenBenchmarking.orgWatts, Fewer Is BetterDuckDB 0.9.1CPU Power Consumption Monitor1102203304405501. Default: Approximate power consumption of 61207 Joules per run.2. Optimized Power Mode: Approximate power consumption of 42916 Joules per run.

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 FastDefaultOptimized Power Mode1326395265SE +/- 0.15, N = 5SE +/- 0.48, N = 457.8750.68
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Super FastDefaultOptimized Power Mode0.03240.06480.09720.12960.1620.1310.144
MinAvgMaxDefault80037193916Optimized Power Mode80036603923OpenBenchmarking.orgMegahertz, More Is Betteruvg266 0.4.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.2440.6569.5Optimized Power Mode188.7351.0476.3OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor1402804205607001. Default: Approximate power consumption of 6521 Joules per run.2. Optimized Power Mode: Approximate power consumption of 5704 Joules per run.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Very FastDefaultOptimized Power Mode1326395265SE +/- 0.07, N = 5SE +/- 0.45, N = 1556.0649.10
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Very FastDefaultOptimized Power Mode0.03020.06040.09060.12080.1510.1250.134
MinAvgMaxDefault80036343920Optimized Power Mode80036673928OpenBenchmarking.orgMegahertz, More Is Betteruvg266 0.4.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault199.1448.8588.6Optimized Power Mode182.9366.0531.1OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor1603204806408001. Default: Approximate power consumption of 6821 Joules per run.2. Optimized Power Mode: Approximate power consumption of 6027 Joules per run.

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: 100000DefaultOptimized Power Mode12002400360048006000SE +/- 32.35, N = 3SE +/- 62.35, N = 354904814
OpenBenchmarking.orgOps per sec Per Watt, More Is BetterApache Hadoop 3.3.6Operation: Create - Threads: 100 - Files: 100000DefaultOptimized Power Mode61218243021.2425.29
MinAvgMaxDefault80035723911Optimized Power Mode80034803907OpenBenchmarking.orgMegahertz, More Is BetterApache Hadoop 3.3.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault205.8258.4316.7Optimized Power Mode93.8190.3214.7OpenBenchmarking.orgWatts, Fewer Is BetterApache Hadoop 3.3.6CPU Power Consumption Monitor801602403204001. Default: Approximate power consumption of 15506 Joules per run.2. Optimized Power Mode: Approximate power consumption of 12182 Joules per run.

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: 500DefaultOptimized Power Mode20K40K60K80K100KSE +/- 265.12, N = 3SE +/- 184.86, N = 390823.0680817.731. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenBenchmarking.orgRequests Per Second Per Watt, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 500DefaultOptimized Power Mode60120180240300216.64266.08
MinAvgMaxDefault80028903905Optimized Power Mode80028893907OpenBenchmarking.orgMegahertz, More Is BetterApache HTTP Server 2.4.56CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault196.4419.2442.6Optimized Power Mode181.8303.7320.2OpenBenchmarking.orgWatts, Fewer Is BetterApache HTTP Server 2.4.56CPU Power Consumption Monitor1202403604806001. Default: Approximate power consumption of 40806 Joules per run.2. Optimized Power Mode: Approximate power consumption of 29664 Joules per run.

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 SpeedDefaultOptimized Power Mode30060090012001500SE +/- 1.19, N = 3SE +/- 2.93, N = 31173.51044.61. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s Per Watt, More Is BetterZstd Compression 1.5.4Compression Level: 19 - Decompression SpeedDefaultOptimized Power Mode1.19682.39363.59044.78725.9843.9535.319
MinAvgMaxDefault80038813905Optimized Power Mode80038603907OpenBenchmarking.orgMegahertz, More Is BetterZstd Compression 1.5.4CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault164.0296.9335.4Optimized Power Mode98.9196.4223.8OpenBenchmarking.orgWatts, Fewer Is BetterZstd Compression 1.5.4CPU Power Consumption Monitor801602403204001. Default: Approximate power consumption of 21176 Joules per run.2. Optimized Power Mode: Approximate power consumption of 15253 Joules per run.

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 8 - Input: Bosphorus 1080pDefaultOptimized Power Mode306090120150SE +/- 0.36, N = 7SE +/- 0.77, N = 6142.27126.701. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 8 - Input: Bosphorus 1080pDefaultOptimized Power Mode0.12620.25240.37860.50480.6310.4780.561
MinAvgMaxDefault80035453922Optimized Power Mode80034393938OpenBenchmarking.orgMegahertz, More Is BetterSVT-AV1 1.8CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault199.8297.7432.0Optimized Power Mode178.6225.9330.4OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.8CPU Power Consumption Monitor1102203304405501. Default: Approximate power consumption of 3020 Joules per run.2. Optimized Power Mode: Approximate power consumption of 2523 Joules per run.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 8 - Input: Bosphorus 4KDefaultOptimized Power Mode1530456075SE +/- 0.39, N = 4SE +/- 0.56, N = 469.4761.911. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 8 - Input: Bosphorus 4KDefaultOptimized Power Mode0.05060.10120.15180.20240.2530.1920.225
MinAvgMaxDefault80038443916Optimized Power Mode80035363907OpenBenchmarking.orgMegahertz, More Is BetterSVT-AV1 1.8CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207.4361.1517.2Optimized Power Mode178.6275.5394.3OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.8CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 5778 Joules per run.2. Optimized Power Mode: Approximate power consumption of 4683 Joules per run.

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: LiveDefaultOptimized Power Mode306090120150SE +/- 1.03, N = 3SE +/- 0.79, N = 3131.81118.241. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFPS Per Watt, More Is BetterFFmpeg 6.1Encoder: libx265 - Scenario: LiveDefaultOptimized Power Mode0.12960.25920.38880.51840.6480.4390.576
MinAvgMaxDefault80038503920Optimized Power Mode80038243917OpenBenchmarking.orgMegahertz, More Is BetterFFmpeg 6.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault198.9300.2361.3Optimized Power Mode129.0205.4267.0OpenBenchmarking.orgWatts, Fewer Is BetterFFmpeg 6.1CPU Power Consumption Monitor1002003004005001. Default: Approximate power consumption of 18311 Joules per run.2. Optimized Power Mode: Approximate power consumption of 13694 Joules per run.

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 1.9Video Input: Bosphorus 1080p - Video Preset: FasterDefaultOptimized Power Mode714212835SE +/- 0.36, N = 3SE +/- 0.30, N = 432.0928.931. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterVVenC 1.9Video Input: Bosphorus 1080p - Video Preset: FasterDefaultOptimized Power Mode0.02590.05180.07770.10360.12950.0920.115
MinAvgMaxDefault80036463910Optimized Power Mode80037233906OpenBenchmarking.orgMegahertz, More Is BetterVVenC 1.9CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.9349.1415.3Optimized Power Mode182.4250.7297.9OpenBenchmarking.orgWatts, Fewer Is BetterVVenC 1.9CPU Power Consumption Monitor1102203304405501. Default: Approximate power consumption of 10009 Joules per run.2. Optimized Power Mode: Approximate power consumption of 7645 Joules per run.

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 SpeedDefaultOptimized Power Mode30060090012001500SE +/- 1.98, N = 3SE +/- 1.74, N = 51422.71282.71. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s Per Watt, More Is BetterZstd Compression 1.5.4Compression Level: 12 - Decompression SpeedDefaultOptimized Power Mode2468104.5166.105
MinAvgMaxDefault80038823910Optimized Power Mode80038443905OpenBenchmarking.orgMegahertz, More Is BetterZstd Compression 1.5.4CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault143.9315.1371.3Optimized Power Mode90.4210.1264.7OpenBenchmarking.orgWatts, Fewer Is BetterZstd Compression 1.5.4CPU Power Consumption Monitor1002003004005001. Default: Approximate power consumption of 21214 Joules per run.2. Optimized Power Mode: Approximate power consumption of 14245 Joules per run.

Y-Cruncher

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.2Pi Digits To Calculate: 1BDefaultOptimized Power Mode1.16982.33963.50944.67925.849SE +/- 0.018, N = 5SE +/- 0.009, N = 55.1994.697
MinAvgMaxDefault80031363912Optimized Power Mode80032353905OpenBenchmarking.orgMegahertz, More Is BetterY-Cruncher 0.8.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207398728Optimized Power Mode97361742OpenBenchmarking.orgWatts, Fewer Is BetterY-Cruncher 0.8.2CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 5100 Joules per run.2. Optimized Power Mode: Approximate power consumption of 4691 Joules per run.

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: 32 - Buffer Length: 256 - Filter Length: 57DefaultOptimized Power Mode300M600M900M1200M1500MSE +/- 9950041.88, N = 3SE +/- 16392511.84, N = 3142910000012933333331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 32 - Buffer Length: 256 - Filter Length: 57DefaultOptimized Power Mode800K1600K2400K3200K4000K3194605.473765376.51
MinAvgMaxDefault80036073906Optimized Power Mode80037763906OpenBenchmarking.orgMegahertz, More Is BetterLiquid-DSP 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207.2447.3490.6Optimized Power Mode184.2343.5374.7OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 15508 Joules per run.2. Optimized Power Mode: Approximate power consumption of 11907 Joules per run.

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 1.9Video Input: Bosphorus 1080p - Video Preset: FastDefaultOptimized Power Mode510152025SE +/- 0.15, N = 3SE +/- 0.11, N = 319.9218.051. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterVVenC 1.9Video Input: Bosphorus 1080p - Video Preset: FastDefaultOptimized Power Mode0.01510.03020.04530.06040.07550.0530.067
MinAvgMaxDefault80038483908Optimized Power Mode80037803905OpenBenchmarking.orgMegahertz, More Is BetterVVenC 1.9CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.6376.5448.9Optimized Power Mode160.6269.5332.4OpenBenchmarking.orgWatts, Fewer Is BetterVVenC 1.9CPU Power Consumption Monitor1202403604806001. Default: Approximate power consumption of 14933 Joules per run.2. Optimized Power Mode: Approximate power consumption of 11677 Joules per run.

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: 64 - Buffer Length: 256 - Filter Length: 57DefaultOptimized Power Mode600M1200M1800M2400M3000MSE +/- 24169149.30, N = 3SE +/- 20734086.19, N = 15281016666725565733331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 64 - Buffer Length: 256 - Filter Length: 57DefaultOptimized Power Mode1000K2000K3000K4000K5000K4449269.384774859.27
MinAvgMaxDefault80038433905Optimized Power Mode80038093907OpenBenchmarking.orgMegahertz, More Is BetterLiquid-DSP 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault208632708Optimized Power Mode183535623OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 21896 Joules per run.2. Optimized Power Mode: Approximate power consumption of 18276 Joules per run.

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 12 - Input: Bosphorus 4KDefaultOptimized Power Mode50100150200250SE +/- 0.90, N = 6SE +/- 2.06, N = 5218.45198.741. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 12 - Input: Bosphorus 4KDefaultOptimized Power Mode0.20180.40360.60540.80721.0090.8080.897
MinAvgMaxDefault80036093920Optimized Power Mode80037003906OpenBenchmarking.orgMegahertz, More Is BetterSVT-AV1 1.8CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault194.7270.4503.0Optimized Power Mode178.4221.4396.1OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.8CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 3380 Joules per run.2. Optimized Power Mode: Approximate power consumption of 2967 Joules per run.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 4 - Input: Bosphorus 4KDefaultOptimized Power Mode246810SE +/- 0.098, N = 3SE +/- 0.006, N = 37.2976.6451. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 4 - Input: Bosphorus 4KDefaultOptimized Power Mode0.00560.01120.01680.02240.0280.0200.025
MinAvgMaxDefault80037303906Optimized Power Mode80037443915OpenBenchmarking.orgMegahertz, More Is BetterSVT-AV1 1.8CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault198.8363.8531.7Optimized Power Mode178.5264.7423.6OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.8CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 11035 Joules per run.2. Optimized Power Mode: Approximate power consumption of 8735 Joules per run.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 13 - Input: Bosphorus 4KDefaultOptimized Power Mode50100150200250SE +/- 1.12, N = 5SE +/- 1.56, N = 9218.76199.551. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 13 - Input: Bosphorus 4KDefaultOptimized Power Mode0.20180.40360.60540.80721.0090.8200.897
MinAvgMaxDefault80034593919Optimized Power Mode80034963921OpenBenchmarking.orgMegahertz, More Is BetterSVT-AV1 1.8CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault188.4266.9496.5Optimized Power Mode94.9222.4429.6OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.8CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 3363 Joules per run.2. Optimized Power Mode: Approximate power consumption of 2891 Joules per run.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 13 - Input: Bosphorus 1080pDefaultOptimized Power Mode140280420560700SE +/- 4.69, N = 11SE +/- 3.16, N = 10628.76574.601. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 13 - Input: Bosphorus 1080pDefaultOptimized Power Mode0.6391.2781.9172.5563.1952.642.84
MinAvgMaxDefault80032053925Optimized Power Mode80032103922OpenBenchmarking.orgMegahertz, More Is BetterSVT-AV1 1.8CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault193.8238.2383.1Optimized Power Mode177.1202.4313.6OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.8CPU Power Consumption Monitor1002003004005001. Default: Approximate power consumption of 1602 Joules per run.2. Optimized Power Mode: Approximate power consumption of 1416 Joules per run.

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 InstallationDefaultOptimized Power Mode20406080100SE +/- 0.03, N = 3SE +/- 0.13, N = 381.6989.37
MinAvgMaxDefault80029943905Optimized Power Mode80029753906OpenBenchmarking.orgMegahertz, More Is BetterOpenRadioss 2023.09.15CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault187640756Optimized Power Mode93580640OpenBenchmarking.orgWatts, Fewer Is BetterOpenRadioss 2023.09.15CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 58006 Joules per run.2. Optimized Power Mode: Approximate power consumption of 56865 Joules per run.

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 FastDefaultOptimized Power Mode20406080100SE +/- 0.23, N = 6SE +/- 0.70, N = 677.5370.871. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Ultra FastDefaultOptimized Power Mode0.05240.10480.15720.20960.2620.2130.233
MinAvgMaxDefault80034253909Optimized Power Mode80036113914OpenBenchmarking.orgMegahertz, More Is BetterKvazaar 2.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault195.7364.2476.1Optimized Power Mode187.6303.7417.2OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor1202403604806001. Default: Approximate power consumption of 4431 Joules per run.2. Optimized Power Mode: Approximate power consumption of 3897 Joules per run.

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 FastDefaultOptimized Power Mode4080120160200SE +/- 0.74, N = 8SE +/- 1.07, N = 11160.63147.08
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Very FastDefaultOptimized Power Mode0.12040.24080.36120.48160.6020.5260.535
MinAvgMaxDefault80032443911Optimized Power Mode80035073923OpenBenchmarking.orgMegahertz, More Is Betteruvg266 0.4.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault149.6305.7483.7Optimized Power Mode183.2274.9402.4OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor1202403604806001. Default: Approximate power consumption of 2483 Joules per run.2. Optimized Power Mode: Approximate power consumption of 2274 Joules per run.

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 1.9Video Input: Bosphorus 4K - Video Preset: FastDefaultOptimized Power Mode246810SE +/- 0.067, N = 3SE +/- 0.066, N = 36.9216.3401. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterVVenC 1.9Video Input: Bosphorus 4K - Video Preset: FastDefaultOptimized Power Mode0.00470.00940.01410.01880.02350.0170.021
MinAvgMaxDefault80038533916Optimized Power Mode80038733908OpenBenchmarking.orgMegahertz, More Is BetterVVenC 1.9CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207.3418.4548.0Optimized Power Mode182.5307.5432.6OpenBenchmarking.orgWatts, Fewer Is BetterVVenC 1.9CPU Power Consumption Monitor1402804205607001. Default: Approximate power consumption of 40025 Joules per run.2. Optimized Power Mode: Approximate power consumption of 32085 Joules per run.

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 FastDefaultOptimized Power Mode4080120160200SE +/- 1.12, N = 9SE +/- 1.12, N = 15166.47153.17
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Ultra FastDefaultOptimized Power Mode0.13250.2650.39750.530.66250.5660.589
MinAvgMaxDefault80034323920Optimized Power Mode80034123925OpenBenchmarking.orgMegahertz, More Is Betteruvg266 0.4.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault195.5294.4446.2Optimized Power Mode187.2260.1393.2OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor1202403604806001. Default: Approximate power consumption of 2322 Joules per run.2. Optimized Power Mode: Approximate power consumption of 2116 Joules per run.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Super FastDefaultOptimized Power Mode306090120150SE +/- 1.01, N = 15SE +/- 1.05, N = 15158.23145.75
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Super FastDefaultOptimized Power Mode0.12310.24620.36930.49240.61550.5260.547
MinAvgMaxDefault80033433917Optimized Power Mode80033043923OpenBenchmarking.orgMegahertz, More Is Betteruvg266 0.4.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault195.4300.9474.5Optimized Power Mode94.6266.6405.6OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor1202403604806001. Default: Approximate power consumption of 2387 Joules per run.2. Optimized Power Mode: Approximate power consumption of 2221 Joules per run.

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 - Mesh TimeDefaultOptimized Power Mode71421283530.0427.711. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -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: 19, Long Mode - Decompression SpeedDefaultOptimized Power Mode30060090012001500SE +/- 16.60, N = 3SE +/- 2.17, N = 31183.81094.21. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s Per Watt, More Is BetterZstd Compression 1.5.4Compression Level: 19, Long Mode - Decompression SpeedDefaultOptimized Power Mode1.30192.60383.90575.20766.50954.1215.786
MinAvgMaxDefault80038363907Optimized Power Mode80038883906OpenBenchmarking.orgMegahertz, More Is BetterZstd Compression 1.5.4CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault199.4287.2314.9Optimized Power Mode182.1189.1202.2OpenBenchmarking.orgWatts, Fewer Is BetterZstd Compression 1.5.4CPU Power Consumption Monitor801602403204001. Default: Approximate power consumption of 23075 Joules per run.2. Optimized Power Mode: Approximate power consumption of 16515 Joules per run.

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 1080pDefaultOptimized Power Mode120240360480600SE +/- 4.34, N = 9SE +/- 3.72, N = 8542.17502.051. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-VP9 0.3Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080pDefaultOptimized Power Mode0.51461.02921.54382.05842.5731.9992.287
MinAvgMaxDefault80035753909Optimized Power Mode80036143917OpenBenchmarking.orgMegahertz, More Is BetterSVT-VP9 0.3CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault199.9271.3493.1Optimized Power Mode178.7219.5408.7OpenBenchmarking.orgWatts, Fewer Is BetterSVT-VP9 0.3CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 2080 Joules per run.2. Optimized Power Mode: Approximate power consumption of 1783 Joules per run.

Timed GCC Compilation

This test times how long it takes to build the GNU Compiler Collection (GCC) open-source compiler. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed GCC Compilation 13.2Time To CompileDefaultOptimized Power Mode160320480640800SE +/- 2.37, N = 3SE +/- 4.73, N = 3706.26759.22
OpenBenchmarking.orgMegahertz, More Is BetterTimed GCC Compilation 13.2CPU Peak Freq (Highest CPU Core Frequency) MonitorDefaultOptimized Power Mode7001400210028003500Min: 800 / Avg: 3844.27 / Max: 3932Min: 800 / Avg: 3848.62 / Max: 3914
OpenBenchmarking.orgWatts, Fewer Is BetterTimed GCC Compilation 13.2CPU Power Consumption MonitorDefaultOptimized Power Mode130260390520650Min: 115.2 / Avg: 332.29 / Max: 762.28Min: 90.55 / Avg: 227.95 / Max: 760.591. Default: Approximate power consumption of 236811 Joules per run.2. Optimized Power Mode: Approximate power consumption of 174682 Joules per run.

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: 1000 - Mode: Read Write - Average LatencyDefaultOptimized Power Mode48121620SE +/- 0.02, N = 3SE +/- 0.06, N = 315.6316.751. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
MinAvgMaxDefault80029813910Optimized Power Mode80029963906OpenBenchmarking.orgMegahertz, More Is BetterPostgreSQL 16CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.2446.2510.9Optimized Power Mode144.0378.6455.4OpenBenchmarking.orgWatts, Fewer Is BetterPostgreSQL 16CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 63206 Joules per run.2. Optimized Power Mode: Approximate power consumption of 53759 Joules per run.

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 1000 - Mode: Read WriteDefaultOptimized Power Mode14K28K42K56K70KSE +/- 70.87, N = 3SE +/- 214.64, N = 363962596951. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -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 1080pDefaultOptimized Power Mode100200300400500SE +/- 1.93, N = 8SE +/- 2.86, N = 8463.27433.531. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-VP9 0.3Tuning: Visual Quality Optimized - Input: Bosphorus 1080pDefaultOptimized Power Mode0.43790.87581.31371.75162.18951.6661.946
MinAvgMaxDefault80037003921Optimized Power Mode80035343907OpenBenchmarking.orgMegahertz, More Is BetterSVT-VP9 0.3CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault199.6278.1500.7Optimized Power Mode179.1222.8433.3OpenBenchmarking.orgWatts, Fewer Is BetterSVT-VP9 0.3CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 2260 Joules per run.2. Optimized Power Mode: Approximate power consumption of 1894 Joules per run.

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 FastDefaultOptimized Power Mode60120180240300SE +/- 2.07, N = 11SE +/- 1.93, N = 10282.05263.961. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Ultra FastDefaultOptimized Power Mode0.25020.50040.75061.00081.2511.0511.112
MinAvgMaxDefault80035233912Optimized Power Mode80035053908OpenBenchmarking.orgMegahertz, More Is BetterKvazaar 2.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault195.4268.4416.5Optimized Power Mode146.9237.4350.4OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor1102203304405501. Default: Approximate power consumption of 1708 Joules per run.2. Optimized Power Mode: Approximate power consumption of 1543 Joules per run.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Very FastDefaultOptimized Power Mode1530456075SE +/- 0.26, N = 5SE +/- 0.44, N = 569.3664.971. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Very FastDefaultOptimized Power Mode0.03940.07880.11820.15760.1970.1640.175
MinAvgMaxDefault80038383907Optimized Power Mode80037503911OpenBenchmarking.orgMegahertz, More Is BetterKvazaar 2.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault195.9424.1563.7Optimized Power Mode185.0372.2523.5OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor1402804205607001. Default: Approximate power consumption of 5599 Joules per run.2. Optimized Power Mode: Approximate power consumption of 5136 Joules per run.

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 4 - Input: Bosphorus 1080pDefaultOptimized Power Mode510152025SE +/- 0.16, N = 5SE +/- 0.11, N = 520.9519.671. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 4 - Input: Bosphorus 1080pDefaultOptimized Power Mode0.01940.03880.05820.07760.0970.0680.086
MinAvgMaxDefault80035513907Optimized Power Mode80036153905OpenBenchmarking.orgMegahertz, More Is BetterSVT-AV1 1.8CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault188.7309.8398.3Optimized Power Mode179.0229.9287.8OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.8CPU Power Consumption Monitor1102203304405501. Default: Approximate power consumption of 4337 Joules per run.2. Optimized Power Mode: Approximate power consumption of 3494 Joules per run.

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: 1000DefaultOptimized Power Mode20K40K60K80K100K80826.7176315.651. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenBenchmarking.orgRequests Per Second Per Watt, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 1000DefaultOptimized Power Mode60120180240300208.91262.58
MinAvgMaxDefault80029533906Optimized Power Mode80028633903OpenBenchmarking.orgMegahertz, More Is BetterApache HTTP Server 2.4.56CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault195.5386.9451.6Optimized Power Mode99.1290.6333.6OpenBenchmarking.orgWatts, Fewer Is BetterApache HTTP Server 2.4.56CPU Power Consumption Monitor120240360480600

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 MakefilesDefaultOptimized Power Mode4080120160200SE +/- 0.30, N = 3SE +/- 0.58, N = 3177.43187.84
MinAvgMaxDefault80037013918Optimized Power Mode80037093912OpenBenchmarking.orgMegahertz, More Is BetterTimed LLVM Compilation 16.0CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault200469762Optimized Power Mode94367758OpenBenchmarking.orgWatts, Fewer Is BetterTimed LLVM Compilation 16.0CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 91316 Joules per run.2. Optimized Power Mode: Approximate power consumption of 75506 Joules per run.

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: TPC-H ParquetDefaultOptimized Power Mode306090120150SE +/- 0.31, N = 3SE +/- 0.38, N = 3148.55157.101. (CXX) g++ options: -O3 -rdynamic -lssl -lcrypto -ldl
MinAvgMaxDefault80038773921Optimized Power Mode80038813918OpenBenchmarking.orgMegahertz, More Is BetterDuckDB 0.9.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault137.7288.3394.2Optimized Power Mode90.3192.7281.6OpenBenchmarking.orgWatts, Fewer Is BetterDuckDB 0.9.1CPU Power Consumption Monitor1102203304405501. Default: Approximate power consumption of 59783 Joules per run.2. Optimized Power Mode: Approximate power consumption of 41243 Joules per run.

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 FastDefaultOptimized Power Mode1632486480SE +/- 0.25, N = 5SE +/- 0.12, N = 571.0567.211. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Super FastDefaultOptimized Power Mode0.04390.08780.13170.17560.21950.1780.195
MinAvgMaxDefault80037003908Optimized Power Mode80036033912OpenBenchmarking.orgMegahertz, More Is BetterKvazaar 2.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207.0398.5519.8Optimized Power Mode186.6344.9484.9OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 5181 Joules per run.2. Optimized Power Mode: Approximate power consumption of 4622 Joules per run.

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: Compression RatingDefaultOptimized Power Mode150K300K450K600K750KSE +/- 869.44, N = 3SE +/- 1440.87, N = 36895166544591. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

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: PlatformDefaultOptimized Power Mode1224364860SE +/- 0.11, N = 3SE +/- 0.14, N = 353.9851.311. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFPS Per Watt, More Is BetterFFmpeg 6.1Encoder: libx265 - Scenario: PlatformDefaultOptimized Power Mode0.05330.10660.15990.21320.26650.1690.237
MinAvgMaxDefault80038683913Optimized Power Mode80038833909OpenBenchmarking.orgMegahertz, More Is BetterFFmpeg 6.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault103.6319.0374.1Optimized Power Mode100.3216.9272.9OpenBenchmarking.orgWatts, Fewer Is BetterFFmpeg 6.1CPU Power Consumption Monitor1002003004005001. Default: Approximate power consumption of 52107 Joules per run.2. Optimized Power Mode: Approximate power consumption of 37236 Joules per run.

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: 1:100DefaultOptimized Power Mode700K1400K2100K2800K3500KSE +/- 25091.90, N = 3SE +/- 42725.42, N = 33441439.713274425.321. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
OpenBenchmarking.orgOps/sec Per Watt, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:100DefaultOptimized Power Mode150030004500600075006692.766801.69
MinAvgMaxDefault80028913910Optimized Power Mode80029083905OpenBenchmarking.orgMegahertz, More Is BetterMemcached 1.6.19CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.7514.2607.5Optimized Power Mode183.2481.4564.7OpenBenchmarking.orgWatts, Fewer Is BetterMemcached 1.6.19CPU Power Consumption Monitor1603204806408001. Default: Approximate power consumption of 35994 Joules per run.2. Optimized Power Mode: Approximate power consumption of 33859 Joules per run.

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 FastDefaultOptimized Power Mode60120180240300SE +/- 1.09, N = 10SE +/- 1.69, N = 10263.88251.331. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Super FastDefaultOptimized Power Mode0.22280.44560.66840.89121.1140.940.99
MinAvgMaxDefault80034063908Optimized Power Mode80032633907OpenBenchmarking.orgMegahertz, More Is BetterKvazaar 2.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault195.0280.7447.3Optimized Power Mode184.6253.9385.9OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor1202403604806001. Default: Approximate power consumption of 1825 Joules per run.2. Optimized Power Mode: Approximate power consumption of 1701 Joules per run.

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: 128 - Buffer Length: 256 - Filter Length: 57DefaultOptimized Power Mode900M1800M2700M3600M4500MSE +/- 11478143.48, N = 3SE +/- 34188302.09, N = 3432986666741250000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 128 - Buffer Length: 256 - Filter Length: 57DefaultOptimized Power Mode1.5M3M4.5M6M7.5M6804882.127072857.52
MinAvgMaxDefault80029563904Optimized Power Mode80028633903OpenBenchmarking.orgMegahertz, More Is BetterLiquid-DSP 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault201636742Optimized Power Mode183583669OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 22058 Joules per run.2. Optimized Power Mode: Approximate power consumption of 20218 Joules per run.

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: Very FastDefaultOptimized Power Mode60120180240300SE +/- 1.16, N = 10SE +/- 1.04, N = 10268.49256.191. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Very FastDefaultOptimized Power Mode0.21560.43120.64680.86241.0780.9250.958
MinAvgMaxDefault80035303907Optimized Power Mode80033103909OpenBenchmarking.orgMegahertz, More Is BetterKvazaar 2.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault195.3290.3469.1Optimized Power Mode188.2267.5439.5OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor1202403604806001. Default: Approximate power consumption of 1887 Joules per run.2. Optimized Power Mode: Approximate power consumption of 1765 Joules per run.

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: 500DefaultOptimized Power Mode500K1000K1500K2000K2500KSE +/- 7546.74, N = 3SE +/- 19286.75, N = 152503177.672389636.671. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3
OpenBenchmarking.orgRequests Per Second Per Watt, More Is BetterRedis 7.0.4Test: SET - Parallel Connections: 500DefaultOptimized Power Mode140028004200560070006160.266737.20
MinAvgMaxDefault80033493905Optimized Power Mode80032053905OpenBenchmarking.orgMegahertz, More Is BetterRedis 7.0.4CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.4406.3549.1Optimized Power Mode93.8354.7522.3OpenBenchmarking.orgWatts, Fewer Is BetterRedis 7.0.4CPU Power Consumption Monitor1402804205607001. Default: Approximate power consumption of 12597 Joules per run.2. Optimized Power Mode: Approximate power consumption of 11161 Joules per run.

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode2K4K6K8K10KSE +/- 7.42, N = 3SE +/- 12.41, N = 311244.7910738.14

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode1.33622.67244.00865.34486.681SE +/- 0.0033, N = 3SE +/- 0.0065, N = 35.67325.9385
MinAvgMaxDefault80028513905Optimized Power Mode80028713907OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207614755Optimized Power Mode178555660OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 26822 Joules per run.2. Optimized Power Mode: Approximate power consumption of 24410 Joules per run.

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 4.0Blend File: Classroom - Compute: CPU-OnlyDefaultOptimized Power Mode816243240SE +/- 0.33, N = 3SE +/- 0.07, N = 334.9433.40
MinAvgMaxDefault80027283905Optimized Power Mode80026963906OpenBenchmarking.orgMegahertz, More Is BetterBlender 4.0CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206621735Optimized Power Mode181618758OpenBenchmarking.orgWatts, Fewer Is BetterBlender 4.0CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 24651 Joules per run.2. Optimized Power Mode: Approximate power consumption of 23896 Joules per run.

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: Video On DemandDefaultOptimized Power Mode1224364860SE +/- 0.04, N = 3SE +/- 0.10, N = 353.6351.451. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFPS Per Watt, More Is BetterFFmpeg 6.1Encoder: libx265 - Scenario: Video On DemandDefaultOptimized Power Mode0.05310.10620.15930.21240.26550.1690.236
MinAvgMaxDefault80038883914Optimized Power Mode80038953912OpenBenchmarking.orgMegahertz, More Is BetterFFmpeg 6.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault162.7318.2364.5Optimized Power Mode109.2218.0280.6OpenBenchmarking.orgWatts, Fewer Is BetterFFmpeg 6.1CPU Power Consumption Monitor1002003004005001. Default: Approximate power consumption of 52403 Joules per run.2. Optimized Power Mode: Approximate power consumption of 37354 Joules per run.

Intel Open Image Denoise

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.1Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-OnlyDefaultOptimized Power Mode1.0622.1243.1864.2485.31SE +/- 0.01, N = 6SE +/- 0.01, N = 64.534.72
OpenBenchmarking.orgImages / Sec Per Watt, More Is BetterIntel Open Image Denoise 2.1Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-OnlyDefaultOptimized Power Mode0.00230.00460.00690.00920.01150.010.01
MinAvgMaxDefault80031493901Optimized Power Mode80029613904OpenBenchmarking.orgMegahertz, More Is BetterIntel Open Image Denoise 2.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault200473676Optimized Power Mode183467708OpenBenchmarking.orgWatts, Fewer Is BetterIntel Open Image Denoise 2.1CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 5519 Joules per run.2. Optimized Power Mode: Approximate power consumption of 5368 Joules per run.

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: UploadDefaultOptimized Power Mode612182430SE +/- 0.06, N = 3SE +/- 0.06, N = 327.0026.021. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFPS Per Watt, More Is BetterFFmpeg 6.1Encoder: libx265 - Scenario: UploadDefaultOptimized Power Mode0.02810.05620.08430.11240.14050.0900.125
MinAvgMaxDefault80038763907Optimized Power Mode80038763907OpenBenchmarking.orgMegahertz, More Is BetterFFmpeg 6.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault138.2299.2415.2Optimized Power Mode89.7207.8317.2OpenBenchmarking.orgWatts, Fewer Is BetterFFmpeg 6.1CPU Power Consumption Monitor1102203304405501. Default: Approximate power consumption of 50564 Joules per run.2. Optimized Power Mode: Approximate power consumption of 35814 Joules per run.

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: 128 - Buffer Length: 256 - Filter Length: 32DefaultOptimized Power Mode800M1600M2400M3200M4000MSE +/- 1386041.53, N = 3SE +/- 3883440.63, N = 3361186666737452333331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 128 - Buffer Length: 256 - Filter Length: 32DefaultOptimized Power Mode1.5M3M4.5M6M7.5M6080886.757101447.97
MinAvgMaxDefault80028973906Optimized Power Mode80029313903OpenBenchmarking.orgMegahertz, More Is BetterLiquid-DSP 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207594674Optimized Power Mode181527585OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 20393 Joules per run.2. Optimized Power Mode: Approximate power consumption of 18107 Joules per run.

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.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUDefaultOptimized Power Mode30K60K90K120K150KSE +/- 775.18, N = 3SE +/- 1361.46, N = 3121577.55117369.551. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

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: 32 - Buffer Length: 256 - Filter Length: 512DefaultOptimized Power Mode110M220M330M440M550MSE +/- 5123258.57, N = 6SE +/- 3588947.54, N = 35082450004914266671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 32 - Buffer Length: 256 - Filter Length: 512DefaultOptimized Power Mode300K600K900K1200K1500K1124172.641384860.49
MinAvgMaxDefault80037743908Optimized Power Mode80038413911OpenBenchmarking.orgMegahertz, More Is BetterLiquid-DSP 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.4452.1497.1Optimized Power Mode183.1354.9394.7OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 15522 Joules per run.2. Optimized Power Mode: Approximate power consumption of 12302 Joules per run.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 256 - Buffer Length: 256 - Filter Length: 57DefaultOptimized Power Mode1200M2400M3600M4800M6000MSE +/- 11767044.38, N = 3SE +/- 19835069.95, N = 3583550000056566000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 256 - Buffer Length: 256 - Filter Length: 57DefaultOptimized Power Mode2M4M6M8M10M9120590.649209885.66
MinAvgMaxDefault80025893900Optimized Power Mode80029313904OpenBenchmarking.orgMegahertz, More Is BetterLiquid-DSP 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault196640754Optimized Power Mode187614691OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 21967 Joules per run.2. Optimized Power Mode: Approximate power consumption of 21087 Joules per run.

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 6.1Build: defconfigDefaultOptimized Power Mode612182430SE +/- 0.21, N = 8SE +/- 0.18, N = 1123.7624.48
MinAvgMaxDefault80034543911Optimized Power Mode80034913908OpenBenchmarking.orgMegahertz, More Is BetterTimed Linux Kernel Compilation 6.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault104388760Optimized Power Mode90320761OpenBenchmarking.orgWatts, Fewer Is BetterTimed Linux Kernel Compilation 6.1CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 12304 Joules per run.2. Optimized Power Mode: Approximate power consumption of 10383 Joules per run.

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: GET - Parallel Connections: 500DefaultOptimized Power Mode700K1400K2100K2800K3500KSE +/- 482.98, N = 3SE +/- 36270.68, N = 33240001.583152359.751. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3
OpenBenchmarking.orgRequests Per Second Per Watt, More Is BetterRedis 7.0.4Test: GET - Parallel Connections: 500DefaultOptimized Power Mode2K4K6K8K10K9025.1710408.05
MinAvgMaxDefault80032493907Optimized Power Mode80033323903OpenBenchmarking.orgMegahertz, More Is BetterRedis 7.0.4CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.6359.0511.5Optimized Power Mode184.3302.9456.0OpenBenchmarking.orgWatts, Fewer Is BetterRedis 7.0.4CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 9573 Joules per run.2. Optimized Power Mode: Approximate power consumption of 8178 Joules per run.

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: 128 - Buffer Length: 256 - Filter Length: 512DefaultOptimized Power Mode300M600M900M1200M1500MSE +/- 4762352.36, N = 3SE +/- 13505101.92, N = 3147390000015134666671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 128 - Buffer Length: 256 - Filter Length: 512DefaultOptimized Power Mode500K1000K1500K2000K2500K2318859.552561825.81
MinAvgMaxDefault80029053905Optimized Power Mode80029143906OpenBenchmarking.orgMegahertz, More Is BetterLiquid-DSP 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207636728Optimized Power Mode182591679OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 21823 Joules per run.2. Optimized Power Mode: Approximate power consumption of 20677 Joules per run.

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 2023.2.devModel: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUDefaultOptimized Power Mode0.09230.18460.27690.36920.4615SE +/- 0.00, N = 3SE +/- 0.01, N = 30.400.41MIN: 0.18 / MAX: 14.47MIN: 0.19 / MAX: 15.841. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie
MinAvgMaxDefault80031683907Optimized Power Mode80032583905OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2023.2.devCPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.6468.0594.9Optimized Power Mode178.9411.8522.8OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.2.devCPU Power Consumption Monitor1603204806408001. Default: Approximate power consumption of 31045 Joules per run.2. Optimized Power Mode: Approximate power consumption of 27319 Joules per run.

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 BeamDefaultOptimized Power Mode20406080100SE +/- 0.18, N = 3SE +/- 0.13, N = 384.7882.73
MinAvgMaxDefault80029453911Optimized Power Mode80029303906OpenBenchmarking.orgMegahertz, More Is BetterOpenRadioss 2023.09.15CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault104651755Optimized Power Mode173641735OpenBenchmarking.orgWatts, Fewer Is BetterOpenRadioss 2023.09.15CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 59209 Joules per run.2. Optimized Power Mode: Approximate power consumption of 57702 Joules per run.

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: NinjaDefaultOptimized Power Mode20406080100SE +/- 0.29, N = 3SE +/- 0.20, N = 396.5298.91
MinAvgMaxDefault249133243908Optimized Power Mode80033183908OpenBenchmarking.orgMegahertz, More Is BetterTimed LLVM Compilation 16.0CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault103521759Optimized Power Mode89460758OpenBenchmarking.orgWatts, Fewer Is BetterTimed LLVM Compilation 16.0CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 57459 Joules per run.2. Optimized Power Mode: Approximate power consumption of 52591 Joules per run.

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: MediumDefaultOptimized Power Mode20406080100SE +/- 0.08, N = 6SE +/- 0.25, N = 688.4686.78
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: MediumDefaultOptimized Power Mode0.05780.11560.17340.23120.2890.2360.257
MinAvgMaxDefault80036493910Optimized Power Mode80036543908OpenBenchmarking.orgMegahertz, More Is Betteruvg266 0.4.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault199.5374.9505.3Optimized Power Mode188.2337.9473.2OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 4186 Joules per run.2. Optimized Power Mode: Approximate power consumption of 3829 Joules per run.

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: 256 - Buffer Length: 256 - Filter Length: 32DefaultOptimized Power Mode1300M2600M3900M5200M6500MSE +/- 7150058.27, N = 3SE +/- 61606333.10, N = 6618560000063014833331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 256 - Buffer Length: 256 - Filter Length: 32DefaultOptimized Power Mode2M4M6M8M10M9739696.8610558400.84
MinAvgMaxDefault80028243908Optimized Power Mode80029443905OpenBenchmarking.orgMegahertz, More Is BetterLiquid-DSP 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault139635755Optimized Power Mode183597744OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 22017 Joules per run.2. Optimized Power Mode: Approximate power consumption of 20391 Joules per run.

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: Li2_STO_aeDefaultOptimized Power Mode20406080100SE +/- 0.40, N = 3SE +/- 1.02, N = 594.3892.671. (CXX) g++ options: -fopenmp -foffload=disable -finline-limit=1000 -fstrict-aliasing -funroll-all-loops -ffast-math -march=native -O3 -lm -ldl
MinAvgMaxDefault80027603909Optimized Power Mode80027393907OpenBenchmarking.orgMegahertz, More Is BetterQMCPACK 3.17.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault193667756Optimized Power Mode175663758OpenBenchmarking.orgWatts, Fewer Is BetterQMCPACK 3.17.1CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 68485 Joules per run.2. Optimized Power Mode: Approximate power consumption of 66661 Joules per run.

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: SlowDefaultOptimized Power Mode612182430SE +/- 0.08, N = 3SE +/- 0.08, N = 327.2026.71
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: SlowDefaultOptimized Power Mode0.01150.0230.03450.0460.05750.0480.051
MinAvgMaxDefault80037283908Optimized Power Mode80037753908OpenBenchmarking.orgMegahertz, More Is Betteruvg266 0.4.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207567719Optimized Power Mode184520693OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 15112 Joules per run.2. Optimized Power Mode: Approximate power consumption of 14226 Joules per run.

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 4.0Blend File: BMW27 - Compute: CPU-OnlyDefaultOptimized Power Mode3691215SE +/- 0.03, N = 4SE +/- 0.04, N = 412.8912.67
MinAvgMaxDefault80028183905Optimized Power Mode80029573909OpenBenchmarking.orgMegahertz, More Is BetterBlender 4.0CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207550738Optimized Power Mode181539759OpenBenchmarking.orgWatts, Fewer Is BetterBlender 4.0CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 9768 Joules per run.2. Optimized Power Mode: Approximate power consumption of 9565 Joules per run.

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: SlowDefaultOptimized Power Mode20406080100SE +/- 0.10, N = 6SE +/- 0.19, N = 680.3078.93
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: SlowDefaultOptimized Power Mode0.05130.10260.15390.20520.25650.2080.228
MinAvgMaxDefault80033793909Optimized Power Mode80035683909OpenBenchmarking.orgMegahertz, More Is Betteruvg266 0.4.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault203.8386.0506.0Optimized Power Mode188.3346.5464.1OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 4568 Joules per run.2. Optimized Power Mode: Approximate power consumption of 4158 Joules per run.

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: KawPow - Hash Count: 1MDefaultOptimized Power Mode15K30K45K60K75KSE +/- 82.14, N = 4SE +/- 943.50, N = 370383.269196.71. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
OpenBenchmarking.orgH/s Per Watt, More Is BetterXmrig 6.21Variant: KawPow - Hash Count: 1MDefaultOptimized Power Mode306090120150125.63128.19
MinAvgMaxDefault80027523908Optimized Power Mode80026733906OpenBenchmarking.orgMegahertz, More Is BetterXmrig 6.21CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault200560759Optimized Power Mode185540759OpenBenchmarking.orgWatts, Fewer Is BetterXmrig 6.21CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 11625 Joules per run.2. Optimized Power Mode: Approximate power consumption of 11696 Joules per run.

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: SlowDefaultOptimized Power Mode306090120150SE +/- 0.38, N = 8SE +/- 0.35, N = 8127.25125.181. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: SlowDefaultOptimized Power Mode0.09810.19620.29430.39240.49050.3600.436
MinAvgMaxDefault80035023906Optimized Power Mode80033833905OpenBenchmarking.orgMegahertz, More Is BetterKvazaar 2.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault205.7353.2497.3Optimized Power Mode142.7287.0393.6OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 3179 Joules per run.2. Optimized Power Mode: Approximate power consumption of 2619 Joules per run.

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 TestDefaultOptimized Power Mode612182430SE +/- 0.14, N = 3SE +/- 0.12, N = 325.7425.37
MinAvgMaxDefault80031803905Optimized Power Mode80032403906OpenBenchmarking.orgMegahertz, More Is BetterOpenRadioss 2023.09.15CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault200518755Optimized Power Mode156477732OpenBenchmarking.orgWatts, Fewer Is BetterOpenRadioss 2023.09.15CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 20380 Joules per run.2. Optimized Power Mode: Approximate power consumption of 18923 Joules per run.

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.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Weld Porosity Detection FP16-INT8 - Device: CPUDefaultOptimized Power Mode11K22K33K44K55KSE +/- 490.25, N = 3SE +/- 229.19, N = 348795.9749494.061. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode4080120160200SE +/- 1.36, N = 3SE +/- 0.33, N = 3160.41162.59

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode90180270360450SE +/- 3.53, N = 3SE +/- 0.71, N = 3397.77392.45
MinAvgMaxDefault80031503907Optimized Power Mode80032443908OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207470759Optimized Power Mode93400759OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 38392 Joules per run.2. Optimized Power Mode: Approximate power consumption of 34803 Joules per run.

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: 64 - Buffer Length: 256 - Filter Length: 512DefaultOptimized Power Mode200M400M600M800M1000MSE +/- 10474795.68, N = 3SE +/- 5205466.14, N = 310123366679996633331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 64 - Buffer Length: 256 - Filter Length: 512DefaultOptimized Power Mode400K800K1200K1600K2000K1627411.291823904.25
MinAvgMaxDefault80037323907Optimized Power Mode80037973906OpenBenchmarking.orgMegahertz, More Is BetterLiquid-DSP 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault200622737Optimized Power Mode181548629OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 21565 Joules per run.2. Optimized Power Mode: Approximate power consumption of 19183 Joules per run.

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 2023.2.devModel: Weld Porosity Detection FP16-INT8 - Device: CPUDefaultOptimized Power Mode0.55131.10261.65392.20522.7565SE +/- 0.01, N = 3SE +/- 0.00, N = 32.452.42MIN: 2.03 / MAX: 23.83MIN: 1.96 / MAX: 28.31. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie
MinAvgMaxDefault80023403905Optimized Power Mode80024073907OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2023.2.devCPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206658766Optimized Power Mode179650730OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.2.devCPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 43207 Joules per run.2. Optimized Power Mode: Approximate power consumption of 42927 Joules per run.

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode80160240320400SE +/- 0.69, N = 3SE +/- 0.05, N = 3350.90346.74
MinAvgMaxDefault80028753907Optimized Power Mode80028443907OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207604750Optimized Power Mode176584755OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 27991 Joules per run.2. Optimized Power Mode: Approximate power consumption of 27250 Joules per run.

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 6.1Build: allmodconfigDefaultOptimized Power Mode306090120150SE +/- 0.51, N = 3SE +/- 0.48, N = 3149.79151.28
MinAvgMaxDefault80030663911Optimized Power Mode80030663911OpenBenchmarking.orgMegahertz, More Is BetterTimed Linux Kernel Compilation 6.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault186584759Optimized Power Mode174553761OpenBenchmarking.orgWatts, Fewer Is BetterTimed Linux Kernel Compilation 6.1CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 94487 Joules per run.2. Optimized Power Mode: Approximate power consumption of 90265 Joules per run.

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 2023.2.devModel: Handwritten English Recognition FP16-INT8 - Device: CPUDefaultOptimized Power Mode918273645SE +/- 0.08, N = 3SE +/- 0.08, N = 338.4138.79MIN: 35.95 / MAX: 58.71MIN: 36.26 / MAX: 60.121. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie
MinAvgMaxDefault80029423912Optimized Power Mode80029573906OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2023.2.devCPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206657735Optimized Power Mode179569617OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.2.devCPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 43114 Joules per run.2. Optimized Power Mode: Approximate power consumption of 37366 Joules per run.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Handwritten English Recognition FP16-INT8 - Device: CPUDefaultOptimized Power Mode7001400210028003500SE +/- 7.06, N = 3SE +/- 6.73, N = 33330.193297.811. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

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: Wownero - Hash Count: 1MDefaultOptimized Power Mode16K32K48K64K80KSE +/- 677.05, N = 4SE +/- 24.14, N = 475002.375701.71. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
OpenBenchmarking.orgH/s Per Watt, More Is BetterXmrig 6.21Variant: Wownero - Hash Count: 1MDefaultOptimized Power Mode306090120150134.73141.96
MinAvgMaxDefault80028903908Optimized Power Mode80029003904OpenBenchmarking.orgMegahertz, More Is BetterXmrig 6.21CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207557766Optimized Power Mode185533759OpenBenchmarking.orgWatts, Fewer Is BetterXmrig 6.21CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 10995 Joules per run.2. Optimized Power Mode: Approximate power consumption of 10799 Joules per run.

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode4080120160200SE +/- 0.43, N = 3SE +/- 0.02, N = 3182.16183.71

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode400800120016002000SE +/- 7.80, N = 3SE +/- 3.31, N = 31791.271776.53

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode400800120016002000SE +/- 1.67, N = 3SE +/- 0.88, N = 31899.221883.66

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode816243240SE +/- 0.16, N = 3SE +/- 0.07, N = 335.6735.96

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode816243240SE +/- 0.03, N = 3SE +/- 0.01, N = 333.6433.90
MinAvgMaxDefault80029493905Optimized Power Mode80029543926OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault145568758Optimized Power Mode174526714OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 29515 Joules per run.2. Optimized Power Mode: Approximate power consumption of 28045 Joules per run.

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: SlowDefaultOptimized Power Mode918273645SE +/- 0.06, N = 4SE +/- 0.06, N = 441.4541.131. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: SlowDefaultOptimized Power Mode0.01820.03640.05460.07280.0910.0770.081
MinAvgMaxDefault80036893920Optimized Power Mode80038163907OpenBenchmarking.orgMegahertz, More Is BetterKvazaar 2.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault200536673Optimized Power Mode188506639OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 10316 Joules per run.2. Optimized Power Mode: Approximate power consumption of 9605 Joules per run.

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-ThreadedDefaultOptimized Power Mode80K160K240K320K400KSE +/- 127.57, N = 3SE +/- 52.43, N = 3388447.2391468.41. (CXX) g++ options: -O3 -march=native -fPIE -pie
OpenBenchmarking.orgMFLOPS Per Watt, More Is BetterQuantLib 1.32Configuration: Multi-ThreadedDefaultOptimized Power Mode130260390520650596.03606.16
MinAvgMaxDefault80026843904Optimized Power Mode80027353906OpenBenchmarking.orgMegahertz, More Is BetterQuantLib 1.32CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault199652758Optimized Power Mode181646759OpenBenchmarking.orgWatts, Fewer Is BetterQuantLib 1.32CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 43014 Joules per run.2. Optimized Power Mode: Approximate power consumption of 42409 Joules per run.

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: MediumDefaultOptimized Power Mode1020304050SE +/- 0.06, N = 4SE +/- 0.05, N = 441.9941.691. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: MediumDefaultOptimized Power Mode0.01890.03780.05670.07560.09450.0780.084
MinAvgMaxDefault80036883907Optimized Power Mode80038143928OpenBenchmarking.orgMegahertz, More Is BetterKvazaar 2.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault203537704Optimized Power Mode188494633OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 10077 Joules per run.2. Optimized Power Mode: Approximate power consumption of 9386 Joules per run.

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: INIVOL and Fluid Structure Interaction Drop ContainerDefaultOptimized Power Mode20406080100SE +/- 0.07, N = 3SE +/- 0.08, N = 397.8798.49
MinAvgMaxDefault80029933905Optimized Power Mode80030243907OpenBenchmarking.orgMegahertz, More Is BetterOpenRadioss 2023.09.15CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault201639756Optimized Power Mode96623713OpenBenchmarking.orgWatts, Fewer Is BetterOpenRadioss 2023.09.15CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 70270 Joules per run.2. Optimized Power Mode: Approximate power consumption of 69121 Joules per run.

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: 64 - Buffer Length: 256 - Filter Length: 32DefaultOptimized Power Mode500M1000M1500M2000M2500MSE +/- 6835284.27, N = 3SE +/- 13974301.81, N = 3247603333324608666671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 64 - Buffer Length: 256 - Filter Length: 32DefaultOptimized Power Mode1.1M2.2M3.3M4.4M5.5M4283633.395140098.56
MinAvgMaxDefault80036863907Optimized Power Mode80037453905OpenBenchmarking.orgMegahertz, More Is BetterLiquid-DSP 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault194578650Optimized Power Mode184479543OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 19653 Joules per run.2. Optimized Power Mode: Approximate power consumption of 16597 Joules per run.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 32 - Buffer Length: 256 - Filter Length: 32DefaultOptimized Power Mode300M600M900M1200M1500MSE +/- 1273228.62, N = 3SE +/- 1193035.34, N = 3121176666712045000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 32 - Buffer Length: 256 - Filter Length: 32DefaultOptimized Power Mode800K1600K2400K3200K4000K2919266.293778050.23
MinAvgMaxDefault80037593905Optimized Power Mode80038603906OpenBenchmarking.orgMegahertz, More Is BetterLiquid-DSP 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.5415.1447.6Optimized Power Mode184.1318.8341.0OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor1202403604806001. Default: Approximate power consumption of 14390 Joules per run.2. Optimized Power Mode: Approximate power consumption of 10946 Joules per run.

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode306090120150SE +/- 0.09, N = 3SE +/- 0.30, N = 3137.35138.15

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: MediumDefaultOptimized Power Mode714212835SE +/- 0.08, N = 3SE +/- 0.19, N = 329.7929.62
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: MediumDefaultOptimized Power Mode0.01310.02620.03930.05240.06550.0540.058
MinAvgMaxDefault80035853904Optimized Power Mode80037523917OpenBenchmarking.orgMegahertz, More Is Betteruvg266 0.4.1CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault201556727Optimized Power Mode185513661OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 13893 Joules per run.2. Optimized Power Mode: Approximate power consumption of 12827 Joules per run.

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 1MDefaultOptimized Power Mode20406080100SE +/- 0.14, N = 3SE +/- 0.17, N = 385.8286.30
MinAvgMaxDefault80031383908Optimized Power Mode80031533907OpenBenchmarking.orgMegahertz, More Is BetterOpenRadioss 2023.09.15CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault144568757Optimized Power Mode89537741OpenBenchmarking.orgWatts, Fewer Is BetterOpenRadioss 2023.09.15CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 67742 Joules per run.2. Optimized Power Mode: Approximate power consumption of 65122 Joules per run.

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode816243240SE +/- 0.08, N = 3SE +/- 0.14, N = 335.9336.13
MinAvgMaxDefault80028353928Xeon Platinun 8592+80027613910Optimized Power Mode80028403914OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207616758Xeon Platinun 8592+207612758Optimized Power Mode179603760OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 26899 Joules per run.2. Optimized Power Mode: Approximate power consumption of 26321 Joules per run.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode400800120016002000SE +/- 3.42, N = 3SE +/- 6.86, N = 31778.591768.91

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 TimeDefaultOptimized Power Mode61218243023.4723.351. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm
MinAvgMaxDefault80030673910Optimized Power Mode280031213907OpenBenchmarking.orgMegahertz, More Is BetterOpenFOAM 10CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206590755Optimized Power Mode169556717OpenBenchmarking.orgWatts, Fewer Is BetterOpenFOAM 10CPU Power Consumption Monitor2004006008001000

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.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Face Detection Retail FP16-INT8 - Device: CPUDefaultOptimized Power Mode5K10K15K20K25KSE +/- 42.09, N = 3SE +/- 6.86, N = 324795.1724922.401. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Machine Translation EN To DE FP16 - Device: CPUDefaultOptimized Power Mode2004006008001000SE +/- 9.12, N = 3SE +/- 3.05, N = 31106.321100.761. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode100200300400500SE +/- 0.03, N = 3SE +/- 0.70, N = 3462.99460.69
MinAvgMaxDefault80025933909Optimized Power Mode80026523911OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206571760Optimized Power Mode175541760OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 29330 Joules per run.2. Optimized Power Mode: Approximate power consumption of 28131 Joules per run.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode306090120150SE +/- 0.21, N = 3SE +/- 0.12, N = 3137.34138.01

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 2023.2.devModel: Machine Translation EN To DE FP16 - Device: CPUDefaultOptimized Power Mode714212835SE +/- 0.23, N = 3SE +/- 0.08, N = 328.8528.99MIN: 21.11 / MAX: 222.12MIN: 22.06 / MAX: 222.831. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie
MinAvgMaxDefault80026963904Optimized Power Mode80029533906OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2023.2.devCPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207655751Optimized Power Mode179638710OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.2.devCPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 43422 Joules per run.2. Optimized Power Mode: Approximate power consumption of 42740 Joules per run.

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 WindshieldDefaultOptimized Power Mode306090120150SE +/- 0.65, N = 3SE +/- 0.28, N = 3111.23111.73
MinAvgMaxDefault80029243915Optimized Power Mode80029523907OpenBenchmarking.orgMegahertz, More Is BetterOpenRadioss 2023.09.15CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault101657757Optimized Power Mode178644730OpenBenchmarking.orgWatts, Fewer Is BetterOpenRadioss 2023.09.15CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 77975 Joules per run.2. Optimized Power Mode: Approximate power consumption of 76688 Joules per run.

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode100200300400500SE +/- 0.27, N = 3SE +/- 0.42, N = 3463.09461.15
MinAvgMaxDefault80026163905Optimized Power Mode80026823914OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207572759Optimized Power Mode177539759OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 29341 Joules per run.2. Optimized Power Mode: Approximate power consumption of 28410 Joules per run.

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 2023.2.devModel: Face Detection Retail FP16-INT8 - Device: CPUDefaultOptimized Power Mode1.15882.31763.47644.63525.794SE +/- 0.01, N = 3SE +/- 0.00, N = 35.155.13MIN: 4.63 / MAX: 27.9MIN: 4.64 / MAX: 29.561. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie
MinAvgMaxDefault80024553908Optimized Power Mode80025043910OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2023.2.devCPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207657771Optimized Power Mode178648743OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.2.devCPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 43563 Joules per run.2. Optimized Power Mode: Approximate power consumption of 42971 Joules per run.

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: MediumDefaultOptimized Power Mode306090120150SE +/- 0.46, N = 8SE +/- 0.58, N = 8131.14130.641. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: MediumDefaultOptimized Power Mode0.10130.20260.30390.40520.50650.3720.450
MinAvgMaxDefault80034903908Optimized Power Mode80034713914OpenBenchmarking.orgMegahertz, More Is BetterKvazaar 2.2CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault199.2352.8497.3Optimized Power Mode184.0290.5415.7OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 3131 Joules per run.2. Optimized Power Mode: Approximate power consumption of 2579 Joules per run.

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.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Person Vehicle Bike Detection FP16 - Device: CPUDefaultOptimized Power Mode2K4K6K8K10KSE +/- 7.22, N = 3SE +/- 6.61, N = 310205.6110168.061. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

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: CryptoNight-Heavy - Hash Count: 1MDefaultOptimized Power Mode15K30K45K60K75KSE +/- 49.63, N = 4SE +/- 96.69, N = 370573.070338.61. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
OpenBenchmarking.orgH/s Per Watt, More Is BetterXmrig 6.21Variant: CryptoNight-Heavy - Hash Count: 1MDefaultOptimized Power Mode306090120150126.16131.15
MinAvgMaxDefault80028103906Optimized Power Mode80028253905OpenBenchmarking.orgMegahertz, More Is BetterXmrig 6.21CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault205559759Optimized Power Mode182536759OpenBenchmarking.orgWatts, Fewer Is BetterXmrig 6.21CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 11607 Joules per run.2. Optimized Power Mode: Approximate power consumption of 11263 Joules per run.

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 2023.2.devModel: Person Vehicle Bike Detection FP16 - Device: CPUDefaultOptimized Power Mode3691215SE +/- 0.01, N = 3SE +/- 0.01, N = 312.5112.55MIN: 10.84 / MAX: 42.36MIN: 10.77 / MAX: 42.911. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie
MinAvgMaxDefault80024923908Optimized Power Mode80025623921OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2023.2.devCPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207644768Optimized Power Mode179628762OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.2.devCPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 44425 Joules per run.2. Optimized Power Mode: Approximate power consumption of 43928 Joules per run.

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: 50DefaultOptimized Power Mode700K1400K2100K2800K3500KSE +/- 12726.60, N = 3SE +/- 1584.15, N = 33060186.423050498.001. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3
OpenBenchmarking.orgRequests Per Second Per Watt, More Is BetterRedis 7.0.4Test: SET - Parallel Connections: 50DefaultOptimized Power Mode4K8K12K16K20K11959.0716432.18
MinAvgMaxDefault80035543904Optimized Power Mode80036253904OpenBenchmarking.orgMegahertz, More Is BetterRedis 7.0.4CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206.7255.9290.9Optimized Power Mode183.2185.6189.5OpenBenchmarking.orgWatts, Fewer Is BetterRedis 7.0.4CPU Power Consumption Monitor701402102803501. Default: Approximate power consumption of 6909 Joules per run.2. Optimized Power Mode: Approximate power consumption of 4950 Joules per run.

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.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Person Detection FP16 - Device: CPUDefaultOptimized Power Mode160320480640800SE +/- 0.71, N = 3SE +/- 0.90, N = 3748.10750.431. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.2.devModel: Person Detection FP16 - Device: CPUDefaultOptimized Power Mode1020304050SE +/- 0.04, N = 3SE +/- 0.05, N = 342.7242.59MIN: 31.92 / MAX: 84.1MIN: 32.12 / MAX: 128.961. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie
MinAvgMaxDefault80026013905Optimized Power Mode80028643906OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2023.2.devCPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207650802Optimized Power Mode181634803OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.2.devCPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 43580 Joules per run.2. Optimized Power Mode: Approximate power consumption of 42470 Joules per run.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.2.devModel: Face Detection FP16-INT8 - Device: CPUDefaultOptimized Power Mode50100150200250SE +/- 0.35, N = 3SE +/- 0.55, N = 3236.70237.36MIN: 157.14 / MAX: 268.19MIN: 160.6 / MAX: 286.171. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie
MinAvgMaxDefault80027063904Optimized Power Mode80027923907OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2023.2.devCPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207635764Optimized Power Mode182615731OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.2.devCPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 45280 Joules per run.2. Optimized Power Mode: Approximate power consumption of 44702 Joules per run.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Face Detection FP16-INT8 - Device: CPUDefaultOptimized Power Mode120240360480600SE +/- 0.65, N = 3SE +/- 1.35, N = 3539.69538.201. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode48121620SE +/- 0.02, N = 3SE +/- 0.01, N = 313.9513.91
MinAvgMaxDefault80030153909Optimized Power Mode80029943907OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206568757Optimized Power Mode176525694OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 29519 Joules per run.2. Optimized Power Mode: Approximate power consumption of 27995 Joules per run.

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-ThreadedDefaultOptimized Power Mode8001600240032004000SE +/- 1.44, N = 3SE +/- 2.92, N = 33650.53641.41. (CXX) g++ options: -O3 -march=native -fPIE -pie
OpenBenchmarking.orgMFLOPS Per Watt, More Is BetterQuantLib 1.32Configuration: Single-ThreadedDefaultOptimized Power Mode51015202518.0320.00
MinAvgMaxDefault80036653905Optimized Power Mode80037833903OpenBenchmarking.orgMegahertz, More Is BetterQuantLib 1.32CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault195.7202.5274.3Optimized Power Mode179.2182.1191.9OpenBenchmarking.orgWatts, Fewer Is BetterQuantLib 1.32CPU Power Consumption Monitor701402102803501. Default: Approximate power consumption of 5332 Joules per run.2. Optimized Power Mode: Approximate power consumption of 4795 Joules per run.

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: CryptoNight-Femto UPX2 - Hash Count: 1MDefaultOptimized Power Mode15K30K45K60K75KSE +/- 844.15, N = 4SE +/- 443.65, N = 369677.869847.91. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
OpenBenchmarking.orgH/s Per Watt, More Is BetterXmrig 6.21Variant: CryptoNight-Femto UPX2 - Hash Count: 1MDefaultOptimized Power Mode306090120150123.55130.14
MinAvgMaxDefault80028623904Optimized Power Mode80029233903OpenBenchmarking.orgMegahertz, More Is BetterXmrig 6.21CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault204564759Optimized Power Mode187537759OpenBenchmarking.orgWatts, Fewer Is BetterXmrig 6.21CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 11843 Joules per run.2. Optimized Power Mode: Approximate power consumption of 11450 Joules per run.

OpenBenchmarking.orgH/s, More Is BetterXmrig 6.21Variant: Monero - Hash Count: 1MDefaultOptimized Power Mode15K30K45K60K75KSE +/- 27.69, N = 4SE +/- 41.44, N = 370469.870298.81. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
OpenBenchmarking.orgH/s Per Watt, More Is BetterXmrig 6.21Variant: Monero - Hash Count: 1MDefaultOptimized Power Mode306090120150125.59130.23
MinAvgMaxDefault80026983904Optimized Power Mode80029143906OpenBenchmarking.orgMegahertz, More Is BetterXmrig 6.21CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207561759Optimized Power Mode183540759OpenBenchmarking.orgWatts, Fewer Is BetterXmrig 6.21CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 11363 Joules per run.2. Optimized Power Mode: Approximate power consumption of 11516 Joules per run.

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode10002000300040005000SE +/- 6.99, N = 3SE +/- 2.64, N = 34580.634590.34

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: 256 - Buffer Length: 256 - Filter Length: 512DefaultOptimized Power Mode500M1000M1500M2000M2500MSE +/- 2968913.01, N = 3SE +/- 1101514.11, N = 3214456666721410000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 256 - Buffer Length: 256 - Filter Length: 512DefaultOptimized Power Mode700K1400K2100K2800K3500K3389797.473430132.22
MinAvgMaxDefault80026103904Optimized Power Mode80027483907OpenBenchmarking.orgMegahertz, More Is BetterLiquid-DSP 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault196633756Optimized Power Mode181624755OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 22143 Joules per run.2. Optimized Power Mode: Approximate power consumption of 21846 Joules per run.

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode1224364860SE +/- 0.05, N = 3SE +/- 0.07, N = 351.7051.62
MinAvgMaxDefault80026303909Optimized Power Mode80028793908OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206593760Optimized Power Mode179575751OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 27295 Joules per run.2. Optimized Power Mode: Approximate power consumption of 26823 Joules per run.

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.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Vehicle Detection FP16-INT8 - Device: CPUDefaultOptimized Power Mode2K4K6K8K10KSE +/- 5.30, N = 3SE +/- 4.32, N = 38930.008943.881. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode2004006008001000SE +/- 0.30, N = 3SE +/- 0.10, N = 3862.14863.35

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 2023.2.devModel: Vehicle Detection FP16-INT8 - Device: CPUDefaultOptimized Power Mode48121620SE +/- 0.01, N = 3SE +/- 0.01, N = 314.3214.30MIN: 12.39 / MAX: 39.16MIN: 12.18 / MAX: 47.31. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie
MinAvgMaxDefault80024383911Optimized Power Mode80024083907OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2023.2.devCPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault206658763Optimized Power Mode178650760OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.2.devCPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 43415 Joules per run.2. Optimized Power Mode: Approximate power consumption of 43323 Joules per run.

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode30060090012001500SE +/- 1.85, N = 3SE +/- 2.51, N = 31235.411237.03

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 RatingDefaultOptimized Power Mode140K280K420K560K700KSE +/- 4040.78, N = 3SE +/- 9574.07, N = 36373116381361. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
OpenBenchmarking.orgMIPS Per Watt, More Is Better7-Zip Compression 22.01Test: Decompression RatingDefaultOptimized Power Mode20040060080010001062.101145.35
MinAvgMaxDefault80029403908Optimized Power Mode80030063909OpenBenchmarking.orgMegahertz, More Is Better7-Zip Compression 22.01CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault205600762Optimized Power Mode179557760OpenBenchmarking.orgWatts, Fewer Is Better7-Zip Compression 22.01CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 29202 Joules per run.2. Optimized Power Mode: Approximate power consumption of 27672 Joules per run.

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 2023.2.devModel: Road Segmentation ADAS FP16-INT8 - Device: CPUDefaultOptimized Power Mode1224364860SE +/- 0.08, N = 3SE +/- 0.16, N = 353.5153.45MIN: 43.94 / MAX: 101.68MIN: 45.01 / MAX: 111.521. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie
MinAvgMaxDefault80025283908Optimized Power Mode80025553906OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2023.2.devCPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault207648756Optimized Power Mode178636758OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.2.devCPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 44067 Joules per run.2. Optimized Power Mode: Approximate power consumption of 43443 Joules per run.

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode1632486480SE +/- 0.05, N = 3SE +/- 0.03, N = 374.0773.99
MinAvgMaxDefault80024443904Optimized Power Mode80024953907OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault105617759Optimized Power Mode179605760OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 26752 Joules per run.2. Optimized Power Mode: Approximate power consumption of 26420 Joules per run.

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.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Road Segmentation ADAS FP16-INT8 - Device: CPUDefaultOptimized Power Mode5001000150020002500SE +/- 3.42, N = 3SE +/- 7.16, N = 32390.252392.381. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode20406080100SE +/- 0.04, N = 3SE +/- 0.04, N = 376.5476.50
MinAvgMaxDefault80024553905Optimized Power Mode80025373904OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.6CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault204612758Optimized Power Mode178602760OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.6CPU Power Consumption Monitor20040060080010001. Default: Approximate power consumption of 26724 Joules per run.2. Optimized Power Mode: Approximate power consumption of 26477 Joules per run.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-StreamDefaultOptimized Power Mode2004006008001000SE +/- 0.97, N = 3SE +/- 0.60, N = 3834.18834.27

CPU Power Consumption Monitor

OpenBenchmarking.orgWattsCPU Power Consumption MonitorPhoronix Test Suite System MonitoringDefaultOptimized Power Mode140280420560700Min: 101.15 / Avg: 445.93 / Max: 802.3Min: 88.73 / Avg: 366.23 / Max: 802.53

CPU Peak Freq (Highest CPU Core Frequency) Monitor

OpenBenchmarking.orgMegahertzCPU Peak Freq (Highest CPU Core Frequency) MonitorPhoronix Test Suite System MonitoringDefaultOptimized Power Mode10002000300040005000Min: 500 / Avg: 3374.53 / Max: 5474Min: 800 / Avg: 3437 / Max: 5154

Blender

OpenBenchmarking.orgWatts, Fewer Is BetterBlender 4.0CPU Power Consumption MonitorDefaultOptimized Power Mode130260390520650Min: 206.7 / Avg: 601.32 / Max: 759.63Min: 179.1 / Avg: 653.77 / Max: 759.95

OpenBenchmarking.orgMegahertz, More Is BetterBlender 4.0CPU Peak Freq (Highest CPU Core Frequency) MonitorDefaultOptimized Power Mode7001400210028003500Min: 500 / Avg: 2658.65 / Max: 3908Min: 800 / Avg: 2769.56 / Max: 3909

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.0Blend File: Barbershop - Compute: CPU-OnlyDefaultOptimized Power Mode306090120150SE +/- 6.68, N = 9SE +/- 0.11, N = 3149.49129.94
OpenBenchmarking.orgMegahertz, More Is BetterBlender 4.0CPU Peak Freq (Highest CPU Core Frequency) MonitorDefaultOptimized Power Mode7001400210028003500Min: 500 / Avg: 2658.65 / Max: 3908Min: 800 / Avg: 2769.56 / Max: 3909
OpenBenchmarking.orgWatts, Fewer Is BetterBlender 4.0CPU Power Consumption MonitorDefaultOptimized Power Mode130260390520650Min: 206.7 / Avg: 601.32 / Max: 759.63Min: 179.1 / Avg: 653.77 / Max: 759.951. Default: Approximate power consumption of 93071 Joules per run.2. Optimized Power Mode: Approximate power consumption of 89349 Joules per run.

PostgreSQL

OpenBenchmarking.orgWatts, Fewer Is BetterPostgreSQL 16CPU Power Consumption MonitorDefaultOptimized Power Mode130260390520650Min: 199.24 / Avg: 508.51 / Max: 759.36Min: 94.15 / Avg: 458.57 / Max: 678.78

OpenBenchmarking.orgMegahertz, More Is BetterPostgreSQL 16CPU Peak Freq (Highest CPU Core Frequency) MonitorDefaultOptimized Power Mode7001400210028003500Min: 800 / Avg: 3000.4 / Max: 3925Min: 800 / Avg: 2988.27 / Max: 3921

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average LatencyDefaultOptimized Power Mode0.25670.51340.77011.02681.2835SE +/- 0.022, N = 10SE +/- 0.016, N = 121.1241.1411. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgMegahertz, More Is BetterPostgreSQL 16CPU Peak Freq (Highest CPU Core Frequency) MonitorDefaultOptimized Power Mode7001400210028003500Min: 800 / Avg: 3000.4 / Max: 3925Min: 800 / Avg: 2988.27 / Max: 3921
OpenBenchmarking.orgWatts, Fewer Is BetterPostgreSQL 16CPU Power Consumption MonitorDefaultOptimized Power Mode130260390520650Min: 199.24 / Avg: 508.51 / Max: 759.36Min: 94.15 / Avg: 458.57 / Max: 678.781. Default: Approximate power consumption of 68242 Joules per run.2. Optimized Power Mode: Approximate power consumption of 63054 Joules per run.

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 1000 - Mode: Read OnlyDefaultOptimized Power Mode200K400K600K800K1000KSE +/- 17460.42, N = 10SE +/- 11877.48, N = 128930898784681. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

Memcached

MinAvgMaxDefault199.2534.2599.5Optimized Power Mode94.3422.8504.5OpenBenchmarking.orgWatts, Fewer Is BetterMemcached 1.6.19CPU Power Consumption Monitor160320480640800

MinAvgMaxDefault80029083906Optimized Power Mode80029253905OpenBenchmarking.orgMegahertz, More Is BetterMemcached 1.6.19CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000

OpenBenchmarking.orgOps/sec Per Watt, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:10DefaultOptimized Power Mode130026003900520065006290.445158.07

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:10DefaultOptimized Power Mode700K1400K2100K2800K3500KSE +/- 29533.59, N = 15SE +/- 43551.88, N = 153360511.122180648.211. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
OpenBenchmarking.orgOps/sec Per Watt, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:10DefaultOptimized Power Mode130026003900520065006290.445158.07
MinAvgMaxDefault80029083906Optimized Power Mode80029253905OpenBenchmarking.orgMegahertz, More Is BetterMemcached 1.6.19CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault199.2534.2599.5Optimized Power Mode94.3422.8504.5OpenBenchmarking.orgWatts, Fewer Is BetterMemcached 1.6.19CPU Power Consumption Monitor1603204806408001. Default: Approximate power consumption of 37289 Joules per run.2. Optimized Power Mode: Approximate power consumption of 29509 Joules per run.

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.

Tuning: Visual Quality Optimized - Input: Bosphorus 4K

Default: The test quit with a non-zero exit status.

Optimized Power Mode: The test quit with a non-zero exit status.

Tuning: PSNR/SSIM Optimized - Input: Bosphorus 4K

Default: The test quit with a non-zero exit status.

Optimized Power Mode: The test quit with a non-zero exit status.

MinAvgMaxDefault195.4271.1486.7Optimized Power Mode94.5216.8399.4OpenBenchmarking.orgWatts, Fewer Is BetterSVT-VP9 0.3CPU Power Consumption Monitor130260390520650

MinAvgMaxDefault80033963909Optimized Power Mode80036123910OpenBenchmarking.orgMegahertz, More Is BetterSVT-VP9 0.3CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000

OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 1080pDefaultOptimized Power Mode0.51191.02381.53572.04762.55952.0032.275

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 1080pDefaultOptimized Power Mode120240360480600SE +/- 2.30, N = 9SE +/- 9.88, N = 15542.98493.261. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 1080pDefaultOptimized Power Mode0.51191.02381.53572.04762.55952.0032.275
MinAvgMaxDefault80033963909Optimized Power Mode80036123910OpenBenchmarking.orgMegahertz, More Is BetterSVT-VP9 0.3CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000
MinAvgMaxDefault195.4271.1486.7Optimized Power Mode94.5216.8399.4OpenBenchmarking.orgWatts, Fewer Is BetterSVT-VP9 0.3CPU Power Consumption Monitor1302603905206501. Default: Approximate power consumption of 2048 Joules per run.2. Optimized Power Mode: Approximate power consumption of 1735 Joules per run.

Tuning: VMAF Optimized - Input: Bosphorus 4K

Default: The test quit with a non-zero exit status.

Optimized Power Mode: The test quit with a non-zero exit status.

easyWave

OpenBenchmarking.orgWatts, Fewer Is BettereasyWave r34CPU Power Consumption MonitorDefaultOptimized Power Mode100200300400500Min: 198.33 / Avg: 407.88 / Max: 539.38Min: 160.89 / Avg: 328.07 / Max: 496.2

OpenBenchmarking.orgMegahertz, More Is BettereasyWave r34CPU Peak Freq (Highest CPU Core Frequency) MonitorDefaultOptimized Power Mode7001400210028003500Min: 800 / Avg: 3436.05 / Max: 3906Min: 800 / Avg: 3485.76 / Max: 3906

OpenBenchmarking.orgSeconds, Fewer Is BettereasyWave r34Input: e2Asean Grid + BengkuluSept2007 Source - Time: 2400DefaultOptimized Power Mode20406080100SE +/- 1.41, N = 15SE +/- 1.07, N = 389.98102.931. (CXX) g++ options: -O3 -fopenmp
OpenBenchmarking.orgMegahertz, More Is BettereasyWave r34CPU Peak Freq (Highest CPU Core Frequency) MonitorDefaultOptimized Power Mode7001400210028003500Min: 800 / Avg: 3436.05 / Max: 3906Min: 800 / Avg: 3485.76 / Max: 3906
OpenBenchmarking.orgWatts, Fewer Is BettereasyWave r34CPU Power Consumption MonitorDefaultOptimized Power Mode100200300400500Min: 198.33 / Avg: 407.88 / Max: 539.38Min: 160.89 / Avg: 328.07 / Max: 496.21. Default: Approximate power consumption of 38994 Joules per run.2. Optimized Power Mode: Approximate power consumption of 35760 Joules per run.

171 Results Shown

PyTorch:
  CPU - 256 - ResNet-50
  CPU - 256 - ResNet-152
  CPU - 64 - ResNet-152
nginx
PyTorch
OpenFOAM
nginx
Redis
Zstd Compression
PyTorch
Xmrig
VVenC
uvg266
PyTorch
Zstd Compression
Apache Spark TPC-H
easyWave
Zstd Compression
SVT-AV1
DuckDB
uvg266:
  Bosphorus 4K - Super Fast
  Bosphorus 4K - Very Fast
Apache Hadoop
Apache HTTP Server
Zstd Compression
SVT-AV1:
  Preset 8 - Bosphorus 1080p
  Preset 8 - Bosphorus 4K
FFmpeg
VVenC
Zstd Compression
Y-Cruncher
Liquid-DSP
VVenC
Liquid-DSP
SVT-AV1:
  Preset 12 - Bosphorus 4K
  Preset 4 - Bosphorus 4K
  Preset 13 - Bosphorus 4K
  Preset 13 - Bosphorus 1080p
OpenRadioss
Kvazaar
uvg266
VVenC
uvg266:
  Bosphorus 1080p - Ultra Fast
  Bosphorus 1080p - Super Fast
OpenFOAM
Zstd Compression
SVT-VP9
Timed GCC Compilation
PostgreSQL:
  100 - 1000 - Read Write - Average Latency
  100 - 1000 - Read Write
SVT-VP9
Kvazaar:
  Bosphorus 1080p - Ultra Fast
  Bosphorus 4K - Very Fast
SVT-AV1
Apache HTTP Server
Timed LLVM Compilation
DuckDB
Kvazaar
7-Zip Compression
FFmpeg
Memcached
Kvazaar
Liquid-DSP
Kvazaar
Redis
Neural Magic DeepSparse:
  ResNet-50, Sparse INT8 - Asynchronous Multi-Stream:
    items/sec
    ms/batch
Blender
FFmpeg
Intel Open Image Denoise
FFmpeg
Liquid-DSP
OpenVINO
Liquid-DSP:
  32 - 256 - 512
  256 - 256 - 57
Timed Linux Kernel Compilation
Redis
Liquid-DSP
OpenVINO
OpenRadioss
Timed LLVM Compilation
uvg266
Liquid-DSP
QMCPACK
uvg266
Blender
uvg266
Xmrig
Kvazaar
OpenRadioss
OpenVINO
Neural Magic DeepSparse:
  BERT-Large, NLP Question Answering - Asynchronous Multi-Stream:
    items/sec
    ms/batch
Liquid-DSP
OpenVINO
Neural Magic DeepSparse
Timed Linux Kernel Compilation
OpenVINO:
  Handwritten English Recognition FP16-INT8 - CPU:
    ms
    FPS
Xmrig
Neural Magic DeepSparse:
  CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Stream
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream
  BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Stream
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream
  BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Stream
Kvazaar
QuantLib
Kvazaar
OpenRadioss
Liquid-DSP:
  64 - 256 - 32
  32 - 256 - 32
Neural Magic DeepSparse
uvg266
OpenRadioss
Neural Magic DeepSparse:
  ResNet-50, Baseline - Asynchronous Multi-Stream:
    ms/batch
    items/sec
OpenFOAM
OpenVINO:
  Face Detection Retail FP16-INT8 - CPU
  Machine Translation EN To DE FP16 - CPU
Neural Magic DeepSparse:
  NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream
  NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream
OpenVINO
OpenRadioss
Neural Magic DeepSparse
OpenVINO
Kvazaar
OpenVINO
Xmrig
OpenVINO
Redis
OpenVINO:
  Person Detection FP16 - CPU:
    FPS
    ms
  Face Detection FP16-INT8 - CPU:
    ms
    FPS
Neural Magic DeepSparse
QuantLib
Xmrig:
  CryptoNight-Femto UPX2 - 1M
  Monero - 1M
Neural Magic DeepSparse
Liquid-DSP
Neural Magic DeepSparse
OpenVINO
Neural Magic DeepSparse
OpenVINO
Neural Magic DeepSparse
7-Zip Compression
OpenVINO
Neural Magic DeepSparse
OpenVINO
Neural Magic DeepSparse:
  CV Detection, YOLOv5s COCO - Asynchronous Multi-Stream:
    ms/batch
    items/sec
CPU Power Consumption Monitor:
  Phoronix Test Suite System Monitoring:
    Watts
    Megahertz
  CPU Power Consumption Monitor:
    Watts
  CPU Peak Freq (Highest CPU Core Frequency) Monitor:
    Megahertz
Blender
PostgreSQL:
  CPU Power Consumption Monitor
  CPU Peak Freq (Highest CPU Core Frequency) Monitor
PostgreSQL:
  100 - 1000 - Read Only - Average Latency
  100 - 1000 - Read Only
Memcached:
  CPU Power Consumption Monitor
  CPU Peak Freq (Highest CPU Core Frequency) Monitor
  1:10
Memcached
SVT-VP9:
  CPU Power Consumption Monitor
  CPU Peak Freq (Highest CPU Core Frequency) Monitor
  VMAF Optimized - Bosphorus 1080p
SVT-VP9
easyWave:
  CPU Power Consumption Monitor
  CPU Peak Freq (Highest CPU Core Frequency) Monitor
easyWave

Default

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --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-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/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-build-config=bootstrap-lto-lean --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
Processor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000161
Java Notes: OpenJDK Runtime Environment (build 11.0.21+9-post-Ubuntu-0ubuntu123.10)
Python Notes: Python 3.11.6
Security Notes: 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

Testing initiated at 14 December 2023 16:53 by user phoronix.

Optimized Power Mode

Processor: 2 x INTEL XEON PLATINUM 8592+ @ 3.90GHz (128 Cores / 256 Threads), Motherboard: Quanta Cloud S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 1008GB, Disk: 3201GB Micron_7450_MTFDKCB3T2TFS, Graphics: ASPEED, Network: 2 x Intel X710 for 10GBASE-T

OS: Ubuntu 23.10, Kernel: 6.5.0-13-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1080

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --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-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/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-build-config=bootstrap-lto-lean --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
Processor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000161
Java Notes: OpenJDK Runtime Environment (build 11.0.21+9-post-Ubuntu-0ubuntu123.10)
Python Notes: Python 3.11.6
Security Notes: 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

Testing initiated at 15 December 2023 10:44 by user phoronix.