7700K Intel 2020

Intel Core i7-7700K testing with a MSI Z270-A PRO (MS-7A71) v1.0 (1.60 BIOS) and MSI Intel HD 630 3GB on Ubuntu 20.04 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2010143-FI-7700KINTE31
Jump To Table - Results

View

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

Limit displaying results to tests within:

AV1 2 Tests
Bioinformatics 2 Tests
BLAS (Basic Linear Algebra Sub-Routine) Tests 3 Tests
Timed Code Compilation 3 Tests
C/C++ Compiler Tests 10 Tests
Compression Tests 4 Tests
CPU Massive 22 Tests
Creator Workloads 17 Tests
Database Test Suite 3 Tests
Encoding 2 Tests
Fortran Tests 5 Tests
HPC - High Performance Computing 25 Tests
Imaging 5 Tests
Java 4 Tests
Common Kernel Benchmarks 2 Tests
Machine Learning 12 Tests
Molecular Dynamics 5 Tests
MPI Benchmarks 5 Tests
Multi-Core 16 Tests
NVIDIA GPU Compute 10 Tests
OCR 2 Tests
Intel oneAPI 2 Tests
OpenMPI Tests 6 Tests
Programmer / Developer System Benchmarks 5 Tests
Python 2 Tests
Renderers 2 Tests
Scientific Computing 11 Tests
Server 3 Tests
Server CPU Tests 11 Tests
Single-Threaded 4 Tests
Speech 2 Tests
Telephony 2 Tests
Video Encoding 2 Tests
Vulkan Compute 3 Tests
Common Workstation Benchmarks 2 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

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
Linux 5.4
October 13 2020
  1 Day, 8 Minutes
Linux 5.8
October 13 2020
  22 Hours, 53 Minutes
Linux 5.9
October 13 2020
  20 Hours, 41 Minutes
Invert Hiding All Results Option
  22 Hours, 34 Minutes

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


7700K Intel 2020ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLOpenCLCompilerFile-SystemScreen ResolutionLinux 5.4Linux 5.8Linux 5.9Intel Core i7-7700K @ 4.50GHz (4 Cores / 8 Threads)MSI Z270-A PRO (MS-7A71) v1.0 (1.60 BIOS)Intel Xeon E3-1200 v6/7th + Z27016GB256GB Samsung SSD 850MSI Intel HD 630 3GB (1150MHz)Realtek ALC892LG Ultra HDRealtek RTL8111/8168/8411Ubuntu 20.045.4.0-28-generic (x86_64)GNOME Shell 3.36.1X Server 1.20.8modesetting 1.20.84.6 Mesa 20.0.4OpenCL 2.1GCC 9.3.0ext43840x21605.8.14-050814-generic (x86_64)GNOME Shell 3.36.45.9.0-050900rc8daily20201011-generic (x86_64) 20201010OpenBenchmarking.orgCompiler Details- Linux 5.4: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,gm2 --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v - Linux 5.8: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,gm2 --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-9-HskZEa/gcc-9-9.3.0/debian/tmp-nvptx/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v - Linux 5.9: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,gm2 --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-9-HskZEa/gcc-9-9.3.0/debian/tmp-nvptx/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Processor Details- Linux 5.4: Scaling Governor: intel_pstate powersave - CPU Microcode: 0xca- Linux 5.8: Scaling Governor: intel_pstate powersave - CPU Microcode: 0xd6- Linux 5.9: Scaling Governor: intel_pstate powersave - CPU Microcode: 0xd6Java Details- Linux 5.4: OpenJDK Runtime Environment (build 11.0.7+10-post-Ubuntu-3ubuntu1)- Linux 5.8: OpenJDK Runtime Environment (build 11.0.8+10-post-Ubuntu-0ubuntu120.04)- Linux 5.9: OpenJDK Runtime Environment (build 11.0.8+10-post-Ubuntu-0ubuntu120.04)Python Details- Linux 5.4: Python 3.8.2- Linux 5.8: Python 3.8.5- Linux 5.9: Python 3.8.5Security Details- Linux 5.4: itlb_multihit: KVM: Mitigation of Split huge pages + l1tf: Mitigation of PTE Inversion; VMX: conditional cache flushes SMT vulnerable + mds: Mitigation of Clear buffers; SMT vulnerable + meltdown: Mitigation of PTI + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full generic retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + tsx_async_abort: Mitigation of Clear buffers; SMT vulnerable - Linux 5.8: itlb_multihit: KVM: Mitigation of VMX disabled + l1tf: Mitigation of PTE Inversion; VMX: conditional cache flushes SMT vulnerable + mds: Mitigation of Clear buffers; SMT vulnerable + meltdown: Mitigation of PTI + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full generic retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Mitigation of Microcode + tsx_async_abort: Mitigation of Clear buffers; SMT vulnerable - Linux 5.9: itlb_multihit: KVM: Mitigation of VMX disabled + l1tf: Mitigation of PTE Inversion; VMX: conditional cache flushes SMT vulnerable + mds: Mitigation of Clear buffers; SMT vulnerable + meltdown: Mitigation of PTI + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full generic retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Mitigation of Microcode + tsx_async_abort: Mitigation of Clear buffers; SMT vulnerable

Linux 5.4Linux 5.8Linux 5.9Logarithmic Result OverviewPhoronix Test SuiteRealSR-NCNNKripkeNCNNApache CouchDBStress-NGeSpeak-NG Speech EngineWireGuard + Linux Networking Stack Stress TestKeyDBVkFFTFFTEGROMACSGLmark2DaCapo BenchmarkNeatBenchRenaissanceMobile Neural NetworkJava Gradle BuildLeelaChessZeroInfluxDB7-Zip CompressionLuxCoreRenderNAMDHuginDarmstadt Automotive Parallel Heterogeneous SuiteMPVSunflow Rendering Systemlibavif avifencBuild2System GZIP DecompressionXZ CompressionTimed Linux Kernel CompilationIncompact3DPyPerformanceTimed MAFFT AlignmentTimed Apache CompilationAI Benchmark AlphaGPAWOpenVINOoneDNNdav1dRodiniaBYTE Unix BenchmarkBRL-CADHierarchical INTegrationMontage Astronomical Image Mosaic EngineLibRawTesseract OCROCRMyPDFDolfynMlpack BenchmarkTimed HMMer SearchASTC EncoderWebP Image EncodeChaos Group V-RAYLAMMPS Molecular Dynamics SimulatorZstd CompressionRNNoiseTensorFlow LiteCaffeOpenCVTNNMonte Carlo Simulations of Ionised Nebulae

7700K Intel 2020stress-ng: RdRandncnn: Vulkan GPU-v3-v3 - mobilenet-v3ncnn: Vulkan GPU - efficientnet-b0ncnn: Vulkan GPU - mnasnetncnn: Vulkan GPU-v2-v2 - mobilenet-v2ncnn: Vulkan GPU - shufflenet-v2ncnn: Vulkan GPU - googlenetncnn: Vulkan GPU - resnet18ncnn: Vulkan GPU - mobilenetncnn: Vulkan GPU - squeezenetncnn: Vulkan GPU - alexnetncnn: Vulkan GPU - resnet50ncnn: Vulkan GPU - yolov4-tinystress-ng: Semaphoresncnn: Vulkan GPU - vgg16stress-ng: NUMAcouchdb: 100 - 1000 - 24renaissance: Savina Reactors.IOstress-ng: Forkingstress-ng: MMAPstress-ng: MEMFDstress-ng: Context Switchingespeak: Text-To-Speech Synthesisinfluxdb: 4 - 10000 - 2,5000,1 - 10000wireguard: keydb: vkfft: stress-ng: Socket Activityffte: N=256, 3D Complex FFT Routinerenaissance: In-Memory Database Shootoutglmark2: 1920 x 1080glmark2: 1920 x 1200stress-ng: SENDFILEcaffe: AlexNet - CPU - 200stress-ng: Atomicmnn: resnet-v2-50onednn: Deconvolution Batch deconv_3d - f32 - CPUlczero: BLAScaffe: GoogleNet - CPU - 200neatbench: CPUcaffe: GoogleNet - CPU - 100glmark2: 2560 x 1440caffe: AlexNet - CPU - 100lczero: Eigenmnn: inception-v3mnn: SqueezeNetV1.0daphne: OpenMP - Points2Imagejava-gradle-perf: Reactormlpack: scikit_qdarodinia: OpenMP Leukocytedacapobench: Jythonmnn: mobilenet-v1-1.0compress-7zip: Compress Speed Testdacapobench: Tradesoapglmark2: 3840 x 2160luxcorerender: DLSCmpv: Big Buck Bunny Sunflower 1080p - Software Onlymnn: MobileNetV2_224onednn: Convolution Batch Shapes Auto - u8s8f32 - CPUluxcorerender: Rainbow Colors and Prismnamd: ATPase Simulation - 327,506 Atomsmlpack: scikit_icancnn: CPU - mnasnetlczero: Randrenaissance: Scala Dottyonednn: Matrix Multiply Batch Shapes Transformer - f32 - CPUavifenc: 8hugin: Panorama Photo Assistant + Stitching Timepyperformance: goopencv: Object Detectionmlpack: scikit_linearridgeregressionpyperformance: crypto_pyaesonednn: Convolution Batch Shapes Auto - f32 - CPUonednn: Deconvolution Batch deconv_1d - f32 - CPUavifenc: 10sunflow: Global Illumination + Image Synthesisonednn: IP Batch 1D - u8s8f32 - CPUstress-ng: Memory Copyingopenvino: Face Detection 0106 FP16 - CPUonednn: IP Batch All - u8s8f32 - CPUopenvino: Face Detection 0106 FP32 - CPUavifenc: 0pyperformance: regex_compileonednn: Recurrent Neural Network Training - f32 - CPUonednn: Recurrent Neural Network Inference - f32 - CPUstress-ng: Glibc C String Functionspyperformance: django_templateonednn: Matrix Multiply Batch Shapes Transformer - u8s8f32 - CPUpyperformance: pathlibopenvino: Person Detection 0106 FP32 - CPUopencv: Features 2Dopenvino: Person Detection 0106 FP16 - CPUstress-ng: CPU Cacheopenvino: Person Detection 0106 FP32 - CPUonednn: IP Batch 1D - f32 - CPUdav1d: Summer Nature 4Kpyperformance: json_loadsonednn: Deconvolution Batch deconv_1d - u8s8f32 - CPUopenvino: Face Detection 0106 FP32 - CPUai-benchmark: Device Inference Scorerodinia: OpenMP LavaMDmpv: Big Buck Bunny Sunflower 4K - Software Onlybuild2: Time To Compileincompact3d: Cylindersystem-decompress-gzip: openvino: Person Detection 0106 FP16 - CPUcompress-xz: Compressing ubuntu-16.04.3-server-i386.img, Compression Level 9onednn: Deconvolution Batch deconv_3d - u8s8f32 - CPUncnn: CPU - squeezenetopenvino: Face Detection 0106 FP16 - CPUncnn: CPU - alexnetdav1d: Chimera 1080p 10-bitstress-ng: Glibc Qsort Data Sortingbuild-linux-kernel: Time To Compilestress-ng: Matrix Mathncnn: CPU - blazefacemafft: Multiple Sequence Alignment - LSU RNAastcenc: Exhaustiveai-benchmark: Device AI Scorebuild-apache: Time To Compilestress-ng: CPU Stressncnn: CPU - mobilenetncnn: CPU - shufflenet-v2gpaw: Carbon Nanotubepyperformance: raytracerenaissance: Akka Unbalanced Cobwebbed Treeonednn: IP Batch All - f32 - CPUdaphne: OpenMP - Euclidean Clusterai-benchmark: Device Training Scoreavifenc: 2dav1d: Summer Nature 1080pncnn: CPU - yolov4-tinyastcenc: Fastopenvino: Age Gender Recognition Retail 0013 FP32 - CPUpyperformance: python_startupdav1d: Chimera 1080pbyte: Dhrystone 2brl-cad: VGR Performance Metrichint: FLOATpyperformance: 2to3stress-ng: Mallocwebp: Quality 100, Losslesspyperformance: chaosncnn: CPU-v2-v2 - mobilenet-v2influxdb: 1024 - 10000 - 2,5000,1 - 10000tesseract-ocr: Time To OCR 7 Imagespyperformance: floatmontage: Mosaic of M17, K band, 1.5 deg x 1.5 deglibraw: Post-Processing Benchmarkinfluxdb: 64 - 10000 - 2,5000,1 - 10000ncnn: CPU - resnet18ocrmypdf: Processing 60 Page PDF Documentastcenc: Thoroughdolfyn: Computational Fluid Dynamicsrodinia: OpenMP CFD Solverncnn: CPU - efficientnet-b0astcenc: Mediumrodinia: OpenMP HotSpot3Dwebp: Quality 100, Lossless, Highest Compressionncnn: CPU - vgg16ncnn: CPU - googlenetmlpack: scikit_svmopenvino: Age Gender Recognition Retail 0013 FP16 - CPUstress-ng: Vector Mathtnn: CPU - MobileNet v2rodinia: OpenMP Streamclusterncnn: CPU - resnet50hmmer: Pfam Database Searchv-ray: CPUwebp: Quality 100compress-zstd: 3tensorflow-lite: SqueezeNetlammps: Rhodopsin Proteinwebp: Quality 100, Highest Compressionstress-ng: Cryptowebp: Defaulttensorflow-lite: Mobilenet Quantrnnoise: tnn: CPU - SqueezeNet v1.1tensorflow-lite: Inception ResNet V2tensorflow-lite: Mobilenet Floattensorflow-lite: NASNet Mobiletensorflow-lite: Inception V4pyperformance: pickle_pure_pythonpyperformance: nbodyopenvino: Age Gender Recognition Retail 0013 FP32 - CPUopenvino: Age Gender Recognition Retail 0013 FP16 - CPUncnn: CPU-v3-v3 - mobilenet-v3compress-zstd: 19mocassin: Dust 2D tau100.0opencv: DNN - Deep Neural Networkkripke: ncnn: Vulkan GPU - blazefacestress-ng: System V Message Passingdaphne: OpenMP - NDT Mappinggromacs: Water Benchmarkrenaissance: Apache Spark PageRankrenaissance: Apache Spark Bayesrenaissance: Apache Spark ALSrenaissance: Rand Forestdacapobench: H2realsr-ncnn: 4x - NoLinux 5.4Linux 5.8Linux 5.9195256.5537.7770.8234.8633.6423.1793.1679.7799.08106.61106.57141.04145.32763926.86250.13125.52152.44617713.73743990.5847.39366.661578574.1426.9991190894.9168.750440142.2711235364.4821538.7314462973052.94159352864128.69102389263212.7342.24711.460995325076110.41256573495129087647.6677.82425141.433877322199.23069.95201.49735325.9992899756341570.84150.374.35116.39020.923.1670256.545.302563471525.6624.128006.76061.699214361043.4894.021.71787.438206.2941.9483.005271367.012694.9140.56051.46147.086151482.653297.060605374.9340.65.7727617.70.891322080.9122.644399.206.54384104.6522.38.181472734.33838457.36237.19193.413600.6389572.8904345.9240.1945.9227120.521.4919.9883.4968.23159.95823746.731.7010.376464.58174726.6712253.8323.753.68484.9804007673.08998.3707990.8790987.881360.9132.667.573700.087.75389.1644618964.461836445325469.7847527930826272.7016.74090.26.341250226.322.62297.176.78433.621239227.018.1444.72257.9218.14938.7738.588.9799.79038.99978.9918.8123.653720.6834720.08308.55221.32937.52102.55377102.3482710.13992333.4417.015978.391.49927847224.077297.608520144327081329282757495603681091.071.075.37232412823228929746.194624409.89797.680.5864012.3482928.2722266.9401611.398295542.1246101.8113.0024.4712.1111.768.3033.4629.5337.2041.2747.0971.5077.65560118.49193.83104.83120.15315098.23842694.5042.37330.201749583.1729.6621084501.1184.544480395.4512114990.2223103.5952977123219.09762755360899.89107567262238.7043.99111.929299126050410.81303893615318890749.1698.06824549.315160252193.05671.94207.37135216.1562875657161610.86153.944.43716.18870.933.1390655.365.402613231521.4694.207416.85262.540210357253.4492.521.48417.473816.3621.9783.050351370.242657.9341.01471.47148.526149485.387297.446612988.3040.15.7068917.80.891307510.9122.554379.116.54102103.7622.18.133432710.49831453.96637.50195.005605.0761922.8804332.3240.0005.9639320.381.519.8682.9667.90159.02323793.601.6910.321461.89173826.5192264.7223.623.69487.4183997656.80298.0709987.2690787.518361.0332.537.593697.197.77389.3944673416.761606444887120.5889927830773901.0016.68389.96.361254110.122.69297.276.91033.521239154.618.1544.77957.7718.15038.6798.568.95100.00638.91579.1618.8523.603713.4034720.29308.57221.29337.49102.45677082.3462708.93988093.4387.017978.231.49827850124.067297.506520281027079129288757489103681091.071.075.37232411777567294172.424899988.40776.820.6194095.8082845.8262605.0711712.8342928258.7226099.9113.0224.4912.1111.778.7533.4929.5737.0441.2946.4471.4977.81560355.50193.3697.27122.08615086.86837586.6042.26333.901737256.0429.8551086068.7183.326474329.5012125084.4422961.9772365793253.71762455661511.86105821274452.4844.19311.667298725994010.81301513625282090149.3308.08424347.011593353197.76072.05206.11636226.1612826457801610.86153.914.45416.01790.943.1007656.065.292616261552.3364.145336.89062.879210354593.4292.421.37237.555726.3921.9563.018761351.142663.5941.12181.48149.099150479.076300.907609670.5240.45.7036317.60.901313760.922.804354.596.60699103.6322.18.206172716.75832453.55137.48194.408603.4241132.8694314.6939.9145.9568720.461.4919.9983.0167.80160.02023890.641.7010.315461.93173726.5362251.9923.623.70485.0334017634.86898.5451986.3890587.900359.5532.597.603711.837.78390.6444505024.961663443682873.6593127830880073.5816.73990.26.361252051.722.65697.477.02033.571242649.218.1044.84357.8418.19438.7208.568.9699.89138.93079.1618.8123.623712.8934648.21307.94721.28637.55102.40077182.3492706.73987493.4377.009979.171.49827861524.074297.517520109727087229283657484333681091.071.075.37232411783387284632.254608826.62798.880.6244135.9322863.1392642.1861685.9292935258.758OpenBenchmarking.org

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: RdRandLinux 5.4Linux 5.8Linux 5.940K80K120K160K200KSE +/- 2.87, N = 3SE +/- 1.96, N = 3SE +/- 1.98, N = 3195256.556101.816099.911. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3Linux 5.4Linux 5.8Linux 5.9918273645SE +/- 0.03, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 337.7713.0013.02MIN: 33.38 / MAX: 38.57MIN: 12.32 / MAX: 13.23MIN: 12.89 / MAX: 13.251. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: efficientnet-b0Linux 5.4Linux 5.8Linux 5.91632486480SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 370.8224.4724.49MIN: 69.67 / MAX: 71.79MIN: 23.73 / MAX: 24.58MIN: 23.83 / MAX: 25.591. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: mnasnetLinux 5.4Linux 5.8Linux 5.9816243240SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 334.8612.1112.11MIN: 34.26 / MAX: 35.46MIN: 12.07 / MAX: 12.17MIN: 11.59 / MAX: 12.261. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2Linux 5.4Linux 5.8Linux 5.9816243240SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 333.6411.7611.77MIN: 33.07 / MAX: 34.23MIN: 11.27 / MAX: 12.75MIN: 11.06 / MAX: 13.041. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: shufflenet-v2Linux 5.4Linux 5.8Linux 5.9612182430SE +/- 0.02, N = 3SE +/- 0.08, N = 3SE +/- 0.25, N = 323.178.308.75MIN: 21.86 / MAX: 25.39MIN: 7.58 / MAX: 8.86MIN: 7.72 / MAX: 9.771. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: googlenetLinux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.23, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 393.1633.4633.49MIN: 62.27 / MAX: 94.88MIN: 33.2 / MAX: 33.72MIN: 32.86 / MAX: 34.591. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: resnet18Linux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 379.7729.5329.57MIN: 76.91 / MAX: 83.23MIN: 29.19 / MAX: 30.71MIN: 29.16 / MAX: 29.671. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: mobilenetLinux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.14, N = 399.0837.2037.04MIN: 96.17 / MAX: 104.21MIN: 34.81 / MAX: 44.93MIN: 34.76 / MAX: 44.881. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: squeezenetLinux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.00, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3106.6141.2741.29MIN: 104.06 / MAX: 109.19MIN: 40.78 / MAX: 42.17MIN: 40.79 / MAX: 42.151. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: alexnetLinux 5.4Linux 5.8Linux 5.920406080100SE +/- 2.09, N = 3SE +/- 0.25, N = 3SE +/- 0.13, N = 3106.5747.0946.44MIN: 45.62 / MAX: 113.54MIN: 44.75 / MAX: 50.3MIN: 44.14 / MAX: 49.491. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: resnet50Linux 5.4Linux 5.8Linux 5.9306090120150SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3141.0471.5071.49MIN: 127.63 / MAX: 144.9MIN: 71.18 / MAX: 72.42MIN: 71.03 / MAX: 71.741. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: yolov4-tinyLinux 5.4Linux 5.8Linux 5.9306090120150SE +/- 0.11, N = 3SE +/- 0.19, N = 3SE +/- 0.11, N = 3145.3277.6577.81MIN: 108.89 / MAX: 162.25MIN: 75.84 / MAX: 78.61MIN: 76.22 / MAX: 114.271. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: SemaphoresLinux 5.4Linux 5.8Linux 5.9160K320K480K640K800KSE +/- 5286.21, N = 3SE +/- 6354.42, N = 3SE +/- 3455.29, N = 3763926.86560118.49560355.501. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: vgg16Linux 5.4Linux 5.8Linux 5.950100150200250SE +/- 0.03, N = 3SE +/- 0.62, N = 3SE +/- 0.15, N = 3250.13193.83193.36MIN: 245.47 / MAX: 262.02MIN: 190.96 / MAX: 196.34MIN: 191.52 / MAX: 195.181. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: NUMALinux 5.4Linux 5.8Linux 5.9306090120150SE +/- 1.10, N = 3SE +/- 0.68, N = 3SE +/- 0.66, N = 3125.52104.8397.271. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

Apache CouchDB

This is a bulk insertion benchmark of Apache CouchDB. CouchDB is a document-oriented NoSQL database implemented in Erlang. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterApache CouchDB 3.1.1Bulk Size: 100 - Inserts: 1000 - Rounds: 24Linux 5.4Linux 5.8Linux 5.9306090120150SE +/- 0.66, N = 3SE +/- 0.55, N = 3SE +/- 0.87, N = 3152.45120.15122.091. (CXX) g++ options: -std=c++14 -lmozjs-68 -lm -lerl_interface -lei -fPIC -MMD

Renaissance

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

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Savina Reactors.IOLinux 5.4Linux 5.8Linux 5.94K8K12K16K20KSE +/- 212.09, N = 5SE +/- 123.78, N = 5SE +/- 151.20, N = 2017713.7415098.2415086.87

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: ForkingLinux 5.4Linux 5.8Linux 5.99K18K27K36K45KSE +/- 32.69, N = 3SE +/- 234.95, N = 3SE +/- 201.96, N = 343990.5842694.5037586.601. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: MMAPLinux 5.4Linux 5.8Linux 5.91122334455SE +/- 0.23, N = 3SE +/- 0.73, N = 3SE +/- 0.14, N = 347.3942.3742.261. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: MEMFDLinux 5.4Linux 5.8Linux 5.980160240320400SE +/- 1.32, N = 3SE +/- 0.87, N = 3SE +/- 0.62, N = 3366.66330.20333.901. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Context SwitchingLinux 5.4Linux 5.8Linux 5.9400K800K1200K1600K2000KSE +/- 2474.15, N = 3SE +/- 2375.17, N = 3SE +/- 12341.27, N = 31578574.141749583.171737256.041. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

eSpeak-NG Speech Engine

This test times how long it takes the eSpeak speech synthesizer to read Project Gutenberg's The Outline of Science and output to a WAV file. This test profile is now tracking the eSpeak-NG version of eSpeak. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BettereSpeak-NG Speech Engine 20200907Text-To-Speech SynthesisLinux 5.4Linux 5.8Linux 5.9714212835SE +/- 0.18, N = 4SE +/- 0.27, N = 20SE +/- 0.21, N = 1727.0029.6629.861. (CC) gcc options: -O2 -std=c99

InfluxDB

This is a benchmark of the InfluxDB open-source time-series database optimized for fast, high-availability storage for IoT and other use-cases. The InfluxDB test profile makes use of InfluxDB Inch for facilitating the benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgval/sec, More Is BetterInfluxDB 1.8.2Concurrent Streams: 4 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000Linux 5.4Linux 5.8Linux 5.9300K600K900K1200K1500KSE +/- 2088.97, N = 3SE +/- 3705.44, N = 3SE +/- 5945.94, N = 31190894.91084501.11086068.7

WireGuard + Linux Networking Stack Stress Test

This is a benchmark of the WireGuard secure VPN tunnel and Linux networking stack stress test. The test runs on the local host but does require root permissions to run. The way it works is it creates three namespaces. ns0 has a loopback device. ns1 and ns2 each have wireguard devices. Those two wireguard devices send traffic through the loopback device of ns0. The end result of this is that tests wind up testing encryption and decryption at the same time -- a pretty CPU and scheduler-heavy workflow. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterWireGuard + Linux Networking Stack Stress TestLinux 5.4Linux 5.8Linux 5.94080120160200SE +/- 1.52, N = 3SE +/- 1.05, N = 3SE +/- 1.26, N = 3168.75184.54183.33

KeyDB

A benchmark of KeyDB as a multi-threaded fork of the Redis server. The KeyDB benchmark is conducted using memtier-benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterKeyDB 6.0.16Linux 5.4Linux 5.8Linux 5.9100K200K300K400K500KSE +/- 5820.40, N = 3SE +/- 552.43, N = 3SE +/- 861.37, N = 3440142.27480395.45474329.501. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

VkFFT

VkFFT is a Fast Fourier Transform (FFT) Library that is GPU accelerated by means of the Vulkan API. The VkFFT benchmark runs FFT performance differences of many different sizes before returning an overall benchmark score. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 2020-09-29Linux 5.4Linux 5.8Linux 5.930060090012001500SE +/- 0.58, N = 3SE +/- 0.67, N = 3112312111212

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Socket ActivityLinux 5.4Linux 5.8Linux 5.911002200330044005500SE +/- 67.02, N = 5SE +/- 53.60, N = 3SE +/- 51.41, N = 35364.484990.225084.441. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

FFTE

FFTE is a package by Daisuke Takahashi to compute Discrete Fourier Transforms of 1-, 2- and 3- dimensional sequences of length (2^p)*(3^q)*(5^r). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMFLOPS, More Is BetterFFTE 7.0N=256, 3D Complex FFT RoutineLinux 5.4Linux 5.8Linux 5.95K10K15K20K25KSE +/- 110.33, N = 3SE +/- 82.87, N = 3SE +/- 83.00, N = 321538.7323103.6022961.981. (F9X) gfortran options: -O3 -fomit-frame-pointer -fopenmp

Renaissance

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

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: In-Memory Database ShootoutLinux 5.4Linux 5.8Linux 5.97001400210028003500SE +/- 34.78, N = 5SE +/- 35.26, N = 25SE +/- 41.32, N = 53052.943219.103253.72

GLmark2

This is a test of Linaro's glmark2 port, currently using the X11 OpenGL 2.0 target. GLmark2 is a basic OpenGL benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterGLmark2 2020.04Resolution: 1920 x 1080Linux 5.4Linux 5.8Linux 5.9140280420560700593627624

OpenBenchmarking.orgScore, More Is BetterGLmark2 2020.04Resolution: 1920 x 1200Linux 5.4Linux 5.8Linux 5.9120240360480600528553556

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: SENDFILELinux 5.4Linux 5.8Linux 5.914K28K42K56K70KSE +/- 48.64, N = 3SE +/- 53.10, N = 3SE +/- 13.69, N = 364128.6960899.8961511.861. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

Caffe

This is a benchmark of the Caffe deep learning framework and currently supports the AlexNet and Googlenet model and execution on both CPUs and NVIDIA GPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: AlexNet - Acceleration: CPU - Iterations: 200Linux 5.4Linux 5.8Linux 5.920K40K60K80K100KSE +/- 85.08, N = 3SE +/- 1277.85, N = 14SE +/- 328.88, N = 31023891075671058211. (CXX) g++ options: -fPIC -O3 -rdynamic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: AtomicLinux 5.4Linux 5.8Linux 5.960K120K180K240K300KSE +/- 3268.99, N = 15SE +/- 3238.76, N = 15SE +/- 481.84, N = 3263212.73262238.70274452.481. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

Mobile Neural Network

MNN is the Mobile Neural Network as a highly efficient, lightweight deep learning framework developed by ALibaba. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2020-09-17Model: resnet-v2-50Linux 5.4Linux 5.8Linux 5.91020304050SE +/- 0.10, N = 3SE +/- 0.13, N = 3SE +/- 0.06, N = 342.2543.9944.19MIN: 41.79 / MAX: 57.27MIN: 43.66 / MAX: 111.84MIN: 43.82 / MAX: 89.331. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Deconvolution Batch deconv_3d - Data Type: f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.93691215SE +/- 0.02, N = 3SE +/- 0.07, N = 3SE +/- 0.04, N = 311.4611.9311.67MIN: 11.3MIN: 11.66MIN: 11.441. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

LeelaChessZero

LeelaChessZero (lc0 / lczero) is a chess engine automated vian neural networks. This test profile can be used for OpenCL, CUDA + cuDNN, and BLAS (CPU-based) benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.26Backend: BLASLinux 5.4Linux 5.8Linux 5.92004006008001000SE +/- 11.67, N = 3SE +/- 3.48, N = 3SE +/- 2.19, N = 39539919871. (CXX) g++ options: -flto -pthread

Caffe

This is a benchmark of the Caffe deep learning framework and currently supports the AlexNet and Googlenet model and execution on both CPUs and NVIDIA GPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: GoogleNet - Acceleration: CPU - Iterations: 200Linux 5.4Linux 5.8Linux 5.960K120K180K240K300KSE +/- 101.17, N = 3SE +/- 225.71, N = 3SE +/- 69.44, N = 32507612605042599401. (CXX) g++ options: -fPIC -O3 -rdynamic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

NeatBench

NeatBench is a benchmark of the cross-platform Neat Video software on the CPU and optional GPU (OpenCL / CUDA) support. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterNeatBench 5Acceleration: CPULinux 5.4Linux 5.8Linux 5.93691215SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.06, N = 310.410.810.8

Caffe

This is a benchmark of the Caffe deep learning framework and currently supports the AlexNet and Googlenet model and execution on both CPUs and NVIDIA GPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: GoogleNet - Acceleration: CPU - Iterations: 100Linux 5.4Linux 5.8Linux 5.930K60K90K120K150KSE +/- 103.41, N = 3SE +/- 63.52, N = 3SE +/- 34.95, N = 31256571303891301511. (CXX) g++ options: -fPIC -O3 -rdynamic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

GLmark2

This is a test of Linaro's glmark2 port, currently using the X11 OpenGL 2.0 target. GLmark2 is a basic OpenGL benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterGLmark2 2020.04Resolution: 2560 x 1440Linux 5.4Linux 5.8Linux 5.980160240320400349361362

Caffe

This is a benchmark of the Caffe deep learning framework and currently supports the AlexNet and Googlenet model and execution on both CPUs and NVIDIA GPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: AlexNet - Acceleration: CPU - Iterations: 100Linux 5.4Linux 5.8Linux 5.911K22K33K44K55KSE +/- 17.36, N = 3SE +/- 24.10, N = 3SE +/- 15.95, N = 35129053188528201. (CXX) g++ options: -fPIC -O3 -rdynamic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

LeelaChessZero

LeelaChessZero (lc0 / lczero) is a chess engine automated vian neural networks. This test profile can be used for OpenCL, CUDA + cuDNN, and BLAS (CPU-based) benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.26Backend: EigenLinux 5.4Linux 5.8Linux 5.92004006008001000SE +/- 4.93, N = 3SE +/- 4.73, N = 38769079011. (CXX) g++ options: -flto -pthread

Mobile Neural Network

MNN is the Mobile Neural Network as a highly efficient, lightweight deep learning framework developed by ALibaba. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2020-09-17Model: inception-v3Linux 5.4Linux 5.8Linux 5.91122334455SE +/- 0.15, N = 3SE +/- 0.07, N = 3SE +/- 0.02, N = 347.6749.1749.33MIN: 47.15 / MAX: 63.21MIN: 48.85 / MAX: 73.83MIN: 49.12 / MAX: 66.431. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2020-09-17Model: SqueezeNetV1.0Linux 5.4Linux 5.8Linux 5.9246810SE +/- 0.023, N = 3SE +/- 0.018, N = 3SE +/- 0.017, N = 37.8248.0688.084MIN: 7.71 / MAX: 27.2MIN: 7.99 / MAX: 12.11MIN: 8.02 / MAX: 11.771. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Darmstadt Automotive Parallel Heterogeneous Suite

DAPHNE is the Darmstadt Automotive Parallel HeterogeNEous Benchmark Suite with OpenCL / CUDA / OpenMP test cases for these automotive benchmarks for evaluating programming models in context to vehicle autonomous driving capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTest Cases Per Minute, More Is BetterDarmstadt Automotive Parallel Heterogeneous SuiteBackend: OpenMP - Kernel: Points2ImageLinux 5.4Linux 5.8Linux 5.95K10K15K20K25KSE +/- 216.15, N = 3SE +/- 235.28, N = 3SE +/- 218.93, N = 325141.4324549.3224347.011. (CXX) g++ options: -O3 -std=c++11 -fopenmp

Java Gradle Build

This test runs Java software project builds using the Gradle build system. It is intended to give developers an idea as to the build performance for development activities and build servers. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterJava Gradle BuildGradle Build: ReactorLinux 5.4Linux 5.8Linux 5.94080120160200SE +/- 2.14, N = 12SE +/- 2.36, N = 12SE +/- 2.86, N = 12199.23193.06197.76

Mlpack Benchmark

Mlpack benchmark scripts for machine learning libraries Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_qdaLinux 5.4Linux 5.8Linux 5.91632486480SE +/- 0.33, N = 3SE +/- 0.80, N = 3SE +/- 0.71, N = 1269.9571.9472.05

Rodinia

Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes select OpenCL, NVIDIA CUDA and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP LeukocyteLinux 5.4Linux 5.8Linux 5.950100150200250SE +/- 0.27, N = 3SE +/- 0.13, N = 3SE +/- 0.83, N = 3201.50207.37206.121. (CXX) g++ options: -O2 -lOpenCL

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: JythonLinux 5.4Linux 5.8Linux 5.98001600240032004000SE +/- 45.57, N = 5SE +/- 28.79, N = 20SE +/- 40.45, N = 20353235213622

Mobile Neural Network

MNN is the Mobile Neural Network as a highly efficient, lightweight deep learning framework developed by ALibaba. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2020-09-17Model: mobilenet-v1-1.0Linux 5.4Linux 5.8Linux 5.9246810SE +/- 0.024, N = 3SE +/- 0.023, N = 3SE +/- 0.008, N = 35.9996.1566.161MIN: 5.91 / MAX: 9.9MIN: 6.04 / MAX: 23.01MIN: 6.1 / MAX: 10.411. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

7-Zip Compression

This is a test of 7-Zip using p7zip with its integrated benchmark feature or upstream 7-Zip for the Windows x64 build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 16.02Compress Speed TestLinux 5.4Linux 5.8Linux 5.96K12K18K24K30KSE +/- 338.75, N = 3SE +/- 332.69, N = 3SE +/- 45.63, N = 32899728756282641. (CXX) g++ options: -pipe -lpthread

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: TradesoapLinux 5.4Linux 5.8Linux 5.912002400360048006000SE +/- 74.35, N = 4SE +/- 44.65, N = 4563457165780

GLmark2

This is a test of Linaro's glmark2 port, currently using the X11 OpenGL 2.0 target. GLmark2 is a basic OpenGL benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterGLmark2 2020.04Resolution: 3840 x 2160Linux 5.4Linux 5.8Linux 5.94080120160200157161161

LuxCoreRender

LuxCoreRender is an open-source physically based renderer. This test profile is focused on running LuxCoreRender on the CPU as opposed to the OpenCL version. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.3Scene: DLSCLinux 5.4Linux 5.8Linux 5.90.19350.3870.58050.7740.9675SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.840.860.86MIN: 0.81 / MAX: 0.86MIN: 0.82 / MAX: 0.87MIN: 0.83 / MAX: 0.87

MPV

MPV is an open-source, cross-platform media player. This test profile tests the frame-rate that can be achieved unsynchronized in a desynchronized mode. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterMPVVideo Input: Big Buck Bunny Sunflower 1080p - Decode: Software OnlyLinux 5.4Linux 5.8Linux 5.9306090120150SE +/- 0.10, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3150.37153.94153.91MIN: 112.18 / MAX: 157.9MIN: 129 / MAX: 157.9MIN: 123.87 / MAX: 157.91. mpv 0.32.0

Mobile Neural Network

MNN is the Mobile Neural Network as a highly efficient, lightweight deep learning framework developed by ALibaba. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2020-09-17Model: MobileNetV2_224Linux 5.4Linux 5.8Linux 5.91.00222.00443.00664.00885.011SE +/- 0.008, N = 3SE +/- 0.021, N = 3SE +/- 0.002, N = 34.3514.4374.454MIN: 4.27 / MAX: 20.32MIN: 4.37 / MAX: 21.73MIN: 4.42 / MAX: 8.171. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.948121620SE +/- 0.04, N = 3SE +/- 0.07, N = 3SE +/- 0.07, N = 316.3916.1916.02MIN: 15.71MIN: 15.69MIN: 15.621. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

LuxCoreRender

LuxCoreRender is an open-source physically based renderer. This test profile is focused on running LuxCoreRender on the CPU as opposed to the OpenCL version. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.3Scene: Rainbow Colors and PrismLinux 5.4Linux 5.8Linux 5.90.21150.4230.63450.8461.0575SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 30.920.930.94MIN: 0.89 / MAX: 0.97MIN: 0.91 / MAX: 0.96MIN: 0.91 / MAX: 1.02

NAMD

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

OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.14ATPase Simulation - 327,506 AtomsLinux 5.4Linux 5.8Linux 5.90.71261.42522.13782.85043.563SE +/- 0.01880, N = 3SE +/- 0.02241, N = 3SE +/- 0.00339, N = 33.167023.139063.10076

Mlpack Benchmark

Mlpack benchmark scripts for machine learning libraries Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_icaLinux 5.4Linux 5.8Linux 5.91326395265SE +/- 0.38, N = 3SE +/- 0.35, N = 3SE +/- 0.69, N = 456.5455.3656.06

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: mnasnetLinux 5.4Linux 5.8Linux 5.91.2152.433.6454.866.075SE +/- 0.01, N = 3SE +/- 0.10, N = 3SE +/- 0.02, N = 35.305.405.29MIN: 5.25 / MAX: 7.75MIN: 5.25 / MAX: 40.64MIN: 5.24 / MAX: 6.681. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

LeelaChessZero

LeelaChessZero (lc0 / lczero) is a chess engine automated vian neural networks. This test profile can be used for OpenCL, CUDA + cuDNN, and BLAS (CPU-based) benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.26Backend: RandomLinux 5.4Linux 5.8Linux 5.960K120K180K240K300KSE +/- 349.60, N = 3SE +/- 136.67, N = 3SE +/- 63.07, N = 32563472613232616261. (CXX) g++ options: -flto -pthread

Renaissance

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

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Scala DottyLinux 5.4Linux 5.8Linux 5.930060090012001500SE +/- 13.80, N = 5SE +/- 8.98, N = 5SE +/- 15.02, N = 51525.661521.471552.34

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.90.94671.89342.84013.78684.7335SE +/- 0.00371, N = 3SE +/- 0.01635, N = 3SE +/- 0.01366, N = 34.128004.207414.14533MIN: 4.07MIN: 4.13MIN: 4.081. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

libavif avifenc

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

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.7.3Encoder Speed: 8Linux 5.4Linux 5.8Linux 5.9246810SE +/- 0.019, N = 3SE +/- 0.022, N = 3SE +/- 0.012, N = 36.7606.8526.8901. (CXX) g++ options: -O3 -fPIC

Hugin

Hugin is an open-source, cross-platform panorama photo stitcher software package. This test profile times how long it takes to run the assistant and panorama photo stitching on a set of images. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterHuginPanorama Photo Assistant + Stitching TimeLinux 5.4Linux 5.8Linux 5.91428425670SE +/- 0.17, N = 3SE +/- 0.46, N = 3SE +/- 0.11, N = 361.7062.5462.88

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: goLinux 5.4Linux 5.8Linux 5.950100150200250214210210

OpenCV

This is a benchmark of the OpenCV (Computer Vision) library's built-in performance tests. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenCV 4.4Test: Object DetectionLinux 5.4Linux 5.8Linux 5.98K16K24K32K40KSE +/- 539.05, N = 15SE +/- 514.75, N = 4SE +/- 481.89, N = 43610435725354591. (CXX) g++ options: -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -ldl -lm -lpthread -lrt

Mlpack Benchmark

Mlpack benchmark scripts for machine learning libraries Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_linearridgeregressionLinux 5.4Linux 5.8Linux 5.90.7831.5662.3493.1323.915SE +/- 0.03, N = 10SE +/- 0.04, N = 3SE +/- 0.05, N = 33.483.443.42

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: crypto_pyaesLinux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.03, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 394.092.592.4

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.9510152025SE +/- 0.07, N = 3SE +/- 0.06, N = 3SE +/- 0.06, N = 321.7221.4821.37MIN: 21.5MIN: 21.34MIN: 21.221. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Deconvolution Batch deconv_1d - Data Type: f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.9246810SE +/- 0.00290, N = 3SE +/- 0.01348, N = 3SE +/- 0.00733, N = 37.438207.473817.55572MIN: 7.4MIN: 7.41MIN: 7.51. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

libavif avifenc

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

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.7.3Encoder Speed: 10Linux 5.4Linux 5.8Linux 5.9246810SE +/- 0.014, N = 3SE +/- 0.010, N = 3SE +/- 0.013, N = 36.2946.3626.3921. (CXX) g++ options: -O3 -fPIC

Sunflow Rendering System

This test runs benchmarks of the Sunflow Rendering System. The Sunflow Rendering System is an open-source render engine for photo-realistic image synthesis with a ray-tracing core. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSunflow Rendering System 0.07.2Global Illumination + Image SynthesisLinux 5.4Linux 5.8Linux 5.90.44510.89021.33531.78042.2255SE +/- 0.003, N = 3SE +/- 0.018, N = 3SE +/- 0.010, N = 31.9481.9781.956MIN: 1.86 / MAX: 2.45MIN: 1.86 / MAX: 2.52MIN: 1.84 / MAX: 2.46

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: IP Batch 1D - Data Type: u8s8f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.90.68631.37262.05892.74523.4315SE +/- 0.00383, N = 3SE +/- 0.00143, N = 3SE +/- 0.00964, N = 33.005273.050353.01876MIN: 2.93MIN: 2.97MIN: 2.931. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Memory CopyingLinux 5.4Linux 5.8Linux 5.930060090012001500SE +/- 7.33, N = 3SE +/- 9.55, N = 3SE +/- 1.27, N = 31367.011370.241351.141. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

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 2021.1Model: Face Detection 0106 FP16 - Device: CPULinux 5.4Linux 5.8Linux 5.96001200180024003000SE +/- 12.61, N = 3SE +/- 5.30, N = 3SE +/- 6.91, N = 32694.912657.932663.59

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: IP Batch All - Data Type: u8s8f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.9918273645SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.10, N = 340.5641.0141.12MIN: 40.09MIN: 40.6MIN: 40.471. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

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 2021.1Model: Face Detection 0106 FP32 - Device: CPULinux 5.4Linux 5.8Linux 5.90.3330.6660.9991.3321.665SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 31.461.471.48

libavif avifenc

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

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.7.3Encoder Speed: 0Linux 5.4Linux 5.8Linux 5.9306090120150SE +/- 0.24, N = 3SE +/- 0.33, N = 3SE +/- 1.40, N = 3147.09148.53149.101. (CXX) g++ options: -O3 -fPIC

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: regex_compileLinux 5.4Linux 5.8Linux 5.9306090120150151149150

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.9110220330440550SE +/- 0.50, N = 3SE +/- 2.50, N = 3SE +/- 5.06, N = 3482.65485.39479.08MIN: 480.58MIN: 480.32MIN: 469.421. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.970140210280350SE +/- 1.86, N = 3SE +/- 0.56, N = 3SE +/- 2.37, N = 3297.06297.45300.91MIN: 294.1MIN: 295.8MIN: 295.131. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Glibc C String FunctionsLinux 5.4Linux 5.8Linux 5.9130K260K390K520K650KSE +/- 312.30, N = 3SE +/- 77.03, N = 3SE +/- 2693.27, N = 3605374.93612988.30609670.521. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: django_templateLinux 5.4Linux 5.8Linux 5.9918273645SE +/- 0.06, N = 3SE +/- 0.12, N = 3SE +/- 0.03, N = 340.640.140.4

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.91.29892.59783.89675.19566.4945SE +/- 0.06972, N = 3SE +/- 0.00567, N = 3SE +/- 0.00520, N = 35.772765.706895.70363MIN: 5.62MIN: 5.61MIN: 5.621. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: pathlibLinux 5.4Linux 5.8Linux 5.948121620SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.00, N = 317.717.817.6

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 2021.1Model: Person Detection 0106 FP32 - Device: CPULinux 5.4Linux 5.8Linux 5.90.20250.4050.60750.811.0125SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 30.890.890.90

OpenCV

This is a benchmark of the OpenCV (Computer Vision) library's built-in performance tests. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenCV 4.4Test: Features 2DLinux 5.4Linux 5.8Linux 5.930K60K90K120K150KSE +/- 1472.64, N = 6SE +/- 415.82, N = 3SE +/- 1357.42, N = 31322081307511313761. (CXX) g++ options: -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -ldl -lm -lpthread -lrt

OpenVINO

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

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Person Detection 0106 FP16 - Device: CPULinux 5.4Linux 5.8Linux 5.90.20480.40960.61440.81921.024SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 30.910.910.90

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: CPU CacheLinux 5.4Linux 5.8Linux 5.9510152025SE +/- 0.05, N = 3SE +/- 0.16, N = 3SE +/- 0.09, N = 322.6422.5522.801. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

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 2021.1Model: Person Detection 0106 FP32 - Device: CPULinux 5.4Linux 5.8Linux 5.99001800270036004500SE +/- 10.14, N = 3SE +/- 20.80, N = 3SE +/- 1.73, N = 34399.204379.114354.59

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: IP Batch 1D - Data Type: f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.9246810SE +/- 0.00427, N = 3SE +/- 0.02516, N = 3SE +/- 0.03872, N = 36.543846.541026.60699MIN: 6.29MIN: 6.33MIN: 6.41. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

dav1d

Dav1d is an open-source, speedy AV1 video decoder. This test profile times how long it takes to decode sample AV1 video content. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Summer Nature 4KLinux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.18, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 3104.65103.76103.63MIN: 98.09 / MAX: 118.79MIN: 97.22 / MAX: 117.82MIN: 97.01 / MAX: 117.671. (CC) gcc options: -pthread

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: json_loadsLinux 5.4Linux 5.8Linux 5.9510152025SE +/- 0.00, N = 3SE +/- 0.06, N = 3SE +/- 0.00, N = 322.322.122.1

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Deconvolution Batch deconv_1d - Data Type: u8s8f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.9246810SE +/- 0.02099, N = 3SE +/- 0.02286, N = 3SE +/- 0.07536, N = 38.181478.133438.20617MIN: 8.06MIN: 8.03MIN: 8.061. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

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 2021.1Model: Face Detection 0106 FP32 - Device: CPULinux 5.4Linux 5.8Linux 5.96001200180024003000SE +/- 6.28, N = 3SE +/- 7.21, N = 3SE +/- 3.35, N = 32734.332710.492716.75

AI Benchmark Alpha

AI Benchmark Alpha is a Python library for evaluating artificial intelligence (AI) performance on diverse hardware platforms and relies upon the TensorFlow machine learning library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterAI Benchmark Alpha 0.1.2Device Inference ScoreLinux 5.4Linux 5.8Linux 5.92004006008001000838831832

Rodinia

Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes select OpenCL, NVIDIA CUDA and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP LavaMDLinux 5.4Linux 5.8Linux 5.9100200300400500SE +/- 1.14, N = 3SE +/- 1.37, N = 3SE +/- 1.07, N = 3457.36453.97453.551. (CXX) g++ options: -O2 -lOpenCL

MPV

MPV is an open-source, cross-platform media player. This test profile tests the frame-rate that can be achieved unsynchronized in a desynchronized mode. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterMPVVideo Input: Big Buck Bunny Sunflower 4K - Decode: Software OnlyLinux 5.4Linux 5.8Linux 5.9918273645SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 337.1937.5037.48MIN: 36.15 / MAX: 37.98MIN: 36.81 / MAX: 38.22MIN: 36.59 / MAX: 38.221. mpv 0.32.0

Build2

This test profile measures the time to bootstrap/install the build2 C++ build toolchain from source. Build2 is a cross-platform build toolchain for C/C++ code and features Cargo-like features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBuild2 0.12Time To CompileLinux 5.4Linux 5.8Linux 5.94080120160200SE +/- 0.23, N = 3SE +/- 0.82, N = 3SE +/- 0.61, N = 3193.41195.01194.41

Incompact3D

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

OpenBenchmarking.orgSeconds, Fewer Is BetterIncompact3D 2020-09-17Input: CylinderLinux 5.4Linux 5.8Linux 5.9130260390520650SE +/- 0.53, N = 3SE +/- 0.21, N = 3SE +/- 1.18, N = 3600.64605.08603.421. (F9X) gfortran options: -cpp -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

System GZIP Decompression

This simple test measures the time to decompress a gzipped tarball (the Qt5 toolkit source package). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSystem GZIP DecompressionLinux 5.4Linux 5.8Linux 5.90.65031.30061.95092.60123.2515SE +/- 0.040, N = 14SE +/- 0.045, N = 14SE +/- 0.034, N = 142.8902.8802.869

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 2021.1Model: Person Detection 0106 FP16 - Device: CPULinux 5.4Linux 5.8Linux 5.99001800270036004500SE +/- 6.10, N = 3SE +/- 11.71, N = 3SE +/- 4.29, N = 34345.924332.324314.69

XZ Compression

This test measures the time needed to compress a sample file (an Ubuntu file-system image) using XZ compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterXZ Compression 5.2.4Compressing ubuntu-16.04.3-server-i386.img, Compression Level 9Linux 5.4Linux 5.8Linux 5.9918273645SE +/- 0.25, N = 3SE +/- 0.08, N = 3SE +/- 0.08, N = 340.1940.0039.911. (CC) gcc options: -pthread -fvisibility=hidden -O2

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Deconvolution Batch deconv_3d - Data Type: u8s8f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.91.34192.68384.02575.36766.7095SE +/- 0.01148, N = 3SE +/- 0.00504, N = 3SE +/- 0.01836, N = 35.922715.963935.95687MIN: 5.88MIN: 5.91MIN: 5.871. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: squeezenetLinux 5.4Linux 5.8Linux 5.9510152025SE +/- 0.06, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 320.5220.3820.46MIN: 20.33 / MAX: 23.11MIN: 20.3 / MAX: 21.63MIN: 20.3 / MAX: 29.271. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenVINO

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

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Face Detection 0106 FP16 - Device: CPULinux 5.4Linux 5.8Linux 5.90.33750.6751.01251.351.6875SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 31.491.501.49

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: alexnetLinux 5.4Linux 5.8Linux 5.9510152025SE +/- 0.18, N = 3SE +/- 0.04, N = 3SE +/- 0.18, N = 319.9819.8619.99MIN: 19.74 / MAX: 20.46MIN: 19.72 / MAX: 29.42MIN: 19.74 / MAX: 20.771. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

dav1d

Dav1d is an open-source, speedy AV1 video decoder. This test profile times how long it takes to decode sample AV1 video content. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Chimera 1080p 10-bitLinux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 383.4982.9683.01MIN: 56.28 / MAX: 184.52MIN: 56.02 / MAX: 182.08MIN: 56.07 / MAX: 181.911. (CC) gcc options: -pthread

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Glibc Qsort Data SortingLinux 5.4Linux 5.8Linux 5.91530456075SE +/- 0.20, N = 3SE +/- 0.42, N = 3SE +/- 0.22, N = 368.2367.9067.801. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 5.4Time To CompileLinux 5.4Linux 5.8Linux 5.94080120160200SE +/- 0.55, N = 3SE +/- 0.73, N = 3SE +/- 0.83, N = 3159.96159.02160.02

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Matrix MathLinux 5.4Linux 5.8Linux 5.95K10K15K20K25KSE +/- 327.37, N = 3SE +/- 82.76, N = 3SE +/- 311.08, N = 323746.7323793.6023890.641. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: blazefaceLinux 5.4Linux 5.8Linux 5.90.38250.7651.14751.531.9125SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 31.701.691.70MIN: 1.67 / MAX: 1.78MIN: 1.67 / MAX: 1.75MIN: 1.68 / MAX: 1.791. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Timed MAFFT Alignment

This test performs an alignment of 100 pyruvate decarboxylase sequences. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed MAFFT Alignment 7.471Multiple Sequence Alignment - LSU RNALinux 5.4Linux 5.8Linux 5.93691215SE +/- 0.10, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 310.3810.3210.321. (CC) gcc options: -std=c99 -O3 -lm -lpthread

ASTC Encoder

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

OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.0Preset: ExhaustiveLinux 5.4Linux 5.8Linux 5.9100200300400500SE +/- 0.07, N = 3SE +/- 0.09, N = 3SE +/- 0.11, N = 3464.58461.89461.931. (CXX) g++ options: -std=c++14 -fvisibility=hidden -O3 -flto -mfpmath=sse -mavx2 -mpopcnt -lpthread

AI Benchmark Alpha

AI Benchmark Alpha is a Python library for evaluating artificial intelligence (AI) performance on diverse hardware platforms and relies upon the TensorFlow machine learning library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterAI Benchmark Alpha 0.1.2Device AI ScoreLinux 5.4Linux 5.8Linux 5.9400800120016002000174717381737

Timed Apache Compilation

This test times how long it takes to build the Apache HTTPD web server. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Apache Compilation 2.4.41Time To CompileLinux 5.4Linux 5.8Linux 5.9612182430SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 326.6726.5226.54

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: CPU StressLinux 5.4Linux 5.8Linux 5.95001000150020002500SE +/- 12.25, N = 3SE +/- 22.94, N = 3SE +/- 17.12, N = 32253.832264.722251.991. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: mobilenetLinux 5.4Linux 5.8Linux 5.9612182430SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 323.7523.6223.62MIN: 23.6 / MAX: 33.68MIN: 23.54 / MAX: 25.8MIN: 23.49 / MAX: 25.451. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: shufflenet-v2Linux 5.4Linux 5.8Linux 5.90.83251.6652.49753.334.1625SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 33.683.693.70MIN: 3.64 / MAX: 5.72MIN: 3.63 / MAX: 5.12MIN: 3.63 / MAX: 23.61. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

GPAW

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

OpenBenchmarking.orgSeconds, Fewer Is BetterGPAW 20.1Input: Carbon NanotubeLinux 5.4Linux 5.8Linux 5.9110220330440550SE +/- 0.10, N = 3SE +/- 1.07, N = 3SE +/- 0.17, N = 3484.98487.42485.031. (CC) gcc options: -pthread -shared -fwrapv -O2 -lxc -lblas -lmpi

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: raytraceLinux 5.4Linux 5.8Linux 5.990180270360450SE +/- 0.33, N = 3SE +/- 0.67, N = 3400399401

Renaissance

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

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Akka Unbalanced Cobwebbed TreeLinux 5.4Linux 5.8Linux 5.916003200480064008000SE +/- 102.37, N = 5SE +/- 52.14, N = 5SE +/- 86.44, N = 57673.097656.807634.87

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: IP Batch All - Data Type: f32 - Engine: CPULinux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.04, N = 3SE +/- 0.24, N = 3SE +/- 0.22, N = 398.3798.0798.55MIN: 96.72MIN: 96.88MIN: 96.81. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Darmstadt Automotive Parallel Heterogeneous Suite

DAPHNE is the Darmstadt Automotive Parallel HeterogeNEous Benchmark Suite with OpenCL / CUDA / OpenMP test cases for these automotive benchmarks for evaluating programming models in context to vehicle autonomous driving capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTest Cases Per Minute, More Is BetterDarmstadt Automotive Parallel Heterogeneous SuiteBackend: OpenMP - Kernel: Euclidean ClusterLinux 5.4Linux 5.8Linux 5.92004006008001000SE +/- 0.19, N = 3SE +/- 0.19, N = 3SE +/- 0.32, N = 3990.87987.26986.381. (CXX) g++ options: -O3 -std=c++11 -fopenmp

AI Benchmark Alpha

AI Benchmark Alpha is a Python library for evaluating artificial intelligence (AI) performance on diverse hardware platforms and relies upon the TensorFlow machine learning library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterAI Benchmark Alpha 0.1.2Device Training ScoreLinux 5.4Linux 5.8Linux 5.92004006008001000909907905

libavif avifenc

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

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.7.3Encoder Speed: 2Linux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.10, N = 3SE +/- 0.23, N = 3SE +/- 0.24, N = 387.8887.5287.901. (CXX) g++ options: -O3 -fPIC

dav1d

Dav1d is an open-source, speedy AV1 video decoder. This test profile times how long it takes to decode sample AV1 video content. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Summer Nature 1080pLinux 5.4Linux 5.8Linux 5.980160240320400SE +/- 0.88, N = 3SE +/- 0.71, N = 3SE +/- 0.04, N = 3360.91361.03359.55MIN: 320.03 / MAX: 393.23MIN: 332.7 / MAX: 396.62MIN: 326.04 / MAX: 3931. (CC) gcc options: -pthread

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: yolov4-tinyLinux 5.4Linux 5.8Linux 5.9816243240SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 332.6632.5332.59MIN: 32.51 / MAX: 34.05MIN: 32.45 / MAX: 35.73MIN: 32.47 / MAX: 34.941. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

ASTC Encoder

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

OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.0Preset: FastLinux 5.4Linux 5.8Linux 5.9246810SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 37.577.597.601. (CXX) g++ options: -std=c++14 -fvisibility=hidden -O3 -flto -mfpmath=sse -mavx2 -mpopcnt -lpthread

OpenVINO

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

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Age Gender Recognition Retail 0013 FP32 - Device: CPULinux 5.4Linux 5.8Linux 5.98001600240032004000SE +/- 8.81, N = 3SE +/- 10.34, N = 3SE +/- 2.42, N = 33700.083697.193711.83

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: python_startupLinux 5.4Linux 5.8Linux 5.9246810SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 37.757.777.78

dav1d

Dav1d is an open-source, speedy AV1 video decoder. This test profile times how long it takes to decode sample AV1 video content. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Chimera 1080pLinux 5.4Linux 5.8Linux 5.980160240320400SE +/- 0.68, N = 3SE +/- 0.78, N = 3SE +/- 0.51, N = 3389.16389.39390.64MIN: 283.33 / MAX: 628.16MIN: 284.04 / MAX: 622.86MIN: 284.57 / MAX: 621.431. (CC) gcc options: -pthread

BYTE Unix Benchmark

This is a test of BYTE. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgLPS, More Is BetterBYTE Unix Benchmark 3.6Computational Test: Dhrystone 2Linux 5.4Linux 5.8Linux 5.910M20M30M40M50MSE +/- 148494.90, N = 3SE +/- 52622.62, N = 3SE +/- 129412.59, N = 344618964.444673416.744505024.9

BRL-CAD

BRL-CAD 7.28.0 is a cross-platform, open-source solid modeling system with built-in benchmark mode. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgVGR Performance Metric, More Is BetterBRL-CAD 7.30.8VGR Performance MetricLinux 5.4Linux 5.8Linux 5.913K26K39K52K65K6183661606616631. (CXX) g++ options: -std=c++11 -pipe -fno-strict-aliasing -fno-common -fexceptions -ftemplate-depth-128 -m64 -ggdb3 -O3 -fipa-pta -fstrength-reduce -finline-functions -flto -pedantic -rdynamic -lSM -lICE -lGLU -lGL -lGLdispatch -lX11 -lXext -lXrender -lpthread -ldl -luuid -lm

Hierarchical INTegration

This test runs the U.S. Department of Energy's Ames Laboratory Hierarchical INTegration (HINT) benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQUIPs, More Is BetterHierarchical INTegration 1.0Test: FLOATLinux 5.4Linux 5.8Linux 5.9100M200M300M400M500MSE +/- 548012.24, N = 3SE +/- 531332.52, N = 3SE +/- 1176140.02, N = 3445325469.78444887120.59443682873.661. (CC) gcc options: -O3 -march=native -lm

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: 2to3Linux 5.4Linux 5.8Linux 5.960120180240300279278278

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: MallocLinux 5.4Linux 5.8Linux 5.97M14M21M28M35MSE +/- 18293.92, N = 3SE +/- 53114.71, N = 3SE +/- 37067.76, N = 330826272.7030773901.0030880073.581. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

WebP Image Encode

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

OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, LosslessLinux 5.4Linux 5.8Linux 5.948121620SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 316.7416.6816.741. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: chaosLinux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.06, N = 3SE +/- 0.06, N = 3SE +/- 0.12, N = 390.289.990.2

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU-v2-v2 - Model: mobilenet-v2Linux 5.4Linux 5.8Linux 5.9246810SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 36.346.366.36MIN: 6.26 / MAX: 7.66MIN: 6.29 / MAX: 7.84MIN: 6.27 / MAX: 8.481. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

InfluxDB

This is a benchmark of the InfluxDB open-source time-series database optimized for fast, high-availability storage for IoT and other use-cases. The InfluxDB test profile makes use of InfluxDB Inch for facilitating the benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgval/sec, More Is BetterInfluxDB 1.8.2Concurrent Streams: 1024 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000Linux 5.4Linux 5.8Linux 5.9300K600K900K1200K1500KSE +/- 1080.06, N = 3SE +/- 1296.96, N = 3SE +/- 1603.00, N = 31250226.31254110.11252051.7

Tesseract OCR

Tesseract-OCR is the open-source optical character recognition (OCR) engine for the conversion of text within images to raw text output. This test profile relies upon a system-supplied Tesseract installation. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTesseract OCR 4.1.1Time To OCR 7 ImagesLinux 5.4Linux 5.8Linux 5.9510152025SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 322.6222.6922.66

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: floatLinux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.13, N = 3SE +/- 0.12, N = 3SE +/- 0.03, N = 397.197.297.4

Montage Astronomical Image Mosaic Engine

Montage is an open-source astronomical image mosaic engine. This BSD-licensed astronomy software is developed by the California Institute of Technology, Pasadena. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMontage Astronomical Image Mosaic Engine 6.0Mosaic of M17, K band, 1.5 deg x 1.5 degLinux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.20, N = 3SE +/- 0.10, N = 3SE +/- 0.13, N = 376.7876.9177.021. (CC) gcc options: -std=gnu99 -lcfitsio -lm -O2

LibRaw

LibRaw is a RAW image decoder for digital camera photos. This test profile runs LibRaw's post-processing benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMpix/sec, More Is BetterLibRaw 0.20Post-Processing BenchmarkLinux 5.4Linux 5.8Linux 5.9816243240SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 333.6233.5233.571. (CXX) g++ options: -O2 -fopenmp -ljpeg -lz -lm

InfluxDB

This is a benchmark of the InfluxDB open-source time-series database optimized for fast, high-availability storage for IoT and other use-cases. The InfluxDB test profile makes use of InfluxDB Inch for facilitating the benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgval/sec, More Is BetterInfluxDB 1.8.2Concurrent Streams: 64 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000Linux 5.4Linux 5.8Linux 5.9300K600K900K1200K1500KSE +/- 588.19, N = 3SE +/- 4430.10, N = 3SE +/- 1100.56, N = 31239227.01239154.61242649.2

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: resnet18Linux 5.4Linux 5.8Linux 5.948121620SE +/- 0.01, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 318.1418.1518.10MIN: 18.03 / MAX: 20.15MIN: 17.99 / MAX: 20.22MIN: 17.98 / MAX: 20.491. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OCRMyPDF

OCRMyPDF is an optical character recognition (OCR) text layer to scanned PDF files, producing new PDFs with the text now selectable/searchable/copy-paste capable. OCRMyPDF leverages the Tesseract OCR engine and is written in Python. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOCRMyPDF 9.6.0+dfsgProcessing 60 Page PDF DocumentLinux 5.4Linux 5.8Linux 5.91020304050SE +/- 0.22, N = 3SE +/- 0.21, N = 3SE +/- 0.35, N = 344.7244.7844.84

ASTC Encoder

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

OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.0Preset: ThoroughLinux 5.4Linux 5.8Linux 5.91326395265SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.06, N = 357.9257.7757.841. (CXX) g++ options: -std=c++14 -fvisibility=hidden -O3 -flto -mfpmath=sse -mavx2 -mpopcnt -lpthread

Dolfyn

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

OpenBenchmarking.orgSeconds, Fewer Is BetterDolfyn 0.527Computational Fluid DynamicsLinux 5.4Linux 5.8Linux 5.948121620SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 318.1518.1518.19

Rodinia

Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes select OpenCL, NVIDIA CUDA and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP CFD SolverLinux 5.4Linux 5.8Linux 5.9918273645SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 338.7738.6838.721. (CXX) g++ options: -O2 -lOpenCL

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: efficientnet-b0Linux 5.4Linux 5.8Linux 5.9246810SE +/- 0.06, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 38.588.568.56MIN: 8.49 / MAX: 21.34MIN: 8.5 / MAX: 9.18MIN: 8.49 / MAX: 9.171. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

ASTC Encoder

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

OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.0Preset: MediumLinux 5.4Linux 5.8Linux 5.93691215SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 38.978.958.961. (CXX) g++ options: -std=c++14 -fvisibility=hidden -O3 -flto -mfpmath=sse -mavx2 -mpopcnt -lpthread

Rodinia

Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes select OpenCL, NVIDIA CUDA and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP HotSpot3DLinux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.05, N = 3SE +/- 0.27, N = 3SE +/- 0.22, N = 399.79100.0199.891. (CXX) g++ options: -O2 -lOpenCL

WebP Image Encode

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

OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Lossless, Highest CompressionLinux 5.4Linux 5.8Linux 5.9918273645SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.08, N = 339.0038.9238.931. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: vgg16Linux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 378.9979.1679.16MIN: 78.84 / MAX: 88.48MIN: 78.88 / MAX: 88.73MIN: 78.91 / MAX: 88.51. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: googlenetLinux 5.4Linux 5.8Linux 5.9510152025SE +/- 0.03, N = 3SE +/- 0.07, N = 3SE +/- 0.03, N = 318.8118.8518.81MIN: 18.71 / MAX: 19.31MIN: 18.69 / MAX: 20.88MIN: 18.71 / MAX: 20.891. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Mlpack Benchmark

Mlpack benchmark scripts for machine learning libraries Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_svmLinux 5.4Linux 5.8Linux 5.9612182430SE +/- 0.09, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 323.6523.6023.62

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 2021.1Model: Age Gender Recognition Retail 0013 FP16 - Device: CPULinux 5.4Linux 5.8Linux 5.98001600240032004000SE +/- 8.11, N = 3SE +/- 7.38, N = 3SE +/- 4.20, N = 33720.683713.403712.89

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Vector MathLinux 5.4Linux 5.8Linux 5.97K14K21K28K35KSE +/- 0.45, N = 3SE +/- 1.65, N = 3SE +/- 73.83, N = 334720.0834720.2934648.211. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

TNN

TNN is an open-source deep learning reasoning framework developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterTNN 0.2.3Target: CPU - Model: MobileNet v2Linux 5.4Linux 5.8Linux 5.970140210280350SE +/- 0.37, N = 3SE +/- 1.05, N = 3SE +/- 0.41, N = 3308.55308.57307.95MIN: 307.18 / MAX: 309.84MIN: 306.49 / MAX: 311.86MIN: 306.42 / MAX: 312.631. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -O3 -rdynamic -ldl

Rodinia

Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes select OpenCL, NVIDIA CUDA and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP StreamclusterLinux 5.4Linux 5.8Linux 5.9510152025SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 321.3321.2921.291. (CXX) g++ options: -O2 -lOpenCL

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: resnet50Linux 5.4Linux 5.8Linux 5.9918273645SE +/- 0.06, N = 3SE +/- 0.02, N = 3SE +/- 0.07, N = 337.5237.4937.55MIN: 37.36 / MAX: 62.22MIN: 37.36 / MAX: 40.61MIN: 37.36 / MAX: 52.241. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Timed HMMer Search

This test searches through the Pfam database of profile hidden markov models. The search finds the domain structure of Drosophila Sevenless protein. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed HMMer Search 3.3.1Pfam Database SearchLinux 5.4Linux 5.8Linux 5.920406080100SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3102.55102.46102.401. (CC) gcc options: -O3 -pthread -lhmmer -leasel -lm

Chaos Group V-RAY

This is a test of Chaos Group's V-RAY benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgKsamples, More Is BetterChaos Group V-RAY 4.10.07Mode: CPULinux 5.4Linux 5.8Linux 5.917003400510068008500SE +/- 14.44, N = 3SE +/- 5.67, N = 3SE +/- 25.64, N = 3771077087718

WebP Image Encode

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

OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100Linux 5.4Linux 5.8Linux 5.90.52851.0571.58552.1142.6425SE +/- 0.000, N = 3SE +/- 0.001, N = 3SE +/- 0.003, N = 32.3482.3462.3491. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff

Zstd Compression

This test measures the time needed to compress a sample file (an Ubuntu ISO) using Zstd compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.5Compression Level: 3Linux 5.4Linux 5.8Linux 5.96001200180024003000SE +/- 7.37, N = 3SE +/- 8.43, N = 3SE +/- 16.61, N = 32710.12708.92706.71. (CC) gcc options: -O3 -pthread -lz -llzma

TensorFlow Lite

This is a benchmark of the TensorFlow Lite implementation. The current Linux support is limited to running on CPUs. This test profile is measuring the average inference time. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: SqueezeNetLinux 5.4Linux 5.8Linux 5.990K180K270K360K450KSE +/- 163.48, N = 3SE +/- 21.36, N = 3SE +/- 34.51, N = 3399233398809398749

LAMMPS Molecular Dynamics Simulator

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

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 24Aug2020Model: Rhodopsin ProteinLinux 5.4Linux 5.8Linux 5.90.77421.54842.32263.09683.871SE +/- 0.009, N = 3SE +/- 0.010, N = 3SE +/- 0.005, N = 33.4413.4383.4371. (CXX) g++ options: -O3 -pthread -lm

WebP Image Encode

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

OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Highest CompressionLinux 5.4Linux 5.8Linux 5.9246810SE +/- 0.007, N = 3SE +/- 0.003, N = 3SE +/- 0.002, N = 37.0157.0177.0091. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: CryptoLinux 5.4Linux 5.8Linux 5.92004006008001000SE +/- 1.00, N = 3SE +/- 0.71, N = 3SE +/- 0.24, N = 3978.39978.23979.171. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

WebP Image Encode

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

OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: DefaultLinux 5.4Linux 5.8Linux 5.90.33730.67461.01191.34921.6865SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.001, N = 31.4991.4981.4981. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff

TensorFlow Lite

This is a benchmark of the TensorFlow Lite implementation. The current Linux support is limited to running on CPUs. This test profile is measuring the average inference time. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet QuantLinux 5.4Linux 5.8Linux 5.960K120K180K240K300KSE +/- 17.75, N = 3SE +/- 79.65, N = 3SE +/- 68.79, N = 3278472278501278615

RNNoise

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

OpenBenchmarking.orgSeconds, Fewer Is BetterRNNoise 2020-06-28Linux 5.4Linux 5.8Linux 5.9612182430SE +/- 0.07, N = 3SE +/- 0.07, N = 3SE +/- 0.08, N = 324.0824.0724.071. (CC) gcc options: -O2 -pedantic -fvisibility=hidden

TNN

TNN is an open-source deep learning reasoning framework developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterTNN 0.2.3Target: CPU - Model: SqueezeNet v1.1Linux 5.4Linux 5.8Linux 5.960120180240300SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 3297.61297.51297.52MIN: 297.15 / MAX: 298.48MIN: 297.11 / MAX: 297.95MIN: 297.06 / MAX: 298.411. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -O3 -rdynamic -ldl

TensorFlow Lite

This is a benchmark of the TensorFlow Lite implementation. The current Linux support is limited to running on CPUs. This test profile is measuring the average inference time. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Inception ResNet V2Linux 5.4Linux 5.8Linux 5.91.1M2.2M3.3M4.4M5.5MSE +/- 294.86, N = 3SE +/- 1330.61, N = 3SE +/- 317.51, N = 3520144352028105201097

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet FloatLinux 5.4Linux 5.8Linux 5.960K120K180K240K300KSE +/- 6.12, N = 3SE +/- 15.28, N = 3SE +/- 47.69, N = 3270813270791270872

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: NASNet MobileLinux 5.4Linux 5.8Linux 5.960K120K180K240K300KSE +/- 56.15, N = 3SE +/- 33.14, N = 3SE +/- 39.95, N = 3292827292887292836

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Inception V4Linux 5.4Linux 5.8Linux 5.91.2M2.4M3.6M4.8M6MSE +/- 1685.06, N = 3SE +/- 51.32, N = 3SE +/- 610.31, N = 3574956057489105748433

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: pickle_pure_pythonLinux 5.4Linux 5.8Linux 5.980160240320400SE +/- 0.67, N = 3SE +/- 0.58, N = 3368368368

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: nbodyLinux 5.4Linux 5.8Linux 5.920406080100109109109

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 2021.1Model: Age Gender Recognition Retail 0013 FP32 - Device: CPULinux 5.4Linux 5.8Linux 5.90.24080.48160.72240.96321.204SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 31.071.071.07

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Age Gender Recognition Retail 0013 FP16 - Device: CPULinux 5.4Linux 5.8Linux 5.90.24080.48160.72240.96321.204SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 31.071.071.07

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU-v3-v3 - Model: mobilenet-v3Linux 5.4Linux 5.8Linux 5.91.20832.41663.62494.83326.0415SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 35.375.375.37MIN: 5.31 / MAX: 6.75MIN: 5.3 / MAX: 6.88MIN: 5.29 / MAX: 6.931. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Zstd Compression

This test measures the time needed to compress a sample file (an Ubuntu ISO) using Zstd compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.5Compression Level: 19Linux 5.4Linux 5.8Linux 5.96121824302323231. (CC) gcc options: -O3 -pthread -lz -llzma

Monte Carlo Simulations of Ionised Nebulae

Mocassin is the Monte Carlo Simulations of Ionised Nebulae. MOCASSIN is a fully 3D or 2D photoionisation and dust radiative transfer code which employs a Monte Carlo approach to the transfer of radiation through media of arbitrary geometry and density distribution. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMonte Carlo Simulations of Ionised Nebulae 2019-03-24Input: Dust 2D tau100.0Linux 5.4Linux 5.8Linux 5.950100150200250SE +/- 0.33, N = 32412412411. (F9X) gfortran options: -cpp -Jsource/ -ffree-line-length-0 -lm -std=legacy -O3 -O2 -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

OpenCV

This is a benchmark of the OpenCV (Computer Vision) library's built-in performance tests. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenCV 4.4Test: DNN - Deep Neural NetworkLinux 5.4Linux 5.8Linux 5.96K12K18K24K30KSE +/- 8295.57, N = 15SE +/- 71.34, N = 3SE +/- 179.08, N = 32823217775178331. (CXX) g++ options: -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -ldl -lm -lpthread -lrt

Kripke

Kripke is a simple, scalable, 3D Sn deterministic particle transport code. Its primary purpose is to research how data layout, programming paradigms and architectures effect the implementation and performance of Sn transport. Kripke is developed by LLNL. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgThroughput FoM, More Is BetterKripke 1.2.4Linux 5.4Linux 5.8Linux 5.92M4M6M8M10MSE +/- 5698.82, N = 3SE +/- 475554.05, N = 9SE +/- 124576.41, N = 32892974672941787284631. (CXX) g++ options: -O3 -fopenmp

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: blazefaceLinux 5.4Linux 5.8Linux 5.9246810SE +/- 0.01, N = 3SE +/- 0.11, N = 3SE +/- 0.08, N = 36.192.422.25MIN: 5.97 / MAX: 6.61MIN: 2.11 / MAX: 2.83MIN: 2.06 / MAX: 2.731. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: System V Message PassingLinux 5.4Linux 5.8Linux 5.91000K2000K3000K4000K5000KSE +/- 144575.68, N = 15SE +/- 6757.38, N = 3SE +/- 6099.30, N = 34624409.894899988.404608826.621. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -lpthread -lc

Darmstadt Automotive Parallel Heterogeneous Suite

DAPHNE is the Darmstadt Automotive Parallel HeterogeNEous Benchmark Suite with OpenCL / CUDA / OpenMP test cases for these automotive benchmarks for evaluating programming models in context to vehicle autonomous driving capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTest Cases Per Minute, More Is BetterDarmstadt Automotive Parallel Heterogeneous SuiteBackend: OpenMP - Kernel: NDT MappingLinux 5.4Linux 5.8Linux 5.92004006008001000SE +/- 2.01, N = 3SE +/- 15.95, N = 15SE +/- 0.69, N = 3797.68776.82798.881. (CXX) g++ options: -O3 -std=c++11 -fopenmp

GROMACS

The GROMACS (GROningen MAchine for Chemical Simulations) molecular dynamics package testing on the CPU with the water_GMX50 data. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2020.3Water BenchmarkLinux 5.4Linux 5.8Linux 5.90.14040.28080.42120.56160.702SE +/- 0.020, N = 9SE +/- 0.001, N = 3SE +/- 0.003, N = 30.5860.6190.6241. (CXX) g++ options: -O3 -pthread -lrt -lpthread -lm

Renaissance

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

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Apache Spark PageRankLinux 5.4Linux 5.8Linux 5.99001800270036004500SE +/- 68.36, N = 21SE +/- 75.44, N = 25SE +/- 44.43, N = 254012.354095.814135.93

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Apache Spark BayesLinux 5.4Linux 5.8Linux 5.96001200180024003000SE +/- 58.62, N = 25SE +/- 55.26, N = 25SE +/- 57.81, N = 252928.272845.832863.14

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Apache Spark ALSLinux 5.4Linux 5.8Linux 5.96001200180024003000SE +/- 19.32, N = 25SE +/- 39.61, N = 25SE +/- 33.82, N = 252266.942605.072642.19

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Random ForestLinux 5.4Linux 5.8Linux 5.9400800120016002000SE +/- 20.88, N = 5SE +/- 24.52, N = 25SE +/- 18.08, N = 251611.401712.831685.93

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: H2Linux 5.4Linux 5.8Linux 5.96001200180024003000SE +/- 33.70, N = 4SE +/- 47.98, N = 20SE +/- 51.89, N = 20295529282935

RealSR-NCNN

RealSR-NCNN is an NCNN neural network implementation of the RealSR project and accelerated using the Vulkan API. RealSR is the Real-World Super Resolution via Kernel Estimation and Noise Injection. NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. This test profile times how long it takes to increase the resolution of a sample image by a scale of 4x with Vulkan. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: NoLinux 5.4Linux 5.8Linux 5.960120180240300SE +/- 2.09, N = 15SE +/- 0.06, N = 3SE +/- 0.08, N = 342.12258.72258.76

202 Results Shown

Stress-NG
NCNN:
  Vulkan GPU-v3-v3 - mobilenet-v3
  Vulkan GPU - efficientnet-b0
  Vulkan GPU - mnasnet
  Vulkan GPU-v2-v2 - mobilenet-v2
  Vulkan GPU - shufflenet-v2
  Vulkan GPU - googlenet
  Vulkan GPU - resnet18
  Vulkan GPU - mobilenet
  Vulkan GPU - squeezenet
  Vulkan GPU - alexnet
  Vulkan GPU - resnet50
  Vulkan GPU - yolov4-tiny
Stress-NG
NCNN
Stress-NG
Apache CouchDB
Renaissance
Stress-NG:
  Forking
  MMAP
  MEMFD
  Context Switching
eSpeak-NG Speech Engine
InfluxDB
WireGuard + Linux Networking Stack Stress Test
KeyDB
VkFFT
Stress-NG
FFTE
Renaissance
GLmark2:
  1920 x 1080
  1920 x 1200
Stress-NG
Caffe
Stress-NG
Mobile Neural Network
oneDNN
LeelaChessZero
Caffe
NeatBench
Caffe
GLmark2
Caffe
LeelaChessZero
Mobile Neural Network:
  inception-v3
  SqueezeNetV1.0
Darmstadt Automotive Parallel Heterogeneous Suite
Java Gradle Build
Mlpack Benchmark
Rodinia
DaCapo Benchmark
Mobile Neural Network
7-Zip Compression
DaCapo Benchmark
GLmark2
LuxCoreRender
MPV
Mobile Neural Network
oneDNN
LuxCoreRender
NAMD
Mlpack Benchmark
NCNN
LeelaChessZero
Renaissance
oneDNN
libavif avifenc
Hugin
PyPerformance
OpenCV
Mlpack Benchmark
PyPerformance
oneDNN:
  Convolution Batch Shapes Auto - f32 - CPU
  Deconvolution Batch deconv_1d - f32 - CPU
libavif avifenc
Sunflow Rendering System
oneDNN
Stress-NG
OpenVINO
oneDNN
OpenVINO
libavif avifenc
PyPerformance
oneDNN:
  Recurrent Neural Network Training - f32 - CPU
  Recurrent Neural Network Inference - f32 - CPU
Stress-NG
PyPerformance
oneDNN
PyPerformance
OpenVINO
OpenCV
OpenVINO
Stress-NG
OpenVINO
oneDNN
dav1d
PyPerformance
oneDNN
OpenVINO
AI Benchmark Alpha
Rodinia
MPV
Build2
Incompact3D
System GZIP Decompression
OpenVINO
XZ Compression
oneDNN
NCNN
OpenVINO
NCNN
dav1d
Stress-NG
Timed Linux Kernel Compilation
Stress-NG
NCNN
Timed MAFFT Alignment
ASTC Encoder
AI Benchmark Alpha
Timed Apache Compilation
Stress-NG
NCNN:
  CPU - mobilenet
  CPU - shufflenet-v2
GPAW
PyPerformance
Renaissance
oneDNN
Darmstadt Automotive Parallel Heterogeneous Suite
AI Benchmark Alpha
libavif avifenc
dav1d
NCNN
ASTC Encoder
OpenVINO
PyPerformance
dav1d
BYTE Unix Benchmark
BRL-CAD
Hierarchical INTegration
PyPerformance
Stress-NG
WebP Image Encode
PyPerformance
NCNN
InfluxDB
Tesseract OCR
PyPerformance
Montage Astronomical Image Mosaic Engine
LibRaw
InfluxDB
NCNN
OCRMyPDF
ASTC Encoder
Dolfyn
Rodinia
NCNN
ASTC Encoder
Rodinia
WebP Image Encode
NCNN:
  CPU - vgg16
  CPU - googlenet
Mlpack Benchmark
OpenVINO
Stress-NG
TNN
Rodinia
NCNN
Timed HMMer Search
Chaos Group V-RAY
WebP Image Encode
Zstd Compression
TensorFlow Lite
LAMMPS Molecular Dynamics Simulator
WebP Image Encode
Stress-NG
WebP Image Encode
TensorFlow Lite
RNNoise
TNN
TensorFlow Lite:
  Inception ResNet V2
  Mobilenet Float
  NASNet Mobile
  Inception V4
PyPerformance:
  pickle_pure_python
  nbody
OpenVINO:
  Age Gender Recognition Retail 0013 FP32 - CPU
  Age Gender Recognition Retail 0013 FP16 - CPU
NCNN
Zstd Compression
Monte Carlo Simulations of Ionised Nebulae
OpenCV
Kripke
NCNN
Stress-NG
Darmstadt Automotive Parallel Heterogeneous Suite
GROMACS
Renaissance:
  Apache Spark PageRank
  Apache Spark Bayes
  Apache Spark ALS
  Rand Forest
DaCapo Benchmark
RealSR-NCNN