7900X New

Intel Core i7-7900X testing with a ASRock X299 Extreme4 (P1.50 BIOS) and Zotac NVIDIA GeForce GT 610 1GB on Ubuntu 19.04 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 2010024-FI-7900XNEW720
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

Limit displaying results to tests within:

AV1 4 Tests
Bioinformatics 2 Tests
BLAS (Basic Linear Algebra Sub-Routine) Tests 2 Tests
C++ Boost Tests 2 Tests
Timed Code Compilation 3 Tests
C/C++ Compiler Tests 12 Tests
Compression Tests 2 Tests
CPU Massive 21 Tests
Creator Workloads 24 Tests
Database Test Suite 2 Tests
Encoding 5 Tests
Game Development 4 Tests
HPC - High Performance Computing 19 Tests
Imaging 6 Tests
Java 3 Tests
Machine Learning 9 Tests
Molecular Dynamics 3 Tests
MPI Benchmarks 2 Tests
Multi-Core 20 Tests
NVIDIA GPU Compute 6 Tests
OCR 2 Tests
Intel oneAPI 4 Tests
OpenCV Tests 2 Tests
OpenMPI Tests 5 Tests
Programmer / Developer System Benchmarks 8 Tests
Python 2 Tests
Renderers 3 Tests
Scientific Computing 9 Tests
Server 2 Tests
Server CPU Tests 13 Tests
Single-Threaded 5 Tests
Speech 2 Tests
Telephony 2 Tests
Video Encoding 5 Tests
Common Workstation Benchmarks 4 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
1
September 29 2020
  23 Hours, 35 Minutes
2
September 30 2020
  18 Hours, 13 Minutes
3
October 01 2020
  17 Hours, 8 Minutes
Invert Hiding All Results Option
  19 Hours, 39 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):

1

Compiler Notes: --disable-multilib --enable-checking=release --enable-languages=c,c++,fortran
Processor Notes: Scaling Governor: intel_pstate powersave - CPU Microcode: 0x2000064
Java Notes: OpenJDK Runtime Environment (build 11.0.5+10-post-Ubuntu-0ubuntu1.119.04)
Python Notes: Python 2.7.16 + Python 3.7.3
Security Notes: 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

2

3

Processor: Intel Core i7-7900X @ 4.50GHz (10 Cores / 20 Threads), Motherboard: ASRock X299 Extreme4 (P1.50 BIOS), Chipset: Intel Sky Lake-E DMI3 Registers, Memory: 16GB, Disk: 120GB Corsair Force MP500, Graphics: Zotac NVIDIA GeForce GT 610 1GB, Audio: Realtek ALC1220, Monitor: LG Ultra HD, Network: Intel I219-V

OS: Ubuntu 19.04, Kernel: 5.0.0-38-generic (x86_64), Desktop: GNOME Shell 3.32.1, Display Server: X Server 1.20.4, Display Driver: modesetting 1.20.4, Compiler: GCC 11.0.0 20200929, File-System: ext4, Screen Resolution: 1920x1080

7900X NewProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverCompilerFile-SystemScreen Resolution123Intel Core i7-7900X @ 4.50GHz (10 Cores / 20 Threads)ASRock X299 Extreme4 (P1.50 BIOS)Intel Sky Lake-E DMI3 Registers16GB120GB Corsair Force MP500Zotac NVIDIA GeForce GT 610 1GBRealtek ALC1220LG Ultra HDIntel I219-VUbuntu 19.045.0.0-38-generic (x86_64)GNOME Shell 3.32.1X Server 1.20.4modesetting 1.20.4GCC 11.0.0 20200929ext41920x1080OpenBenchmarking.orgCompiler Details- --disable-multilib --enable-checking=release --enable-languages=c,c++,fortran Processor Details- Scaling Governor: intel_pstate powersave - CPU Microcode: 0x2000064Java Details- OpenJDK Runtime Environment (build 11.0.5+10-post-Ubuntu-0ubuntu1.119.04)Python Details- Python 2.7.16 + Python 3.7.3Security Details- 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

123Result OverviewPhoronix Test Suite100%102%105%107%Algebraic Multi-Grid BenchmarkeSpeak-NG Speech EngineJava Gradle BuildNeatBenchLULESHOCRMyPDFYafaRayTimed MAFFT AlignmentRNNoiseTimed Linux Kernel CompilationoneDNNRodiniaNCNNKeyDBIntel Open Image DenoiseBRL-CADHuginC-BloscSVT-VP9LAMMPS Molecular Dynamics SimulatorLuxCoreRenderInfluxDBTimed LLVM CompilationZstd CompressionG'MICWebP Image EncodeHierarchical INTegrationEmbreeLibRawMlpack BenchmarkTimed Apache CompilationOpenVKLNAMDMobile Neural NetworkPyPerformanceTimed HMMer SearchTesseract OCRTensorFlow LiteRawTherapeelibavif avifencAOM AV1dav1dBasis UniversalGNU Octave BenchmarkGitBYTE Unix BenchmarkCaffeSVT-AV1RenaissanceDaCapo BenchmarkBlenderGPAWTNNOpenCV

7900X Newamg: espeak: Text-To-Speech Synthesisjava-gradle-perf: Reactoronednn: Recurrent Neural Network Inference - f32 - CPUncnn: CPU - googlenetrodinia: OpenMP HotSpot3Donednn: Deconvolution Batch deconv_3d - u8s8f32 - CPUneatbench: CPUncnn: CPU - blazefacerenaissance: Savina Reactors.IOrodinia: OpenMP Streamclusteronednn: Deconvolution Batch deconv_1d - f32 - CPUrenaissance: Genetic Algorithm Using Jenetics + Futuresrenaissance: Rand Forestonednn: Matrix Multiply Batch Shapes Transformer - u8s8f32 - CPUdacapobench: Tradesoaplulesh: renaissance: In-Memory Database Shootoutonednn: IP Batch All - f32 - CPUdacapobench: Tradebeansncnn: CPU - mobilenetrenaissance: Apache Spark PageRankocrmypdf: Processing 60 Page PDF Documentncnn: CPU-v2-v2 - mobilenet-v2yafaray: Total Time For Sample Scenencnn: CPU - mnasnetmafft: Multiple Sequence Alignment - LSU RNAncnn: CPU-v3-v3 - mobilenet-v3renaissance: Scala Dottyncnn: CPU - efficientnet-b0ncnn: CPU - yolov4-tinyonednn: Recurrent Neural Network Training - f32 - CPUpyperformance: nbodydacapobench: Jythononednn: Deconvolution Batch deconv_3d - bf16bf16bf16 - CPUluxcorerender: DLSCrnnoise: ncnn: CPU - squeezenetonednn: IP Batch 1D - u8s8f32 - CPUmlpack: scikit_linearridgeregressionncnn: CPU - resnet18renaissance: Twitter HTTP Requestsembree: Pathtracer - Crownonednn: Convolution Batch Shapes Auto - u8s8f32 - CPUgmic: 2D Function Plotting, 1000 Timesmnn: MobileNetV2_224build-linux-kernel: Time To Compilepyperformance: 2to3dacapobench: H2pyperformance: json_loadsncnn: CPU - resnet50pyperformance: chaosembree: Pathtracer - Asian Dragonlammps: 20k Atomsonednn: IP Batch All - u8s8f32 - CPUopenvkl: vklBenchmarkStructuredVolumencnn: CPU - vgg16avifenc: 10webp: Defaultsvt-vp9: VMAF Optimized - Bosphorus 1080prenaissance: Apache Spark ALSembree: Pathtracer ISPC - Crownncnn: CPU - shufflenet-v2onednn: Deconvolution Batch deconv_3d - f32 - CPUkeydb: webp: Quality 100oidn: Memorialbrl-cad: VGR Performance Metrichugin: Panorama Photo Assistant + Stitching Timencnn: CPU - alexnetrodinia: OpenMP LavaMDblosc: blosclzembree: Pathtracer ISPC - Asian Dragonavifenc: 2webp: Quality 100, Lossless, Highest Compressionavifenc: 8pyperformance: pickle_pure_pythoncompress-zstd: 19svt-vp9: PSNR/SSIM Optimized - Bosphorus 1080pmnn: SqueezeNetV1.0gmic: Plotting Isosurface Of A 3D Volume, 1000 Timesbuild-llvm: Time To Compileavifenc: 0onednn: Convolution Batch Shapes Auto - f32 - CPUbasis: ETC1Scompress-zstd: 3rodinia: OpenMP CFD Solverinfluxdb: 4 - 10000 - 2,5000,1 - 10000mlpack: scikit_icamnn: inception-v3webp: Quality 100, Highest Compressionluxcorerender: Rainbow Colors and Prismonednn: IP Batch 1D - bf16bf16bf16 - CPUdav1d: Summer Nature 4Kpyperformance: floataom-av1: Speed 8 Realtimehint: FLOATmlpack: scikit_qdasvt-vp9: Visual Quality Optimized - Bosphorus 1080ponednn: Matrix Multiply Batch Shapes Transformer - f32 - CPUaom-av1: Speed 4 Two-Passtensorflow-lite: Mobilenet Floatlibraw: Post-Processing Benchmarktensorflow-lite: Inception ResNet V2build-apache: Time To Compilesvt-av1: Enc Mode 4 - 1080pnamd: ATPase Simulation - 327,506 Atomscaffe: GoogleNet - CPU - 100dav1d: Summer Nature 1080ponednn: Convolution Batch Shapes Auto - bf16bf16bf16 - CPUopenvkl: vklBenchmarkbasis: UASTC Level 0embree: Pathtracer ISPC - Asian Dragon Objpyperformance: crypto_pyaesonednn: Matrix Multiply Batch Shapes Transformer - bf16bf16bf16 - CPUaom-av1: Speed 6 Two-Passrodinia: OpenMP Leukocytecaffe: AlexNet - CPU - 100lammps: Rhodopsin Proteinhmmer: Pfam Database Searchmnn: resnet-v2-50rawtherapee: Total Benchmark Timecaffe: AlexNet - CPU - 200blender: BMW27 - CPU-Onlyonednn: Deconvolution Batch deconv_1d - bf16bf16bf16 - CPUtesseract-ocr: Time To OCR 7 Imagesaom-av1: Speed 6 Realtimedav1d: Chimera 1080pblender: Fishy Cat - CPU-Onlyblender: Classroom - CPU-Onlygpaw: Carbon Nanotubepyperformance: django_templateonednn: Deconvolution Batch deconv_1d - u8s8f32 - CPUtensorflow-lite: Inception V4webp: Quality 100, Losslesstnn: CPU - MobileNet v2mnn: mobilenet-v1-1.0caffe: AlexNet - CPU - 1000gmic: 3D Elevated Function In Rand Colors, 100 Timestensorflow-lite: NASNet Mobileblender: Pabellon Barcelona - CPU-Onlyembree: Pathtracer - Asian Dragon Objsvt-av1: Enc Mode 8 - 1080pcaffe: GoogleNet - CPU - 200mlpack: scikit_svmoctave-benchmark: basis: UASTC Level 2git: Time To Complete Common Git Commandsbyte: Dhrystone 2basis: UASTC Level 3tensorflow-lite: Mobilenet Quantdav1d: Chimera 1080p 10-bitonednn: IP Batch All - bf16bf16bf16 - CPUblender: Barbershop - CPU-Onlytensorflow-lite: SqueezeNettnn: CPU - SqueezeNet v1.1caffe: GoogleNet - CPU - 1000basis: UASTC Level 2 + RDO Post-Processingkripke: pyperformance: python_startuppyperformance: regex_compilepyperformance: raytracepyperformance: pathlibpyperformance: gosvt-av1: Enc Mode 0 - 1080paom-av1: Speed 0 Two-Passopencv: DNN - Deep Neural Networkonednn: IP Batch 1D - f32 - CPUrenaissance: Akka Unbalanced Cobwebbed Treerenaissance: Apache Spark Bayes12321811.2930.389285.59585.758814.1098.0492.2037718.41.9423109.93514.0822.513324831.3462184.9790.870165407911.4801014791.11443.4009650716.843941.90030.9805.50162.8254.9710.5214.641800.1166.8225.07202.214102380616.25431.6826.77615.111.189442.3712.122290.01111.606710.06068114.1564.60582.580302446223.823.0995.913.92057.46516.859756744630.22522545.415.3851.466187.942367.48113.53434.343.44890653144.312.26519.7112385555.08210.99225.5269271.217.667454.46341.2165.66940354.3200.607.55719.277605.55790.00810.608750.7094366.520.1381201483.251.4940.9227.0511.798.06587178.7994.632.78430439779.8902743.02156.082.095072.2413396738.66256399325.5093.9631.59855122110496.0412.5830175.677.95615.695897.72.828633.52109.679479717.510133.45135.63153.94896232151.1414.524926.63617.89610.06208.93451.27249.72246.91.46860283685017.909315.6834.21347860463.181181809509.6112.892033.35824379615.168.11630.99450.94141673100.158.72113660986.7495.1659605.44191171296.3661220000721.816291405210.515440817.82130.1330.3143933.1751910146.3761531.73222928.8630.463289.45989.674513.6396.2052.2558517.81.8823820.78313.6632.583574702.2982121.9620.845101411011.8105214861.35444.6064642117.264038.93030.6115.37166.5404.8610.2904.551768.7756.6824.58205.072102373316.52891.7027.07915.211.209932.3612.122328.08411.78619.97412115.8154.66881.472300448123.623.2197.113.95887.55617.064656075820.36036145.775.4431.450189.662370.42613.54694.333.47977650558.982.26719.6612301955.54210.92223.6949346.117.808454.0440.9025.68640054.4201.187.52419.317606.31690.18410.646150.9334338.320.1181194436.451.8141.1727.0111.788.11044177.8295.132.76429449066.3595743.22155.672.096432.2313347838.82255417725.5003.9671.59226122299495.7512.5949176.287.98215.664497.62.836613.52109.642481007.514133.26435.60253.85696039151.5114.545326.69517.92610.40209.39452.25249.96047.01.47076283260717.884315.3984.22147835363.188181669508.8712.911833.33924415415.148.12630.95750.96741712858.158.76213647986.7395.1430605.36191148296.4201220130721.87310.515440817.82130.1330.3142673.311219866.1301534.53023819.1728.249301.16786.170914.20100.0462.1821618.31.9323326.52813.7912.508314691.5582172.2680.859670419811.7050654926.07143.4983634817.044028.32231.3575.44164.8484.9410.4974.651762.7746.7024.91201.086104376516.21881.6726.61014.951.192452.4011.922311.68411.70949.91643114.5334.62082.110304452023.922.9295.914.09227.49716.935456461883.81982045.255.4461.455187.642390.51713.42124.373.45026647412.542.28519.5412279755.21610.90224.7349318.617.718454.17941.0445.71240354.0199.747.50319.179602.00289.54310.574951.0454345.220.0101201966.851.6841.1057.0191.798.08919178.6494.732.61428315485.2755243.12155.362.087052.2313337738.83255323725.6073.9791.59601121854494.2812.6267176.147.97215.646797.92.828113.51109.373480197.494133.10035.54053.81195991151.2414.510026.63417.93609.04209.24451.81250.26147.01.46784283124317.919315.0784.21547916963.287181512508.7912.902533.30724395715.148.11830.99150.90941700019.858.77713648286.6795.0964605.01191041296.3141220333721.67810.515440817.82130.1330.3143013.1771110202.4761526.989OpenBenchmarking.org

Algebraic Multi-Grid Benchmark

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

OpenBenchmarking.orgFigure Of Merit, More Is BetterAlgebraic Multi-Grid Benchmark1235K10K15K20K25KSE +/- 181.95, N = 3SE +/- 292.58, N = 4SE +/- 145.70, N = 321811.2922928.8623819.171. (CC) gcc options: -lparcsr_ls -lparcsr_mv -lseq_mv -lIJ_mv -lkrylov -lHYPRE_utilities -lm -fopenmp -pthread -lmpi
OpenBenchmarking.orgFigure Of Merit, More Is BetterAlgebraic Multi-Grid Benchmark1234K8K12K16K20KMin: 21563.43 / Avg: 21811.29 / Max: 22165.96Min: 22123.89 / Avg: 22928.86 / Max: 23520.07Min: 23634.6 / Avg: 23819.17 / Max: 24106.751. (CC) gcc options: -lparcsr_ls -lparcsr_mv -lseq_mv -lIJ_mv -lkrylov -lHYPRE_utilities -lm -fopenmp -pthread -lmpi

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 Synthesis123714212835SE +/- 0.33, N = 16SE +/- 0.24, N = 4SE +/- 0.14, N = 430.3930.4628.251. (CC) gcc options: -O2 -std=c99
OpenBenchmarking.orgSeconds, Fewer Is BettereSpeak-NG Speech Engine 20200907Text-To-Speech Synthesis123714212835Min: 26.44 / Avg: 30.39 / Max: 32.85Min: 30.15 / Avg: 30.46 / Max: 31.17Min: 27.86 / Avg: 28.25 / Max: 28.51. (CC) gcc options: -O2 -std=c99

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: Reactor12370140210280350SE +/- 4.07, N = 3SE +/- 4.60, N = 9SE +/- 4.29, N = 3285.60289.46301.17
OpenBenchmarking.orgSeconds, Fewer Is BetterJava Gradle BuildGradle Build: Reactor12350100150200250Min: 278.25 / Avg: 285.6 / Max: 292.32Min: 274.36 / Avg: 289.46 / Max: 316.31Min: 296.17 / Avg: 301.17 / Max: 309.7

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 Inference - Data Type: f32 - Engine: CPU12320406080100SE +/- 0.23, N = 3SE +/- 0.13, N = 3SE +/- 0.16, N = 385.7689.6786.17MIN: 84.62MIN: 88.46MIN: 84.951. (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: CPU12320406080100Min: 85.42 / Avg: 85.76 / Max: 86.19Min: 89.42 / Avg: 89.67 / Max: 89.83Min: 85.85 / Avg: 86.17 / Max: 86.341. (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: googlenet12348121620SE +/- 0.25, N = 3SE +/- 0.01, N = 3SE +/- 0.40, N = 314.1013.6314.20MIN: 13.53 / MAX: 56.4MIN: 13.5 / MAX: 14.99MIN: 13.48 / MAX: 63.481. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: googlenet12348121620Min: 13.63 / Avg: 14.1 / Max: 14.48Min: 13.61 / Avg: 13.63 / Max: 13.64Min: 13.58 / Avg: 14.2 / Max: 14.951. (CXX) g++ options: -O3 -rdynamic -lgomp -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 HotSpot3D12320406080100SE +/- 1.37, N = 4SE +/- 0.01, N = 3SE +/- 1.43, N = 498.0596.21100.051. (CXX) g++ options: -O2 -lOpenCL
OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP HotSpot3D12320406080100Min: 96.56 / Avg: 98.05 / Max: 102.15Min: 96.2 / Avg: 96.21 / Max: 96.22Min: 95.95 / Avg: 100.05 / Max: 102.041. (CXX) g++ options: -O2 -lOpenCL

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: CPU1230.50761.01521.52282.03042.538SE +/- 0.02337, N = 3SE +/- 0.02099, N = 3SE +/- 0.01102, N = 32.203772.255852.18216MIN: 2.16MIN: 2.21MIN: 2.151. (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_3d - Data Type: u8s8f32 - Engine: CPU123246810Min: 2.17 / Avg: 2.2 / Max: 2.25Min: 2.23 / Avg: 2.26 / Max: 2.3Min: 2.17 / Avg: 2.18 / Max: 2.21. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

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: CPU123510152025SE +/- 0.17, N = 10SE +/- 0.28, N = 3SE +/- 0.25, N = 318.417.818.3
OpenBenchmarking.orgFPS, More Is BetterNeatBench 5Acceleration: CPU123510152025Min: 17.3 / Avg: 18.4 / Max: 19.1Min: 17.5 / Avg: 17.83 / Max: 18.4Min: 17.8 / Avg: 18.3 / Max: 18.6

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: blazeface1230.43650.8731.30951.7462.1825SE +/- 0.05, N = 3SE +/- 0.00, N = 3SE +/- 0.05, N = 31.941.881.93MIN: 1.85 / MAX: 2.12MIN: 1.85 / MAX: 1.96MIN: 1.84 / MAX: 2.071. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: blazeface123246810Min: 1.88 / Avg: 1.94 / Max: 2.03Min: 1.88 / Avg: 1.88 / Max: 1.88Min: 1.88 / Avg: 1.93 / Max: 2.031. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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.IO1235K10K15K20K25KSE +/- 256.36, N = 5SE +/- 217.82, N = 5SE +/- 179.08, N = 2023109.9423820.7823326.53
OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Savina Reactors.IO1234K8K12K16K20KMin: 22294.75 / Avg: 23109.94 / Max: 23692.46Min: 23186.83 / Avg: 23820.78 / Max: 24354.79Min: 21735.57 / Avg: 23326.53 / Max: 25042.49

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 Streamcluster12348121620SE +/- 0.15, N = 8SE +/- 0.17, N = 15SE +/- 0.16, N = 1514.0813.6613.791. (CXX) g++ options: -O2 -lOpenCL
OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP Streamcluster12348121620Min: 13.06 / Avg: 14.08 / Max: 14.36Min: 12.8 / Avg: 13.66 / Max: 14.35Min: 12.84 / Avg: 13.79 / Max: 14.331. (CXX) g++ options: -O2 -lOpenCL

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: f32 - Engine: CPU1230.58131.16261.74392.32522.9065SE +/- 0.00797, N = 3SE +/- 0.00820, N = 3SE +/- 0.00718, N = 32.513322.583572.50831MIN: 2.45MIN: 2.51MIN: 2.451. (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: CPU123246810Min: 2.5 / Avg: 2.51 / Max: 2.53Min: 2.57 / Avg: 2.58 / Max: 2.6Min: 2.5 / Avg: 2.51 / Max: 2.521. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

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: Genetic Algorithm Using Jenetics + Futures12310002000300040005000SE +/- 51.97, N = 15SE +/- 44.75, N = 20SE +/- 55.22, N = 204831.354702.304691.56
OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Genetic Algorithm Using Jenetics + Futures1238001600240032004000Min: 4468.08 / Avg: 4831.35 / Max: 5038.96Min: 4400.86 / Avg: 4702.3 / Max: 5067.51Min: 4393.15 / Avg: 4691.56 / Max: 5099.84

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Random Forest1235001000150020002500SE +/- 22.85, N = 5SE +/- 15.11, N = 17SE +/- 25.98, N = 52184.982121.962172.27
OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Random Forest123400800120016002000Min: 2124.54 / Avg: 2184.98 / Max: 2232.77Min: 2029.11 / Avg: 2121.96 / Max: 2216.38Min: 2070.25 / Avg: 2172.27 / Max: 2211.46

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: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU1230.19580.39160.58740.78320.979SE +/- 0.002444, N = 3SE +/- 0.001415, N = 3SE +/- 0.000700, N = 30.8701650.8451010.859670MIN: 0.84MIN: 0.82MIN: 0.831. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU123246810Min: 0.87 / Avg: 0.87 / Max: 0.87Min: 0.84 / Avg: 0.85 / Max: 0.85Min: 0.86 / Avg: 0.86 / Max: 0.861. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

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: Tradesoap1239001800270036004500SE +/- 27.44, N = 4SE +/- 37.49, N = 4SE +/- 37.21, N = 3407941104198
OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: Tradesoap1237001400210028003500Min: 4020 / Avg: 4078.5 / Max: 4152Min: 4004 / Avg: 4110 / Max: 4178Min: 4125 / Avg: 4198.33 / Max: 4246

LULESH

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

OpenBenchmarking.orgz/s, More Is BetterLULESH 2.0.31233691215SE +/- 0.16, N = 3SE +/- 0.06, N = 3SE +/- 0.17, N = 411.4811.8111.711. (CXX) g++ options: -O3 -fopenmp -lm -pthread -lmpi_cxx -lmpi
OpenBenchmarking.orgz/s, More Is BetterLULESH 2.0.31233691215Min: 11.22 / Avg: 11.48 / Max: 11.76Min: 11.74 / Avg: 11.81 / Max: 11.92Min: 11.27 / Avg: 11.71 / Max: 12.071. (CXX) g++ options: -O3 -fopenmp -lm -pthread -lmpi_cxx -lmpi

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 Shootout12311002200330044005500SE +/- 60.85, N = 5SE +/- 49.84, N = 25SE +/- 36.91, N = 54791.114861.354926.07
OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: In-Memory Database Shootout1239001800270036004500Min: 4648.93 / Avg: 4791.11 / Max: 5011.95Min: 4238 / Avg: 4861.35 / Max: 5163.14Min: 4841.4 / Avg: 4926.07 / Max: 5059.2

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: CPU1231020304050SE +/- 0.04, N = 3SE +/- 0.56, N = 3SE +/- 0.07, N = 343.4044.6143.50MIN: 41.11MIN: 41.49MIN: 41.121. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: IP Batch All - Data Type: f32 - Engine: CPU123918273645Min: 43.35 / Avg: 43.4 / Max: 43.48Min: 43.84 / Avg: 44.61 / Max: 45.69Min: 43.37 / Avg: 43.5 / Max: 43.571. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

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: Tradebeans12314002800420056007000SE +/- 63.74, N = 4SE +/- 62.97, N = 9SE +/- 37.45, N = 4650764216348
OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: Tradebeans12311002200330044005500Min: 6317 / Avg: 6506.75 / Max: 6592Min: 6172 / Avg: 6421.22 / Max: 6750Min: 6254 / Avg: 6348.25 / Max: 6435

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: mobilenet12348121620SE +/- 0.02, N = 3SE +/- 0.32, N = 3SE +/- 0.22, N = 316.8417.2617.04MIN: 16.72 / MAX: 17.69MIN: 16.74 / MAX: 63.4MIN: 16.72 / MAX: 17.561. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: mobilenet12348121620Min: 16.8 / Avg: 16.84 / Max: 16.88Min: 16.82 / Avg: 17.26 / Max: 17.89Min: 16.81 / Avg: 17.04 / Max: 17.481. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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 PageRank1239001800270036004500SE +/- 43.45, N = 25SE +/- 42.51, N = 25SE +/- 42.16, N = 253941.904038.934028.32
OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Apache Spark PageRank1237001400210028003500Min: 3596.17 / Avg: 3941.9 / Max: 4331.25Min: 3699.03 / Avg: 4038.93 / Max: 4442.62Min: 3665.95 / Avg: 4028.32 / Max: 4422.16

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 8.0.1+dfsgProcessing 60 Page PDF Document123714212835SE +/- 0.33, N = 3SE +/- 0.11, N = 3SE +/- 0.08, N = 330.9830.6131.36
OpenBenchmarking.orgSeconds, Fewer Is BetterOCRMyPDF 8.0.1+dfsgProcessing 60 Page PDF Document123714212835Min: 30.64 / Avg: 30.98 / Max: 31.64Min: 30.47 / Avg: 30.61 / Max: 30.82Min: 31.27 / Avg: 31.36 / Max: 31.52

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-v21231.23752.4753.71254.956.1875SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.09, N = 35.505.375.44MIN: 5.24 / MAX: 5.69MIN: 5.17 / MAX: 6.06MIN: 5.24 / MAX: 5.71. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU-v2-v2 - Model: mobilenet-v2123246810Min: 5.41 / Avg: 5.5 / Max: 5.57Min: 5.31 / Avg: 5.37 / Max: 5.42Min: 5.31 / Avg: 5.44 / Max: 5.621. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

YafaRay

YafaRay is an open-source physically based montecarlo ray-tracing engine. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterYafaRay 3.4.1Total Time For Sample Scene1234080120160200SE +/- 2.59, N = 3SE +/- 2.58, N = 12SE +/- 2.55, N = 3162.83166.54164.851. (CXX) g++ options: -std=c++11 -O3 -ffast-math -rdynamic -ldl -lImath -lIlmImf -lIex -lHalf -lz -lIlmThread -lxml2 -lfreetype -lboost_system -lboost_filesystem -lboost_locale
OpenBenchmarking.orgSeconds, Fewer Is BetterYafaRay 3.4.1Total Time For Sample Scene123306090120150Min: 158.16 / Avg: 162.83 / Max: 167.09Min: 159.18 / Avg: 166.54 / Max: 193.11Min: 160.33 / Avg: 164.85 / Max: 169.161. (CXX) g++ options: -std=c++11 -O3 -ffast-math -rdynamic -ldl -lImath -lIlmImf -lIex -lHalf -lz -lIlmThread -lxml2 -lfreetype -lboost_system -lboost_filesystem -lboost_locale

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: mnasnet1231.11832.23663.35494.47325.5915SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.10, N = 34.974.864.94MIN: 4.71 / MAX: 6.29MIN: 4.69 / MAX: 5MIN: 4.73 / MAX: 5.51. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: mnasnet123246810Min: 4.86 / Avg: 4.97 / Max: 5.03Min: 4.8 / Avg: 4.86 / Max: 4.91Min: 4.84 / Avg: 4.94 / Max: 5.141. (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 RNA1233691215SE +/- 0.06, N = 3SE +/- 0.13, N = 3SE +/- 0.06, N = 310.5210.2910.501. (CC) gcc options: -std=c99 -O3 -lm -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed MAFFT Alignment 7.471Multiple Sequence Alignment - LSU RNA1233691215Min: 10.45 / Avg: 10.52 / Max: 10.64Min: 10.04 / Avg: 10.29 / Max: 10.42Min: 10.44 / Avg: 10.5 / Max: 10.611. (CC) gcc options: -std=c99 -O3 -lm -lpthread

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-v31231.04632.09263.13894.18525.2315SE +/- 0.06, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 34.644.554.65MIN: 4.47 / MAX: 4.79MIN: 4.45 / MAX: 5.99MIN: 4.44 / MAX: 5.591. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU-v3-v3 - Model: mobilenet-v3123246810Min: 4.53 / Avg: 4.64 / Max: 4.7Min: 4.48 / Avg: 4.55 / Max: 4.6Min: 4.61 / Avg: 4.65 / Max: 4.741. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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 Dotty123400800120016002000SE +/- 16.25, N = 5SE +/- 12.30, N = 5SE +/- 10.01, N = 51800.121768.781762.77
OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Scala Dotty12330060090012001500Min: 1774.67 / Avg: 1800.12 / Max: 1860.33Min: 1748.79 / Avg: 1768.77 / Max: 1816.68Min: 1726.83 / Avg: 1762.77 / Max: 1786.79

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-b0123246810SE +/- 0.12, N = 3SE +/- 0.02, N = 3SE +/- 0.08, N = 36.826.686.70MIN: 6.51 / MAX: 7.14MIN: 6.54 / MAX: 6.83MIN: 6.49 / MAX: 6.941. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: efficientnet-b01233691215Min: 6.6 / Avg: 6.82 / Max: 7.01Min: 6.65 / Avg: 6.68 / Max: 6.7Min: 6.6 / Avg: 6.7 / Max: 6.851. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: yolov4-tiny123612182430SE +/- 0.31, N = 3SE +/- 0.14, N = 3SE +/- 0.40, N = 325.0724.5824.91MIN: 24.64 / MAX: 25.79MIN: 24.22 / MAX: 26.04MIN: 24.24 / MAX: 26.591. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: yolov4-tiny123612182430Min: 24.71 / Avg: 25.07 / Max: 25.68Min: 24.31 / Avg: 24.58 / Max: 24.78Min: 24.33 / Avg: 24.91 / Max: 25.671. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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: CPU12350100150200250SE +/- 0.24, N = 3SE +/- 0.48, N = 3SE +/- 0.29, N = 3202.21205.07201.09MIN: 200.84MIN: 203.25MIN: 199.441. (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 Training - Data Type: f32 - Engine: CPU1234080120160200Min: 201.78 / Avg: 202.21 / Max: 202.61Min: 204.18 / Avg: 205.07 / Max: 205.84Min: 200.53 / Avg: 201.09 / Max: 201.521. (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: nbody12320406080100SE +/- 0.88, N = 3SE +/- 1.00, N = 3SE +/- 1.20, N = 3102102104
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: nbody12320406080100Min: 101 / Avg: 102.33 / Max: 104Min: 101 / Avg: 102 / Max: 104Min: 102 / Avg: 103.67 / Max: 106

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: Jython1238001600240032004000SE +/- 7.72, N = 4SE +/- 25.83, N = 4SE +/- 30.00, N = 4380637333765
OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: Jython1237001400210028003500Min: 3791 / Avg: 3805.75 / Max: 3826Min: 3660 / Avg: 3732.5 / Max: 3781Min: 3720 / Avg: 3765.25 / Max: 3848

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: bf16bf16bf16 - Engine: CPU12348121620SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 316.2516.5316.22MIN: 16.17MIN: 16.44MIN: 16.141. (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_3d - Data Type: bf16bf16bf16 - Engine: CPU12348121620Min: 16.22 / Avg: 16.25 / Max: 16.31Min: 16.5 / Avg: 16.53 / Max: 16.58Min: 16.21 / Avg: 16.22 / Max: 16.221. (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: DLSC1230.38250.7651.14751.531.9125SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 31.681.701.67MIN: 1.61 / MAX: 1.72MIN: 1.63 / MAX: 1.78MIN: 1.6 / MAX: 1.72
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.3Scene: DLSC123246810Min: 1.67 / Avg: 1.68 / Max: 1.69Min: 1.67 / Avg: 1.7 / Max: 1.73Min: 1.67 / Avg: 1.67 / Max: 1.69

RNNoise

OpenBenchmarking.orgSeconds, Fewer Is BetterRNNoise 2020-06-28123612182430SE +/- 0.11, N = 3SE +/- 0.33, N = 13SE +/- 0.01, N = 326.7827.0826.611. (CC) gcc options: -O2 -pedantic -fvisibility=hidden
OpenBenchmarking.orgSeconds, Fewer Is BetterRNNoise 2020-06-28123612182430Min: 26.56 / Avg: 26.78 / Max: 26.89Min: 26.58 / Avg: 27.08 / Max: 30.99Min: 26.59 / Avg: 26.61 / Max: 26.621. (CC) gcc options: -O2 -pedantic -fvisibility=hidden

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: squeezenet12348121620SE +/- 0.15, N = 3SE +/- 0.18, N = 3SE +/- 0.18, N = 315.1115.2114.95MIN: 14.85 / MAX: 61.59MIN: 14.81 / MAX: 15.89MIN: 14.57 / MAX: 19.071. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: squeezenet12348121620Min: 14.93 / Avg: 15.11 / Max: 15.41Min: 14.88 / Avg: 15.21 / Max: 15.51Min: 14.65 / Avg: 14.95 / Max: 15.271. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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: CPU1230.27220.54440.81661.08881.361SE +/- 0.00077, N = 3SE +/- 0.00275, N = 3SE +/- 0.00566, N = 31.189441.209931.19245MIN: 1.16MIN: 1.18MIN: 1.161. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: IP Batch 1D - Data Type: u8s8f32 - Engine: CPU123246810Min: 1.19 / Avg: 1.19 / Max: 1.19Min: 1.2 / Avg: 1.21 / Max: 1.21Min: 1.19 / Avg: 1.19 / Max: 1.21. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Mlpack Benchmark

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

OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_linearridgeregression1230.541.081.622.162.7SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 32.372.362.40
OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_linearridgeregression123246810Min: 2.37 / Avg: 2.37 / Max: 2.38Min: 2.33 / Avg: 2.36 / Max: 2.39Min: 2.37 / Avg: 2.4 / Max: 2.41

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: resnet181233691215SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.22, N = 312.1212.1211.92MIN: 11.98 / MAX: 13.14MIN: 11.99 / MAX: 12.25MIN: 11.41 / MAX: 12.561. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: resnet1812348121620Min: 12.02 / Avg: 12.12 / Max: 12.23Min: 12.03 / Avg: 12.12 / Max: 12.17Min: 11.47 / Avg: 11.92 / Max: 12.141. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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: Twitter HTTP Requests1235001000150020002500SE +/- 7.20, N = 5SE +/- 12.66, N = 5SE +/- 10.38, N = 52290.012328.082311.68
OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Twitter HTTP Requests123400800120016002000Min: 2262.92 / Avg: 2290.01 / Max: 2304.01Min: 2283.3 / Avg: 2328.08 / Max: 2358.67Min: 2275.09 / Avg: 2311.68 / Max: 2334.68

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.9.0Binary: Pathtracer - Model: Crown1233691215SE +/- 0.14, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 311.6111.7911.71MIN: 11.24 / MAX: 12MIN: 11.66 / MAX: 11.96MIN: 11.61 / MAX: 11.87
OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.9.0Binary: Pathtracer - Model: Crown1233691215Min: 11.43 / Avg: 11.61 / Max: 11.88Min: 11.72 / Avg: 11.79 / Max: 11.83Min: 11.67 / Avg: 11.71 / Max: 11.74

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: CPU1233691215SE +/- 0.06923, N = 3SE +/- 0.01762, N = 3SE +/- 0.03654, N = 310.060689.974129.91643MIN: 9.92MIN: 9.89MIN: 9.821. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU1233691215Min: 9.99 / Avg: 10.06 / Max: 10.2Min: 9.95 / Avg: 9.97 / Max: 10.01Min: 9.87 / Avg: 9.92 / Max: 9.991. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

G'MIC

G'MIC is an open-source framework for image processing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterG'MICTest: 2D Function Plotting, 1000 Times123306090120150SE +/- 1.20, N = 8SE +/- 1.22, N = 3SE +/- 1.02, N = 15114.16115.82114.531. Version 2.4.5, Copyright (c) 2008-2019, David Tschumperle.
OpenBenchmarking.orgSeconds, Fewer Is BetterG'MICTest: 2D Function Plotting, 1000 Times12320406080100Min: 110.93 / Avg: 114.16 / Max: 120.9Min: 113.87 / Avg: 115.81 / Max: 118.06Min: 110.9 / Avg: 114.53 / Max: 122.641. Version 2.4.5, Copyright (c) 2008-2019, David Tschumperle.

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_2241231.05032.10063.15094.20125.2515SE +/- 0.005, N = 3SE +/- 0.058, N = 3SE +/- 0.020, N = 34.6054.6684.620MIN: 4.46 / MAX: 5.98MIN: 4.47 / MAX: 67.24MIN: 4.44 / MAX: 6.021. (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: MobileNetV2_224123246810Min: 4.6 / Avg: 4.6 / Max: 4.61Min: 4.6 / Avg: 4.67 / Max: 4.78Min: 4.58 / Avg: 4.62 / Max: 4.651. (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

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 Compile12320406080100SE +/- 0.67, N = 3SE +/- 0.75, N = 3SE +/- 0.86, N = 382.5881.4782.11
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 5.4Time To Compile1231632486480Min: 81.85 / Avg: 82.58 / Max: 83.92Min: 80.55 / Avg: 81.47 / Max: 82.96Min: 81.07 / Avg: 82.11 / Max: 83.81

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: 2to312370140210280350SE +/- 0.67, N = 3SE +/- 1.33, N = 3302300304
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: 2to312350100150200250Min: 301 / Avg: 301.67 / Max: 303Min: 301 / Avg: 303.67 / Max: 305

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: H212310002000300040005000SE +/- 31.29, N = 18SE +/- 41.85, N = 20SE +/- 41.45, N = 4446244814520
OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: H21238001600240032004000Min: 4144 / Avg: 4462.44 / Max: 4757Min: 4163 / Avg: 4480.9 / Max: 4868Min: 4439 / Avg: 4520.25 / Max: 4632

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_loads123612182430SE +/- 0.09, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 323.823.623.9
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: json_loads123612182430Min: 23.6 / Avg: 23.77 / Max: 23.9Min: 23.5 / Avg: 23.63 / Max: 23.7Min: 23.8 / Avg: 23.9 / Max: 24

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: resnet50123612182430SE +/- 0.04, N = 3SE +/- 0.07, N = 3SE +/- 0.24, N = 323.0923.2122.92MIN: 22.94 / MAX: 23.66MIN: 22.97 / MAX: 24.73MIN: 22.34 / MAX: 24.561. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: resnet50123510152025Min: 23.04 / Avg: 23.09 / Max: 23.16Min: 23.09 / Avg: 23.21 / Max: 23.33Min: 22.45 / Avg: 22.92 / Max: 23.261. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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: chaos12320406080100SE +/- 0.12, N = 3SE +/- 1.12, N = 3SE +/- 0.09, N = 395.997.195.9
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: chaos12320406080100Min: 95.7 / Avg: 95.87 / Max: 96.1Min: 95.8 / Avg: 97.07 / Max: 99.3Min: 95.8 / Avg: 95.93 / Max: 96.1

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.9.0Binary: Pathtracer - Model: Asian Dragon12348121620SE +/- 0.12, N = 3SE +/- 0.05, N = 3SE +/- 0.11, N = 313.9213.9614.09MIN: 13.7 / MAX: 14.23MIN: 13.81 / MAX: 14.14MIN: 13.84 / MAX: 14.36
OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.9.0Binary: Pathtracer - Model: Asian Dragon12348121620Min: 13.73 / Avg: 13.92 / Max: 14.13Min: 13.86 / Avg: 13.96 / Max: 14.04Min: 13.88 / Avg: 14.09 / Max: 14.24

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: 20k Atoms123246810SE +/- 0.015, N = 3SE +/- 0.015, N = 3SE +/- 0.011, N = 37.4657.5567.4971. (CXX) g++ options: -O3 -pthread -lm
OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 24Aug2020Model: 20k Atoms1233691215Min: 7.44 / Avg: 7.46 / Max: 7.49Min: 7.54 / Avg: 7.56 / Max: 7.59Min: 7.48 / Avg: 7.5 / Max: 7.521. (CXX) g++ options: -O3 -pthread -lm

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: IP Batch All - Data Type: u8s8f32 - Engine: CPU12348121620SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 316.8617.0616.94MIN: 16.37MIN: 16.72MIN: 16.431. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: IP Batch All - Data Type: u8s8f32 - Engine: CPU12348121620Min: 16.8 / Avg: 16.86 / Max: 16.94Min: 17.05 / Avg: 17.06 / Max: 17.09Min: 16.86 / Avg: 16.94 / Max: 171. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenVKL

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

OpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 0.9Benchmark: vklBenchmarkStructuredVolume12312M24M36M48M60MSE +/- 640455.21, N = 3SE +/- 102024.09, N = 3SE +/- 254970.05, N = 356744630.2356075820.3656461883.82MIN: 1223481 / MAX: 389287872MIN: 1231969 / MAX: 373823712MIN: 1241223 / MAX: 371099664
OpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 0.9Benchmark: vklBenchmarkStructuredVolume12310M20M30M40M50MMin: 55595335.19 / Avg: 56744630.23 / Max: 57809062.7Min: 55898283.73 / Avg: 56075820.36 / Max: 56251693.81Min: 56104698.62 / Avg: 56461883.82 / Max: 56955664.03

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: vgg161231020304050SE +/- 0.20, N = 3SE +/- 0.28, N = 3SE +/- 0.47, N = 345.4145.7745.25MIN: 45.02 / MAX: 46.7MIN: 45.15 / MAX: 93MIN: 44.04 / MAX: 85.021. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: vgg16123918273645Min: 45.11 / Avg: 45.41 / Max: 45.8Min: 45.23 / Avg: 45.77 / Max: 46.19Min: 44.37 / Avg: 45.25 / Max: 45.981. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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: 101231.22542.45083.67624.90166.127SE +/- 0.011, N = 3SE +/- 0.011, N = 3SE +/- 0.001, N = 35.3855.4435.4461. (CXX) g++ options: -O3 -fPIC
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.7.3Encoder Speed: 10123246810Min: 5.37 / Avg: 5.39 / Max: 5.4Min: 5.43 / Avg: 5.44 / Max: 5.46Min: 5.45 / Avg: 5.45 / Max: 5.451. (CXX) g++ options: -O3 -fPIC

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: Default1230.32990.65980.98971.31961.6495SE +/- 0.009, N = 3SE +/- 0.009, N = 3SE +/- 0.013, N = 31.4661.4501.4551. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -lpng16 -ljpeg
OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Default123246810Min: 1.45 / Avg: 1.47 / Max: 1.48Min: 1.43 / Avg: 1.45 / Max: 1.46Min: 1.43 / Avg: 1.45 / Max: 1.471. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -lpng16 -ljpeg

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.1Tuning: VMAF Optimized - Input: Bosphorus 1080p1234080120160200SE +/- 3.14, N = 3SE +/- 2.67, N = 4SE +/- 3.06, N = 3187.94189.66187.641. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.1Tuning: VMAF Optimized - Input: Bosphorus 1080p123306090120150Min: 181.65 / Avg: 187.94 / Max: 191.2Min: 181.71 / Avg: 189.66 / Max: 193.05Min: 181.54 / Avg: 187.64 / Max: 191.081. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -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 ALS1235001000150020002500SE +/- 19.20, N = 5SE +/- 13.21, N = 5SE +/- 17.15, N = 172367.482370.432390.52
OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Apache Spark ALS123400800120016002000Min: 2310.92 / Avg: 2367.48 / Max: 2415.06Min: 2341.42 / Avg: 2370.43 / Max: 2415.49Min: 2282.05 / Avg: 2390.52 / Max: 2529.15

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.9.0Binary: Pathtracer ISPC - Model: Crown1233691215SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 313.5313.5513.42MIN: 13.4 / MAX: 13.76MIN: 13.41 / MAX: 13.78MIN: 13.28 / MAX: 13.65
OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.9.0Binary: Pathtracer ISPC - Model: Crown12348121620Min: 13.49 / Avg: 13.53 / Max: 13.58Min: 13.49 / Avg: 13.55 / Max: 13.59Min: 13.36 / Avg: 13.42 / Max: 13.46

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: shufflenet-v21230.98331.96662.94993.93324.9165SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 34.344.334.37MIN: 4.24 / MAX: 4.94MIN: 4.27 / MAX: 4.4MIN: 4.25 / MAX: 5.791. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: shufflenet-v2123246810Min: 4.29 / Avg: 4.34 / Max: 4.4Min: 4.32 / Avg: 4.33 / Max: 4.34Min: 4.32 / Avg: 4.37 / Max: 4.431. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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: CPU1230.78291.56582.34873.13163.9145SE +/- 0.00242, N = 3SE +/- 0.01044, N = 3SE +/- 0.00663, N = 33.448903.479773.45026MIN: 3.4MIN: 3.43MIN: 3.41. (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_3d - Data Type: f32 - Engine: CPU123246810Min: 3.45 / Avg: 3.45 / Max: 3.45Min: 3.46 / Avg: 3.48 / Max: 3.5Min: 3.44 / Avg: 3.45 / Max: 3.461. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

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 6.0.16123140K280K420K560K700KSE +/- 2274.07, N = 3SE +/- 837.37, N = 3SE +/- 422.72, N = 3653144.31650558.98647412.541. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
OpenBenchmarking.orgOps/sec, More Is BetterKeyDB 6.0.16123110K220K330K440K550KMin: 649939.36 / Avg: 653144.31 / Max: 657541.49Min: 648914.86 / Avg: 650558.98 / Max: 651657.14Min: 646854.48 / Avg: 647412.54 / Max: 648241.571. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

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 1001230.51411.02821.54232.05642.5705SE +/- 0.009, N = 3SE +/- 0.010, N = 3SE +/- 0.015, N = 32.2652.2672.2851. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -lpng16 -ljpeg
OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100123246810Min: 2.25 / Avg: 2.27 / Max: 2.28Min: 2.25 / Avg: 2.27 / Max: 2.29Min: 2.26 / Avg: 2.29 / Max: 2.311. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -lpng16 -ljpeg

Intel Open Image Denoise

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

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.2.0Scene: Memorial123510152025SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 319.7119.6619.54
OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.2.0Scene: Memorial123510152025Min: 19.66 / Avg: 19.71 / Max: 19.75Min: 19.61 / Avg: 19.66 / Max: 19.69Min: 19.46 / Avg: 19.54 / Max: 19.63

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 Metric12330K60K90K120K150K1238551230191227971. (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 -lXi -lGLU -lGL -lGLdispatch -lX11 -lXext -lXrender -lpthread -ldl -luuid -lm

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 Time1231224364860SE +/- 0.35, N = 3SE +/- 0.78, N = 4SE +/- 0.42, N = 355.0855.5455.22
OpenBenchmarking.orgSeconds, Fewer Is BetterHuginPanorama Photo Assistant + Stitching Time1231122334455Min: 54.38 / Avg: 55.08 / Max: 55.46Min: 53.98 / Avg: 55.54 / Max: 57.56Min: 54.73 / Avg: 55.22 / Max: 56.04

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: alexnet1233691215SE +/- 0.07, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 310.9910.9210.90MIN: 10.85 / MAX: 37.5MIN: 10.86 / MAX: 11.62MIN: 10.83 / MAX: 111. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: alexnet1233691215Min: 10.9 / Avg: 10.99 / Max: 11.13Min: 10.9 / Avg: 10.92 / Max: 10.93Min: 10.86 / Avg: 10.9 / Max: 10.921. (CXX) g++ options: -O3 -rdynamic -lgomp -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 LavaMD12350100150200250SE +/- 0.96, N = 3SE +/- 0.55, N = 3SE +/- 1.60, N = 3225.53223.69224.731. (CXX) g++ options: -O2 -lOpenCL
OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP LavaMD1234080120160200Min: 224.01 / Avg: 225.53 / Max: 227.3Min: 222.64 / Avg: 223.69 / Max: 224.48Min: 222.22 / Avg: 224.73 / Max: 227.711. (CXX) g++ options: -O2 -lOpenCL

C-Blosc

A simple, compressed, fast and persistent data store library for C. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterC-Blosc 2.0 Beta 5Compressor: blosclz1232K4K6K8K10KSE +/- 3.65, N = 3SE +/- 9.60, N = 3SE +/- 22.72, N = 39271.29346.19318.61. (CXX) g++ options: -rdynamic
OpenBenchmarking.orgMB/s, More Is BetterC-Blosc 2.0 Beta 5Compressor: blosclz12316003200480064008000Min: 9266.6 / Avg: 9271.2 / Max: 9278.4Min: 9329 / Avg: 9346.07 / Max: 9362.2Min: 9273.5 / Avg: 9318.63 / Max: 9345.81. (CXX) g++ options: -rdynamic

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.9.0Binary: Pathtracer ISPC - Model: Asian Dragon12348121620SE +/- 0.06, N = 3SE +/- 0.14, N = 3SE +/- 0.05, N = 317.6717.8117.72MIN: 17.49 / MAX: 17.87MIN: 17.44 / MAX: 18.21MIN: 17.53 / MAX: 17.97
OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.9.0Binary: Pathtracer ISPC - Model: Asian Dragon123510152025Min: 17.56 / Avg: 17.67 / Max: 17.76Min: 17.57 / Avg: 17.81 / Max: 18.04Min: 17.63 / Avg: 17.72 / Max: 17.8

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: 21231224364860SE +/- 0.31, N = 3SE +/- 0.24, N = 3SE +/- 0.11, N = 354.4654.0454.181. (CXX) g++ options: -O3 -fPIC
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.7.3Encoder Speed: 21231122334455Min: 53.88 / Avg: 54.46 / Max: 54.96Min: 53.56 / Avg: 54.04 / Max: 54.32Min: 54.03 / Avg: 54.18 / Max: 54.391. (CXX) g++ options: -O3 -fPIC

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 Compression123918273645SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 341.2240.9041.041. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -lpng16 -ljpeg
OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Lossless, Highest Compression123918273645Min: 41.19 / Avg: 41.22 / Max: 41.26Min: 40.85 / Avg: 40.9 / Max: 40.99Min: 40.98 / Avg: 41.04 / Max: 41.111. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -lpng16 -ljpeg

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: 81231.28522.57043.85565.14086.426SE +/- 0.007, N = 3SE +/- 0.027, N = 3SE +/- 0.019, N = 35.6695.6865.7121. (CXX) g++ options: -O3 -fPIC
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.7.3Encoder Speed: 8123246810Min: 5.66 / Avg: 5.67 / Max: 5.68Min: 5.64 / Avg: 5.69 / Max: 5.73Min: 5.69 / Avg: 5.71 / Max: 5.751. (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: pickle_pure_python12390180270360450SE +/- 2.40, N = 3SE +/- 0.67, N = 3403400403
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: pickle_pure_python12370140210280350Min: 397 / Avg: 400.33 / Max: 405Min: 402 / Avg: 402.67 / Max: 404

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: 191231224364860SE +/- 0.00, N = 3SE +/- 0.32, N = 3SE +/- 0.38, N = 354.354.454.01. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.5Compression Level: 191231122334455Min: 54.3 / Avg: 54.3 / Max: 54.3Min: 53.8 / Avg: 54.43 / Max: 54.8Min: 53.4 / Avg: 53.97 / Max: 54.71. (CC) gcc options: -O3 -pthread -lz -llzma

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.1Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p1234080120160200SE +/- 0.14, N = 3SE +/- 0.33, N = 3SE +/- 0.81, N = 3200.60201.18199.741. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.1Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p1234080120160200Min: 200.33 / Avg: 200.6 / Max: 200.8Min: 200.53 / Avg: 201.18 / Max: 201.61Min: 198.68 / Avg: 199.74 / Max: 201.341. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

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: SqueezeNetV1.0123246810SE +/- 0.036, N = 3SE +/- 0.021, N = 3SE +/- 0.018, N = 37.5577.5247.503MIN: 7.35 / MAX: 28.35MIN: 7.37 / MAX: 8.84MIN: 7.35 / MAX: 9.051. (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: SqueezeNetV1.01233691215Min: 7.52 / Avg: 7.56 / Max: 7.63Min: 7.48 / Avg: 7.52 / Max: 7.55Min: 7.47 / Avg: 7.5 / Max: 7.521. (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

G'MIC

G'MIC is an open-source framework for image processing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterG'MICTest: Plotting Isosurface Of A 3D Volume, 1000 Times123510152025SE +/- 0.17, N = 3SE +/- 0.08, N = 3SE +/- 0.03, N = 319.2819.3219.181. Version 2.4.5, Copyright (c) 2008-2019, David Tschumperle.
OpenBenchmarking.orgSeconds, Fewer Is BetterG'MICTest: Plotting Isosurface Of A 3D Volume, 1000 Times123510152025Min: 19.1 / Avg: 19.28 / Max: 19.61Min: 19.24 / Avg: 19.32 / Max: 19.47Min: 19.13 / Avg: 19.18 / Max: 19.221. Version 2.4.5, Copyright (c) 2008-2019, David Tschumperle.

Timed LLVM Compilation

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

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 10.0Time To Compile123130260390520650SE +/- 2.27, N = 3SE +/- 4.02, N = 3SE +/- 3.87, N = 3605.56606.32602.00
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 10.0Time To Compile123110220330440550Min: 603.14 / Avg: 605.56 / Max: 610.09Min: 598.48 / Avg: 606.32 / Max: 611.77Min: 594.33 / Avg: 602 / Max: 606.7

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: 012320406080100SE +/- 0.59, N = 3SE +/- 0.07, N = 3SE +/- 0.30, N = 390.0190.1889.541. (CXX) g++ options: -O3 -fPIC
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.7.3Encoder Speed: 012320406080100Min: 88.97 / Avg: 90.01 / Max: 91Min: 90.07 / Avg: 90.18 / Max: 90.3Min: 89.02 / Avg: 89.54 / Max: 90.051. (CXX) g++ options: -O3 -fPIC

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: CPU1233691215SE +/- 0.00, N = 3SE +/- 0.07, N = 3SE +/- 0.01, N = 310.6110.6510.57MIN: 10.55MIN: 10.52MIN: 10.51. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU1233691215Min: 10.61 / Avg: 10.61 / Max: 10.61Min: 10.57 / Avg: 10.65 / Max: 10.79Min: 10.57 / Avg: 10.57 / Max: 10.591. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Basis Universal

Basis Universal is a GPU texture codoec. This test times how long it takes to convert sRGB PNGs into Basis Univeral assets with various settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.12Settings: ETC1S1231224364860SE +/- 0.24, N = 3SE +/- 0.32, N = 3SE +/- 0.34, N = 350.7150.9351.051. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.12Settings: ETC1S1231020304050Min: 50.37 / Avg: 50.71 / Max: 51.18Min: 50.3 / Avg: 50.93 / Max: 51.34Min: 50.37 / Avg: 51.05 / Max: 51.381. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -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: 31239001800270036004500SE +/- 5.94, N = 3SE +/- 2.76, N = 3SE +/- 4.38, N = 34366.54338.34345.21. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.5Compression Level: 31238001600240032004000Min: 4359 / Avg: 4366.47 / Max: 4378.2Min: 4332.8 / Avg: 4338.3 / Max: 4341.4Min: 4336.6 / Avg: 4345.17 / Max: 43511. (CC) gcc options: -O3 -pthread -lz -llzma

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 Solver123510152025SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 320.1420.1220.011. (CXX) g++ options: -O2 -lOpenCL
OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP CFD Solver123510152025Min: 20.07 / Avg: 20.14 / Max: 20.2Min: 20.04 / Avg: 20.12 / Max: 20.17Min: 19.92 / Avg: 20.01 / Max: 20.081. (CXX) g++ options: -O2 -lOpenCL

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: 10000123300K600K900K1200K1500KSE +/- 1855.52, N = 3SE +/- 5496.75, N = 3SE +/- 2799.34, N = 31201483.21194436.41201966.8
OpenBenchmarking.orgval/sec, More Is BetterInfluxDB 1.8.2Concurrent Streams: 4 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000123200K400K600K800K1000KMin: 1198489.9 / Avg: 1201483.23 / Max: 1204879.6Min: 1183557.8 / Avg: 1194436.43 / Max: 1201248.4Min: 1198179.7 / Avg: 1201966.8 / Max: 1207431.4

Mlpack Benchmark

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

OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_ica1231224364860SE +/- 0.13, N = 3SE +/- 0.16, N = 3SE +/- 0.08, N = 351.4951.8151.68
OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_ica1231020304050Min: 51.26 / Avg: 51.49 / Max: 51.72Min: 51.49 / Avg: 51.81 / Max: 51.98Min: 51.54 / Avg: 51.68 / Max: 51.79

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-v3123918273645SE +/- 0.32, N = 3SE +/- 0.15, N = 3SE +/- 0.17, N = 340.9241.1741.11MIN: 40.1 / MAX: 98.32MIN: 40.61 / MAX: 100.93MIN: 40.44 / MAX: 104.581. (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: inception-v3123918273645Min: 40.32 / Avg: 40.92 / Max: 41.42Min: 40.88 / Avg: 41.17 / Max: 41.4Min: 40.83 / Avg: 41.11 / Max: 41.411. (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

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 Compression123246810SE +/- 0.019, N = 3SE +/- 0.017, N = 3SE +/- 0.016, N = 37.0517.0117.0191. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -lpng16 -ljpeg
OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Highest Compression1233691215Min: 7.02 / Avg: 7.05 / Max: 7.08Min: 6.99 / Avg: 7.01 / Max: 7.05Min: 6.99 / Avg: 7.02 / Max: 7.041. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -lpng16 -ljpeg

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 Prism1230.40280.80561.20841.61122.014SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 31.791.781.79MIN: 1.78 / MAX: 1.82MIN: 1.76 / MAX: 1.82MIN: 1.77 / MAX: 1.85
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.3Scene: Rainbow Colors and Prism123246810Min: 1.78 / Avg: 1.79 / Max: 1.8Min: 1.76 / Avg: 1.78 / Max: 1.8Min: 1.77 / Avg: 1.79 / Max: 1.81

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: IP Batch 1D - Data Type: bf16bf16bf16 - Engine: CPU123246810SE +/- 0.00100, N = 3SE +/- 0.03696, N = 3SE +/- 0.02158, N = 38.065878.110448.08919MIN: 8MIN: 8MIN: 81. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: IP Batch 1D - Data Type: bf16bf16bf16 - Engine: CPU1233691215Min: 8.06 / Avg: 8.07 / Max: 8.07Min: 8.07 / Avg: 8.11 / Max: 8.18Min: 8.06 / Avg: 8.09 / Max: 8.131. (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 4K1234080120160200SE +/- 0.11, N = 3SE +/- 0.15, N = 3SE +/- 0.31, N = 3178.79177.82178.64MIN: 156.78 / MAX: 198.38MIN: 155.22 / MAX: 197.63MIN: 154.31 / MAX: 198.671. (CC) gcc options: -pthread
OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Summer Nature 4K123306090120150Min: 178.63 / Avg: 178.79 / Max: 179Min: 177.55 / Avg: 177.82 / Max: 178.06Min: 178.13 / Avg: 178.64 / Max: 179.21. (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: float12320406080100SE +/- 0.30, N = 3SE +/- 0.07, N = 3SE +/- 0.41, N = 394.695.194.7
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: float12320406080100Min: 94 / Avg: 94.6 / Max: 94.9Min: 95 / Avg: 95.07 / Max: 95.2Min: 94 / Avg: 94.73 / Max: 95.4

AOM AV1

This is a simple test of the AOMedia AV1 encoder run on the CPU with a sample video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 2.0Encoder Mode: Speed 8 Realtime123816243240SE +/- 0.04, N = 3SE +/- 0.05, N = 3SE +/- 0.11, N = 332.7832.7632.611. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 2.0Encoder Mode: Speed 8 Realtime123714212835Min: 32.71 / Avg: 32.78 / Max: 32.86Min: 32.68 / Avg: 32.76 / Max: 32.85Min: 32.43 / Avg: 32.61 / Max: 32.81. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

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: FLOAT12390M180M270M360M450MSE +/- 393118.30, N = 3SE +/- 325962.67, N = 3SE +/- 645266.62, N = 3430439779.89429449066.36428315485.281. (CC) gcc options: -O3 -march=native -lm
OpenBenchmarking.orgQUIPs, More Is BetterHierarchical INTegration 1.0Test: FLOAT12370M140M210M280M350MMin: 429654049.97 / Avg: 430439779.89 / Max: 430857085.8Min: 428806891.72 / Avg: 429449066.36 / Max: 429867435.64Min: 427449845.58 / Avg: 428315485.28 / Max: 429577224.891. (CC) gcc options: -O3 -march=native -lm

Mlpack Benchmark

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

OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_qda1231020304050SE +/- 0.11, N = 3SE +/- 0.32, N = 14SE +/- 0.30, N = 343.0243.2243.12
OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_qda123918273645Min: 42.83 / Avg: 43.02 / Max: 43.21Min: 42.58 / Avg: 43.22 / Max: 47.27Min: 42.59 / Avg: 43.12 / Max: 43.64

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.1Tuning: Visual Quality Optimized - Input: Bosphorus 1080p123306090120150SE +/- 0.25, N = 3SE +/- 0.36, N = 3SE +/- 0.51, N = 3156.08155.67155.361. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.1Tuning: Visual Quality Optimized - Input: Bosphorus 1080p123306090120150Min: 155.68 / Avg: 156.08 / Max: 156.54Min: 154.96 / Avg: 155.67 / Max: 156.09Min: 154.48 / Avg: 155.36 / Max: 156.251. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the 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: CPU1230.47170.94341.41511.88682.3585SE +/- 0.01389, N = 3SE +/- 0.00344, N = 3SE +/- 0.00361, N = 32.095072.096432.08705MIN: 2.04MIN: 2.06MIN: 2.061. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU123246810Min: 2.07 / Avg: 2.1 / Max: 2.12Min: 2.09 / Avg: 2.1 / Max: 2.1Min: 2.08 / Avg: 2.09 / Max: 2.091. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

AOM AV1

This is a simple test of the AOMedia AV1 encoder run on the CPU with a sample video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 2.0Encoder Mode: Speed 4 Two-Pass1230.5041.0081.5122.0162.52SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 32.242.232.231. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 2.0Encoder Mode: Speed 4 Two-Pass123246810Min: 2.23 / Avg: 2.24 / Max: 2.24Min: 2.22 / Avg: 2.23 / Max: 2.25Min: 2.23 / Avg: 2.23 / Max: 2.231. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

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 Float12330K60K90K120K150KSE +/- 99.22, N = 3SE +/- 36.88, N = 3SE +/- 74.51, N = 3133967133478133377
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet Float12320K40K60K80K100KMin: 133769 / Avg: 133967.33 / Max: 134072Min: 133429 / Avg: 133477.67 / Max: 133550Min: 133274 / Avg: 133377.33 / Max: 133522

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 Benchmark123918273645SE +/- 0.07, N = 3SE +/- 0.08, N = 3SE +/- 0.01, N = 338.6638.8238.831. (CXX) g++ options: -O2 -fopenmp -ljpeg -lz -lm
OpenBenchmarking.orgMpix/sec, More Is BetterLibRaw 0.20Post-Processing Benchmark123816243240Min: 38.55 / Avg: 38.66 / Max: 38.79Min: 38.71 / Avg: 38.82 / Max: 38.98Min: 38.82 / Avg: 38.83 / Max: 38.841. (CXX) g++ options: -O2 -fopenmp -ljpeg -lz -lm

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 V2123500K1000K1500K2000K2500KSE +/- 1919.48, N = 3SE +/- 1370.60, N = 3SE +/- 2061.83, N = 3256399325541772553237
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Inception ResNet V2123400K800K1200K1600K2000KMin: 2560160 / Avg: 2563993.33 / Max: 2566090Min: 2551520 / Avg: 2554176.67 / Max: 2556090Min: 2549210 / Avg: 2553236.67 / Max: 2556020

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 Compile123612182430SE +/- 0.02, N = 3SE +/- 0.07, N = 3SE +/- 0.00, N = 325.5125.5025.61
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Apache Compilation 2.4.41Time To Compile123612182430Min: 25.47 / Avg: 25.51 / Max: 25.55Min: 25.41 / Avg: 25.5 / Max: 25.64Min: 25.6 / Avg: 25.61 / Max: 25.61

SVT-AV1

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-AV1 CPU-based multi-threaded video encoder for the AV1 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8Encoder Mode: Enc Mode 4 - Input: 1080p1230.89531.79062.68593.58124.4765SE +/- 0.011, N = 3SE +/- 0.010, N = 3SE +/- 0.012, N = 33.9633.9673.9791. (CXX) g++ options: -O3 -fcommon -fPIE -fPIC -pie
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8Encoder Mode: Enc Mode 4 - Input: 1080p123246810Min: 3.95 / Avg: 3.96 / Max: 3.98Min: 3.95 / Avg: 3.97 / Max: 3.98Min: 3.96 / Avg: 3.98 / Max: 41. (CXX) g++ options: -O3 -fcommon -fPIE -fPIC -pie

NAMD

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

OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.14ATPase Simulation - 327,506 Atoms1230.35970.71941.07911.43881.7985SE +/- 0.00295, N = 3SE +/- 0.00324, N = 3SE +/- 0.00436, N = 31.598551.592261.59601
OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.14ATPase Simulation - 327,506 Atoms123246810Min: 1.59 / Avg: 1.6 / Max: 1.6Min: 1.59 / Avg: 1.59 / Max: 1.6Min: 1.59 / Avg: 1.6 / Max: 1.6

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: 10012330K60K90K120K150KSE +/- 72.07, N = 3SE +/- 196.94, N = 3SE +/- 207.96, N = 31221101222991218541. (CXX) g++ options: -fPIC -O3 -rdynamic -lboost_system -lboost_thread -lboost_filesystem -lboost_chrono -lboost_date_time -lboost_atomic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas
OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: GoogleNet - Acceleration: CPU - Iterations: 10012320K40K60K80K100KMin: 122001 / Avg: 122109.67 / Max: 122246Min: 122009 / Avg: 122299.33 / Max: 122675Min: 121445 / Avg: 121853.67 / Max: 1221251. (CXX) g++ options: -fPIC -O3 -rdynamic -lboost_system -lboost_thread -lboost_filesystem -lboost_chrono -lboost_date_time -lboost_atomic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

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 1080p123110220330440550SE +/- 0.66, N = 3SE +/- 1.82, N = 3SE +/- 1.11, N = 3496.04495.75494.28MIN: 384.41 / MAX: 542.46MIN: 384.94 / MAX: 544.39MIN: 380.44 / MAX: 541.51. (CC) gcc options: -pthread
OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Summer Nature 1080p12390180270360450Min: 494.8 / Avg: 496.04 / Max: 497.05Min: 492.1 / Avg: 495.75 / Max: 497.62Min: 492.36 / Avg: 494.28 / Max: 496.191. (CC) gcc options: -pthread

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: bf16bf16bf16 - Engine: CPU1233691215SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 312.5812.5912.63MIN: 12.36MIN: 12.37MIN: 12.431. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU12348121620Min: 12.58 / Avg: 12.58 / Max: 12.58Min: 12.59 / Avg: 12.59 / Max: 12.61Min: 12.6 / Avg: 12.63 / Max: 12.651. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenVKL

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

OpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 0.9Benchmark: vklBenchmark1234080120160200SE +/- 0.76, N = 3SE +/- 1.33, N = 3SE +/- 0.45, N = 3175.67176.28176.14MIN: 1 / MAX: 755MIN: 1 / MAX: 749MIN: 1 / MAX: 759
OpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 0.9Benchmark: vklBenchmark123306090120150Min: 174.17 / Avg: 175.67 / Max: 176.58Min: 173.83 / Avg: 176.28 / Max: 178.42Min: 175.25 / Avg: 176.14 / Max: 176.67

Basis Universal

Basis Universal is a GPU texture codoec. This test times how long it takes to convert sRGB PNGs into Basis Univeral assets with various settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.12Settings: UASTC Level 0123246810SE +/- 0.014, N = 3SE +/- 0.010, N = 3SE +/- 0.008, N = 37.9567.9827.9721. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.12Settings: UASTC Level 01233691215Min: 7.94 / Avg: 7.96 / Max: 7.98Min: 7.96 / Avg: 7.98 / Max: 8Min: 7.96 / Avg: 7.97 / Max: 7.981. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.9.0Binary: Pathtracer ISPC - Model: Asian Dragon Obj12348121620SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 315.7015.6615.65MIN: 15.6 / MAX: 15.88MIN: 15.58 / MAX: 15.85MIN: 15.57 / MAX: 15.82
OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.9.0Binary: Pathtracer ISPC - Model: Asian Dragon Obj12348121620Min: 15.65 / Avg: 15.7 / Max: 15.75Min: 15.64 / Avg: 15.66 / Max: 15.68Min: 15.62 / Avg: 15.65 / Max: 15.68

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_pyaes12320406080100SE +/- 0.13, N = 3SE +/- 0.12, N = 3SE +/- 0.10, N = 397.797.697.9
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: crypto_pyaes12320406080100Min: 97.4 / Avg: 97.67 / Max: 97.8Min: 97.4 / Avg: 97.57 / Max: 97.8Min: 97.7 / Avg: 97.9 / Max: 98

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: bf16bf16bf16 - Engine: CPU1230.63821.27641.91462.55283.191SE +/- 0.00103, N = 3SE +/- 0.00212, N = 3SE +/- 0.02581, N = 32.828632.836612.82811MIN: 2.68MIN: 2.68MIN: 2.651. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU123246810Min: 2.83 / Avg: 2.83 / Max: 2.83Min: 2.83 / Avg: 2.84 / Max: 2.84Min: 2.79 / Avg: 2.83 / Max: 2.881. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

AOM AV1

This is a simple test of the AOMedia AV1 encoder run on the CPU with a sample video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 2.0Encoder Mode: Speed 6 Two-Pass1230.7921.5842.3763.1683.96SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 33.523.523.511. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 2.0Encoder Mode: Speed 6 Two-Pass123246810Min: 3.51 / Avg: 3.52 / Max: 3.52Min: 3.52 / Avg: 3.52 / Max: 3.53Min: 3.49 / Avg: 3.51 / Max: 3.531. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -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 Leukocyte12320406080100SE +/- 0.16, N = 3SE +/- 0.17, N = 3SE +/- 0.15, N = 3109.68109.64109.371. (CXX) g++ options: -O2 -lOpenCL
OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP Leukocyte12320406080100Min: 109.5 / Avg: 109.68 / Max: 110Min: 109.34 / Avg: 109.64 / Max: 109.94Min: 109.22 / Avg: 109.37 / Max: 109.681. (CXX) g++ options: -O2 -lOpenCL

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: 10012310K20K30K40K50KSE +/- 101.10, N = 3SE +/- 130.16, N = 3SE +/- 74.48, N = 34797148100480191. (CXX) g++ options: -fPIC -O3 -rdynamic -lboost_system -lboost_thread -lboost_filesystem -lboost_chrono -lboost_date_time -lboost_atomic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas
OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: AlexNet - Acceleration: CPU - Iterations: 1001238K16K24K32K40KMin: 47808 / Avg: 47970.67 / Max: 48156Min: 47941 / Avg: 48100 / Max: 48358Min: 47891 / Avg: 48019.33 / Max: 481491. (CXX) g++ options: -fPIC -O3 -rdynamic -lboost_system -lboost_thread -lboost_filesystem -lboost_chrono -lboost_date_time -lboost_atomic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

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 Protein123246810SE +/- 0.021, N = 3SE +/- 0.008, N = 3SE +/- 0.012, N = 37.5107.5147.4941. (CXX) g++ options: -O3 -pthread -lm
OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 24Aug2020Model: Rhodopsin Protein1233691215Min: 7.47 / Avg: 7.51 / Max: 7.54Min: 7.5 / Avg: 7.51 / Max: 7.52Min: 7.48 / Avg: 7.49 / Max: 7.521. (CXX) g++ options: -O3 -pthread -lm

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 Search123306090120150SE +/- 0.11, N = 3SE +/- 0.28, N = 3SE +/- 0.27, N = 3133.45133.26133.101. (CC) gcc options: -O3 -pthread -lhmmer -leasel -lm
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed HMMer Search 3.3.1Pfam Database Search123306090120150Min: 133.23 / Avg: 133.45 / Max: 133.59Min: 132.88 / Avg: 133.26 / Max: 133.82Min: 132.62 / Avg: 133.1 / Max: 133.571. (CC) gcc options: -O3 -pthread -lhmmer -leasel -lm

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-v2-50123816243240SE +/- 0.09, N = 3SE +/- 0.14, N = 3SE +/- 0.07, N = 335.6335.6035.54MIN: 35.04 / MAX: 97.88MIN: 35.19 / MAX: 80.6MIN: 35.2 / MAX: 71.131. (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: resnet-v2-50123816243240Min: 35.47 / Avg: 35.63 / Max: 35.78Min: 35.36 / Avg: 35.6 / Max: 35.84Min: 35.41 / Avg: 35.54 / Max: 35.651. (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

RawTherapee

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

OpenBenchmarking.orgSeconds, Fewer Is BetterRawTherapeeTotal Benchmark Time1231224364860SE +/- 0.09, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 353.9553.8653.811. RawTherapee, version 5.5, command line. An advanced, cross-platform program for developing raw photos. Website: http://www.rawtherapee.com/ Documentation: http://rawpedia.rawtherapee.com/ Forum: https://discuss.pixls.us/c/software/rawtherapee Code and bug reports: https://github.com/Beep6581/RawTherapee Symbols: <Chevrons> indicate parameters you can change. [Square brackets] mean the parameter is optional. The pipe symbol | indicates a choice of one or the other. The dash symbol - denotes a range of possible values from one to the other. Usage: rawtherapee-cli -c <dir>|<files> Convert files in batch with default parameters. rawtherapee-cli <other options> -c <dir>|<files> Convert files in batch with your own settings. Options: rawtherapee-cli[-o <output>|-O <output>] [-q] [-a] [-s|-S] [-p <one.pp3> [-p <two.pp3> ...] ] [-d] [ -j[1-100] -js<1-3> | -t[z] -b<8|16|16f|32> | -n -b<8|16> ] [-Y] [-f] -c <input> -c <files> Specify one or more input files or folders. When specifying folders, Rawtherapee will look for image file types which comply with the selected extensions (see also '-a'). -c must be the last option. -o <file>|<dir> Set output file or folder. Saves output file alongside input file if -o is not specified. -O <file>|<dir> Set output file or folder and copy pp3 file into it. Saves output file alongside input file if -O is not specified. -q Quick-start mode. Does not load cached files to speedup start time. -a Process all supported image file types when specifying a folder, even those not currently selected in Preferences > File Browser > Parsed Extensions. -s Use the existing sidecar file to build the processing parameters, e.g. for photo.raw there should be a photo.raw.pp3 file in the same folder. If the sidecar file does not exist, neutral values will be used. -S Like -s but skip if the sidecar file does not exist. -p <file.pp3> Specify processing profile to be used for all conversions. You can specify as many sets of "-p <file.pp3>" options as you like, each will be built on top of the previous one, as explained below. -d Use the default raw or non-raw processing profile as set in Preferences > Image Processing > Default Processing Profile -j[1-100] Specify output to be JPEG (default, if -t and -n are not set). Optionally, specify compression 1-100 (default value: 92). -js<1-3> Specify the JPEG chroma subsampling parameter, where: 1 = Best compression: 2x2, 1x1, 1x1 (4:2:0) Chroma halved vertically and horizontally. 2 = Balanced (default): 2x1, 1x1, 1x1 (4:2:2) Chroma halved horizontally. 3 = Best quality: 1x1, 1x1, 1x1 (4:4:4) No chroma subsampling. -b<8|16|16f|32> Specify bit depth per channel. 8 = 8-bit integer. Applies to JPEG, PNG and TIFF. Default for JPEG and PNG. 16 = 16-bit integer. Applies to TIFF and PNG. Default for TIFF. 16f = 16-bit float. Applies to TIFF. 32 = 32-bit float. Applies to TIFF. -t[z] Specify output to be TIFF. Uncompressed by default, or deflate compression with 'z'. -n Specify output to be compressed PNG. Compression is hard-coded to PNG_FILTER_PAETH, Z_RLE. -Y Overwrite output if present. -f Use the custom fast-export processing pipeline. Your pp3 files can be incomplete, RawTherapee will build the final values as follows: 1- A new processing profile is created using neutral values, 2- If the "-d" option is set, the values are overridden by those found in the default raw or non-raw processing profile. 3- If one or more "-p" options are set, the values are overridden by those found in these processing profiles. 4- If the "-s" or "-S" options are set, the values are finally overridden by those found in the sidecar files. The processing profiles are processed in the order specified on the command line.
OpenBenchmarking.orgSeconds, Fewer Is BetterRawTherapeeTotal Benchmark Time1231122334455Min: 53.78 / Avg: 53.95 / Max: 54.1Min: 53.81 / Avg: 53.86 / Max: 53.94Min: 53.78 / Avg: 53.81 / Max: 53.841. RawTherapee, version 5.5, command line. An advanced, cross-platform program for developing raw photos. Website: http://www.rawtherapee.com/ Documentation: http://rawpedia.rawtherapee.com/ Forum: https://discuss.pixls.us/c/software/rawtherapee Code and bug reports: https://github.com/Beep6581/RawTherapee Symbols: <Chevrons> indicate parameters you can change. [Square brackets] mean the parameter is optional. The pipe symbol | indicates a choice of one or the other. The dash symbol - denotes a range of possible values from one to the other. Usage: rawtherapee-cli -c <dir>|<files> Convert files in batch with default parameters. rawtherapee-cli <other options> -c <dir>|<files> Convert files in batch with your own settings. Options: rawtherapee-cli[-o <output>|-O <output>] [-q] [-a] [-s|-S] [-p <one.pp3> [-p <two.pp3> ...] ] [-d] [ -j[1-100] -js<1-3> | -t[z] -b<8|16|16f|32> | -n -b<8|16> ] [-Y] [-f] -c <input> -c <files> Specify one or more input files or folders. When specifying folders, Rawtherapee will look for image file types which comply with the selected extensions (see also '-a'). -c must be the last option. -o <file>|<dir> Set output file or folder. Saves output file alongside input file if -o is not specified. -O <file>|<dir> Set output file or folder and copy pp3 file into it. Saves output file alongside input file if -O is not specified. -q Quick-start mode. Does not load cached files to speedup start time. -a Process all supported image file types when specifying a folder, even those not currently selected in Preferences > File Browser > Parsed Extensions. -s Use the existing sidecar file to build the processing parameters, e.g. for photo.raw there should be a photo.raw.pp3 file in the same folder. If the sidecar file does not exist, neutral values will be used. -S Like -s but skip if the sidecar file does not exist. -p <file.pp3> Specify processing profile to be used for all conversions. You can specify as many sets of "-p <file.pp3>" options as you like, each will be built on top of the previous one, as explained below. -d Use the default raw or non-raw processing profile as set in Preferences > Image Processing > Default Processing Profile -j[1-100] Specify output to be JPEG (default, if -t and -n are not set). Optionally, specify compression 1-100 (default value: 92). -js<1-3> Specify the JPEG chroma subsampling parameter, where: 1 = Best compression: 2x2, 1x1, 1x1 (4:2:0) Chroma halved vertically and horizontally. 2 = Balanced (default): 2x1, 1x1, 1x1 (4:2:2) Chroma halved horizontally. 3 = Best quality: 1x1, 1x1, 1x1 (4:4:4) No chroma subsampling. -b<8|16|16f|32> Specify bit depth per channel. 8 = 8-bit integer. Applies to JPEG, PNG and TIFF. Default for JPEG and PNG. 16 = 16-bit integer. Applies to TIFF and PNG. Default for TIFF. 16f = 16-bit float. Applies to TIFF. 32 = 32-bit float. Applies to TIFF. -t[z] Specify output to be TIFF. Uncompressed by default, or deflate compression with 'z'. -n Specify output to be compressed PNG. Compression is hard-coded to PNG_FILTER_PAETH, Z_RLE. -Y Overwrite output if present. -f Use the custom fast-export processing pipeline. Your pp3 files can be incomplete, RawTherapee will build the final values as follows: 1- A new processing profile is created using neutral values, 2- If the "-d" option is set, the values are overridden by those found in the default raw or non-raw processing profile. 3- If one or more "-p" options are set, the values are overridden by those found in these processing profiles. 4- If the "-s" or "-S" options are set, the values are finally overridden by those found in the sidecar files. The processing profiles are processed in the order specified on the command line.

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: 20012320K40K60K80K100KSE +/- 100.00, N = 3SE +/- 34.60, N = 3SE +/- 123.16, N = 39623296039959911. (CXX) g++ options: -fPIC -O3 -rdynamic -lboost_system -lboost_thread -lboost_filesystem -lboost_chrono -lboost_date_time -lboost_atomic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas
OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: AlexNet - Acceleration: CPU - Iterations: 20012320K40K60K80K100KMin: 96037 / Avg: 96232 / Max: 96368Min: 96000 / Avg: 96039 / Max: 96108Min: 95759 / Avg: 95990.67 / Max: 961791. (CXX) g++ options: -fPIC -O3 -rdynamic -lboost_system -lboost_thread -lboost_filesystem -lboost_chrono -lboost_date_time -lboost_atomic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: BMW27 - Compute: CPU-Only123306090120150SE +/- 0.15, N = 3SE +/- 0.28, N = 3SE +/- 0.04, N = 3151.14151.51151.24
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: BMW27 - Compute: CPU-Only123306090120150Min: 150.94 / Avg: 151.14 / Max: 151.42Min: 150.95 / Avg: 151.51 / Max: 151.84Min: 151.17 / Avg: 151.24 / Max: 151.31

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: bf16bf16bf16 - Engine: CPU12348121620SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 314.5214.5514.51MIN: 14.31MIN: 14.29MIN: 14.311. (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: bf16bf16bf16 - Engine: CPU12348121620Min: 14.5 / Avg: 14.52 / Max: 14.57Min: 14.49 / Avg: 14.55 / Max: 14.65Min: 14.5 / Avg: 14.51 / Max: 14.541. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

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 4.0.0Time To OCR 7 Images123612182430SE +/- 0.04, N = 3SE +/- 0.14, N = 3SE +/- 0.02, N = 326.6426.7026.63
OpenBenchmarking.orgSeconds, Fewer Is BetterTesseract OCR 4.0.0Time To OCR 7 Images123612182430Min: 26.56 / Avg: 26.64 / Max: 26.68Min: 26.42 / Avg: 26.69 / Max: 26.84Min: 26.6 / Avg: 26.63 / Max: 26.66

AOM AV1

This is a simple test of the AOMedia AV1 encoder run on the CPU with a sample video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 2.0Encoder Mode: Speed 6 Realtime12348121620SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 317.8917.9217.931. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 2.0Encoder Mode: Speed 6 Realtime123510152025Min: 17.82 / Avg: 17.89 / Max: 17.95Min: 17.89 / Avg: 17.92 / Max: 17.97Min: 17.88 / Avg: 17.93 / Max: 17.981. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -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 1080p123130260390520650SE +/- 1.75, N = 3SE +/- 1.88, N = 3SE +/- 3.14, N = 3610.06610.40609.04MIN: 465.77 / MAX: 779.74MIN: 465.82 / MAX: 795.4MIN: 463.93 / MAX: 783.911. (CC) gcc options: -pthread
OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Chimera 1080p123110220330440550Min: 608.17 / Avg: 610.06 / Max: 613.56Min: 606.83 / Avg: 610.4 / Max: 613.21Min: 605.54 / Avg: 609.04 / Max: 615.31. (CC) gcc options: -pthread

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Fishy Cat - Compute: CPU-Only12350100150200250SE +/- 0.17, N = 3SE +/- 0.27, N = 3SE +/- 0.06, N = 3208.93209.39209.24
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Fishy Cat - Compute: CPU-Only1234080120160200Min: 208.6 / Avg: 208.93 / Max: 209.17Min: 209 / Avg: 209.39 / Max: 209.92Min: 209.12 / Avg: 209.24 / Max: 209.32

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Classroom - Compute: CPU-Only123100200300400500SE +/- 0.93, N = 3SE +/- 0.31, N = 3SE +/- 0.76, N = 3451.27452.25451.81
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Classroom - Compute: CPU-Only12380160240320400Min: 449.76 / Avg: 451.27 / Max: 452.95Min: 451.64 / Avg: 452.25 / Max: 452.62Min: 450.34 / Avg: 451.81 / Max: 452.86

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 Nanotube12350100150200250SE +/- 0.71, N = 3SE +/- 0.35, N = 3SE +/- 0.28, N = 3249.72249.96250.261. (CC) gcc options: -pthread -shared -lxc -lblas -lmpi
OpenBenchmarking.orgSeconds, Fewer Is BetterGPAW 20.1Input: Carbon Nanotube12350100150200250Min: 248.8 / Avg: 249.72 / Max: 251.11Min: 249.33 / Avg: 249.96 / Max: 250.54Min: 249.72 / Avg: 250.26 / Max: 250.631. (CC) gcc options: -pthread -shared -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: django_template1231122334455SE +/- 0.09, N = 3SE +/- 0.06, N = 3SE +/- 0.03, N = 346.947.047.0
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: django_template1231020304050Min: 46.8 / Avg: 46.93 / Max: 47.1Min: 46.9 / Avg: 47 / Max: 47.1Min: 46.9 / Avg: 46.97 / Max: 47

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: CPU1230.33090.66180.99271.32361.6545SE +/- 0.00183, N = 3SE +/- 0.00089, N = 3SE +/- 0.00104, N = 31.468601.470761.46784MIN: 1.45MIN: 1.46MIN: 1.451. (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: u8s8f32 - Engine: CPU123246810Min: 1.47 / Avg: 1.47 / Max: 1.47Min: 1.47 / Avg: 1.47 / Max: 1.47Min: 1.47 / Avg: 1.47 / Max: 1.471. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -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 V4123600K1200K1800K2400K3000KSE +/- 4029.11, N = 3SE +/- 2935.10, N = 3SE +/- 2268.55, N = 3283685028326072831243
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Inception V4123500K1000K1500K2000K2500KMin: 2829510 / Avg: 2836850 / Max: 2843400Min: 2827200 / Avg: 2832606.67 / Max: 2837290Min: 2826710 / Avg: 2831243.33 / Max: 2833670

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, Lossless12348121620SE +/- 0.03, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 317.9117.8817.921. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -lpng16 -ljpeg
OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Lossless123510152025Min: 17.86 / Avg: 17.91 / Max: 17.95Min: 17.88 / Avg: 17.88 / Max: 17.89Min: 17.85 / Avg: 17.92 / Max: 17.961. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -lpng16 -ljpeg

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 v212370140210280350SE +/- 0.40, N = 3SE +/- 0.24, N = 3SE +/- 0.04, N = 3315.68315.40315.08MIN: 314.48 / MAX: 332.45MIN: 314.51 / MAX: 326.8MIN: 314.62 / MAX: 316.041. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -O3 -rdynamic -ldl
OpenBenchmarking.orgms, Fewer Is BetterTNN 0.2.3Target: CPU - Model: MobileNet v212360120180240300Min: 315.04 / Avg: 315.68 / Max: 316.4Min: 315.08 / Avg: 315.4 / Max: 315.88Min: 315 / Avg: 315.08 / Max: 315.121. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -O3 -rdynamic -ldl

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-v1-1.01230.94971.89942.84913.79884.7485SE +/- 0.008, N = 3SE +/- 0.014, N = 3SE +/- 0.017, N = 34.2134.2214.215MIN: 4.13 / MAX: 5.67MIN: 4.14 / MAX: 5.06MIN: 4.12 / MAX: 8.761. (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: mobilenet-v1-1.0123246810Min: 4.2 / Avg: 4.21 / Max: 4.23Min: 4.2 / Avg: 4.22 / Max: 4.25Min: 4.18 / Avg: 4.22 / Max: 4.241. (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

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: 1000123100K200K300K400K500KSE +/- 159.61, N = 3SE +/- 301.22, N = 3SE +/- 305.87, N = 34786044783534791691. (CXX) g++ options: -fPIC -O3 -rdynamic -lboost_system -lboost_thread -lboost_filesystem -lboost_chrono -lboost_date_time -lboost_atomic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas
OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: AlexNet - Acceleration: CPU - Iterations: 100012380K160K240K320K400KMin: 478415 / Avg: 478603.67 / Max: 478921Min: 477776 / Avg: 478352.67 / Max: 478792Min: 478649 / Avg: 479168.67 / Max: 4797081. (CXX) g++ options: -fPIC -O3 -rdynamic -lboost_system -lboost_thread -lboost_filesystem -lboost_chrono -lboost_date_time -lboost_atomic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

G'MIC

G'MIC is an open-source framework for image processing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterG'MICTest: 3D Elevated Function In Random Colors, 100 Times1231428425670SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.07, N = 363.1863.1963.291. Version 2.4.5, Copyright (c) 2008-2019, David Tschumperle.
OpenBenchmarking.orgSeconds, Fewer Is BetterG'MICTest: 3D Elevated Function In Random Colors, 100 Times1231224364860Min: 63.13 / Avg: 63.18 / Max: 63.22Min: 63.17 / Avg: 63.19 / Max: 63.21Min: 63.19 / Avg: 63.29 / Max: 63.421. Version 2.4.5, Copyright (c) 2008-2019, David Tschumperle.

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: NASNet Mobile12340K80K120K160K200KSE +/- 113.93, N = 3SE +/- 369.09, N = 3SE +/- 248.04, N = 3181809181669181512
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: NASNet Mobile12330K60K90K120K150KMin: 181619 / Avg: 181809.33 / Max: 182013Min: 181174 / Avg: 181669.33 / Max: 182391Min: 181126 / Avg: 181512.33 / Max: 181975

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Pabellon Barcelona - Compute: CPU-Only123110220330440550SE +/- 0.34, N = 3SE +/- 0.39, N = 3SE +/- 0.19, N = 3509.61508.87508.79
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Pabellon Barcelona - Compute: CPU-Only12390180270360450Min: 508.93 / Avg: 509.61 / Max: 509.96Min: 508.25 / Avg: 508.87 / Max: 509.6Min: 508.41 / Avg: 508.79 / Max: 509.02

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.9.0Binary: Pathtracer - Model: Asian Dragon Obj1233691215SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 312.8912.9112.90MIN: 12.85 / MAX: 13MIN: 12.87 / MAX: 13.02MIN: 12.82 / MAX: 13.01
OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.9.0Binary: Pathtracer - Model: Asian Dragon Obj12348121620Min: 12.88 / Avg: 12.89 / Max: 12.9Min: 12.9 / Avg: 12.91 / Max: 12.92Min: 12.85 / Avg: 12.9 / Max: 12.94

SVT-AV1

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-AV1 CPU-based multi-threaded video encoder for the AV1 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8Encoder Mode: Enc Mode 8 - Input: 1080p123816243240SE +/- 0.11, N = 3SE +/- 0.14, N = 3SE +/- 0.04, N = 333.3633.3433.311. (CXX) g++ options: -O3 -fcommon -fPIE -fPIC -pie
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8Encoder Mode: Enc Mode 8 - Input: 1080p123714212835Min: 33.24 / Avg: 33.36 / Max: 33.57Min: 33.07 / Avg: 33.34 / Max: 33.51Min: 33.24 / Avg: 33.31 / Max: 33.391. (CXX) g++ options: -O3 -fcommon -fPIE -fPIC -pie

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: 20012350K100K150K200K250KSE +/- 275.20, N = 3SE +/- 136.15, N = 3SE +/- 490.17, N = 32437962441542439571. (CXX) g++ options: -fPIC -O3 -rdynamic -lboost_system -lboost_thread -lboost_filesystem -lboost_chrono -lboost_date_time -lboost_atomic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas
OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: GoogleNet - Acceleration: CPU - Iterations: 20012340K80K120K160K200KMin: 243357 / Avg: 243796 / Max: 244303Min: 243889 / Avg: 244154.33 / Max: 244340Min: 243033 / Avg: 243956.67 / Max: 2447031. (CXX) g++ options: -fPIC -O3 -rdynamic -lboost_system -lboost_thread -lboost_filesystem -lboost_chrono -lboost_date_time -lboost_atomic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

Mlpack Benchmark

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

OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_svm12348121620SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 315.1615.1415.14
OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_svm12348121620Min: 15.14 / Avg: 15.16 / Max: 15.17Min: 15.13 / Avg: 15.14 / Max: 15.15Min: 15.12 / Avg: 15.14 / Max: 15.17

GNU Octave Benchmark

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

OpenBenchmarking.orgSeconds, Fewer Is BetterGNU Octave Benchmark 4.4.1123246810SE +/- 0.023, N = 5SE +/- 0.020, N = 5SE +/- 0.031, N = 58.1168.1268.118
OpenBenchmarking.orgSeconds, Fewer Is BetterGNU Octave Benchmark 4.4.11233691215Min: 8.07 / Avg: 8.12 / Max: 8.2Min: 8.07 / Avg: 8.13 / Max: 8.19Min: 8.07 / Avg: 8.12 / Max: 8.24

Basis Universal

Basis Universal is a GPU texture codoec. This test times how long it takes to convert sRGB PNGs into Basis Univeral assets with various settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.12Settings: UASTC Level 2123714212835SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 330.9930.9630.991. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.12Settings: UASTC Level 2123714212835Min: 30.95 / Avg: 30.99 / Max: 31.03Min: 30.94 / Avg: 30.96 / Max: 30.99Min: 30.99 / Avg: 30.99 / Max: 30.991. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread

Git

This test measures the time needed to carry out some sample Git operations on an example, static repository that happens to be a copy of the GNOME GTK tool-kit repository. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGitTime To Complete Common Git Commands1231122334455SE +/- 0.03, N = 3SE +/- 0.08, N = 3SE +/- 0.06, N = 350.9450.9750.911. git version 2.20.1
OpenBenchmarking.orgSeconds, Fewer Is BetterGitTime To Complete Common Git Commands1231020304050Min: 50.88 / Avg: 50.94 / Max: 50.98Min: 50.87 / Avg: 50.97 / Max: 51.13Min: 50.79 / Avg: 50.91 / Max: 511. git version 2.20.1

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 21239M18M27M36M45MSE +/- 41963.68, N = 3SE +/- 123890.70, N = 3SE +/- 18764.26, N = 341673100.141712858.141700019.8
OpenBenchmarking.orgLPS, More Is BetterBYTE Unix Benchmark 3.6Computational Test: Dhrystone 21237M14M21M28M35MMin: 41625344 / Avg: 41673100.07 / Max: 41756747.4Min: 41468122.8 / Avg: 41712858.07 / Max: 41868769.7Min: 41679916.5 / Avg: 41700019.83 / Max: 41737515.7

Basis Universal

Basis Universal is a GPU texture codoec. This test times how long it takes to convert sRGB PNGs into Basis Univeral assets with various settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.12Settings: UASTC Level 31231326395265SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 358.7258.7658.781. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.12Settings: UASTC Level 31231224364860Min: 58.7 / Avg: 58.72 / Max: 58.77Min: 58.74 / Avg: 58.76 / Max: 58.78Min: 58.77 / Avg: 58.78 / Max: 58.781. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread

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 Quant12330K60K90K120K150KSE +/- 53.26, N = 3SE +/- 46.16, N = 3SE +/- 32.25, N = 3136609136479136482
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet Quant12320K40K60K80K100KMin: 136503 / Avg: 136609.33 / Max: 136668Min: 136424 / Avg: 136479.33 / Max: 136571Min: 136418 / Avg: 136482 / Max: 136521

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-bit12320406080100SE +/- 0.10, N = 3SE +/- 0.08, N = 3SE +/- 0.05, N = 386.7486.7386.67MIN: 57.94 / MAX: 211.3MIN: 57.89 / MAX: 206.91MIN: 57.82 / MAX: 213.791. (CC) gcc options: -pthread
OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Chimera 1080p 10-bit1231632486480Min: 86.62 / Avg: 86.74 / Max: 86.93Min: 86.57 / Avg: 86.73 / Max: 86.86Min: 86.57 / Avg: 86.67 / Max: 86.731. (CC) gcc options: -pthread

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: bf16bf16bf16 - Engine: CPU12320406080100SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 395.1795.1495.10MIN: 94.26MIN: 94.32MIN: 94.271. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: IP Batch All - Data Type: bf16bf16bf16 - Engine: CPU12320406080100Min: 95.11 / Avg: 95.17 / Max: 95.21Min: 95.08 / Avg: 95.14 / Max: 95.18Min: 95.07 / Avg: 95.1 / Max: 95.151. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Barbershop - Compute: CPU-Only123130260390520650SE +/- 0.50, N = 3SE +/- 0.34, N = 3SE +/- 0.33, N = 3605.44605.36605.01
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Barbershop - Compute: CPU-Only123110220330440550Min: 604.47 / Avg: 605.44 / Max: 606.14Min: 604.96 / Avg: 605.36 / Max: 606.04Min: 604.37 / Avg: 605.01 / Max: 605.5

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: SqueezeNet12340K80K120K160K200KSE +/- 377.66, N = 3SE +/- 337.48, N = 3SE +/- 359.21, N = 3191171191148191041
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: SqueezeNet12330K60K90K120K150KMin: 190584 / Avg: 191170.67 / Max: 191876Min: 190524 / Avg: 191147.67 / Max: 191683Min: 190442 / Avg: 191041 / Max: 191684

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 v1.112360120180240300SE +/- 0.15, N = 3SE +/- 0.22, N = 3SE +/- 0.09, N = 3296.37296.42296.31MIN: 295.7 / MAX: 311.12MIN: 295.78 / MAX: 308.68MIN: 295.84 / MAX: 301.861. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -O3 -rdynamic -ldl
OpenBenchmarking.orgms, Fewer Is BetterTNN 0.2.3Target: CPU - Model: SqueezeNet v1.112350100150200250Min: 296.09 / Avg: 296.37 / Max: 296.62Min: 296.19 / Avg: 296.42 / Max: 296.85Min: 296.18 / Avg: 296.31 / Max: 296.51. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -O3 -rdynamic -ldl

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: 1000123300K600K900K1200K1500KSE +/- 1694.74, N = 3SE +/- 576.40, N = 3SE +/- 638.13, N = 31220000122013012203331. (CXX) g++ options: -fPIC -O3 -rdynamic -lboost_system -lboost_thread -lboost_filesystem -lboost_chrono -lboost_date_time -lboost_atomic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas
OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: GoogleNet - Acceleration: CPU - Iterations: 1000123200K400K600K800K1000KMin: 1217680 / Avg: 1220000 / Max: 1223300Min: 1219260 / Avg: 1220130 / Max: 1221220Min: 1219370 / Avg: 1220333.33 / Max: 12215401. (CXX) g++ options: -fPIC -O3 -rdynamic -lboost_system -lboost_thread -lboost_filesystem -lboost_chrono -lboost_date_time -lboost_atomic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

Basis Universal

Basis Universal is a GPU texture codoec. This test times how long it takes to convert sRGB PNGs into Basis Univeral assets with various settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.12Settings: UASTC Level 2 + RDO Post-Processing123160320480640800SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.12, N = 3721.82721.87721.681. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.12Settings: UASTC Level 2 + RDO Post-Processing123130260390520650Min: 721.79 / Avg: 721.82 / Max: 721.84Min: 721.78 / Avg: 721.87 / Max: 721.97Min: 721.45 / Avg: 721.68 / Max: 721.851. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread

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 1.2.41600K1200K1800K2400K3000KSE +/- 49600.17, N = 929140521. (CXX) g++ options: -O3 -fopenmp

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_startup1233691215SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 310.510.510.5
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: python_startup1233691215Min: 10.4 / Avg: 10.47 / Max: 10.5Min: 10.4 / Avg: 10.47 / Max: 10.5Min: 10.5 / Avg: 10.53 / Max: 10.6

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: regex_compile123306090120150SE +/- 0.58, N = 3SE +/- 0.33, N = 3154154154
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: regex_compile123306090120150Min: 153 / Avg: 154 / Max: 155Min: 153 / Avg: 153.67 / Max: 154

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: raytrace12390180270360450SE +/- 0.67, N = 3408408408
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: raytrace12370140210280350Min: 407 / Avg: 407.67 / Max: 409

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: pathlib12348121620SE +/- 0.00, N = 3SE +/- 0.09, N = 3SE +/- 0.03, N = 317.817.817.8
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: pathlib123510152025Min: 17.8 / Avg: 17.8 / Max: 17.8Min: 17.6 / Avg: 17.77 / Max: 17.9Min: 17.7 / Avg: 17.77 / Max: 17.8

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: go12350100150200250SE +/- 0.67, N = 3213213213
OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: go1234080120160200Min: 212 / Avg: 212.67 / Max: 214

SVT-AV1

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-AV1 CPU-based multi-threaded video encoder for the AV1 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8Encoder Mode: Enc Mode 0 - Input: 1080p1230.02990.05980.08970.11960.1495SE +/- 0.000, N = 3SE +/- 0.000, N = 3SE +/- 0.000, N = 30.1330.1330.1331. (CXX) g++ options: -O3 -fcommon -fPIE -fPIC -pie
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8Encoder Mode: Enc Mode 0 - Input: 1080p12312345Min: 0.13 / Avg: 0.13 / Max: 0.13Min: 0.13 / Avg: 0.13 / Max: 0.13Min: 0.13 / Avg: 0.13 / Max: 0.131. (CXX) g++ options: -O3 -fcommon -fPIE -fPIC -pie

AOM AV1

This is a simple test of the AOMedia AV1 encoder run on the CPU with a sample video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 2.0Encoder Mode: Speed 0 Two-Pass1230.06980.13960.20940.27920.349SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.310.310.311. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 2.0Encoder Mode: Speed 0 Two-Pass12312345Min: 0.31 / Avg: 0.31 / Max: 0.31Min: 0.31 / Avg: 0.31 / Max: 0.31Min: 0.31 / Avg: 0.31 / Max: 0.311. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

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 Network1239001800270036004500SE +/- 129.03, N = 15SE +/- 128.33, N = 12SE +/- 67.44, N = 124393426743011. (CXX) g++ options: -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -ldl -lm -lpthread -lrt
OpenBenchmarking.orgms, Fewer Is BetterOpenCV 4.4Test: DNN - Deep Neural Network1238001600240032004000Min: 3448 / Avg: 4393.27 / Max: 5157Min: 3475 / Avg: 4266.75 / Max: 4984Min: 3927 / Avg: 4300.92 / Max: 46381. (CXX) g++ options: -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -ldl -lm -lpthread -lrt

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: CPU1230.7451.492.2352.983.725SE +/- 0.00204, N = 3SE +/- 0.07364, N = 14SE +/- 0.00502, N = 33.175193.311213.17711MIN: 3.07MIN: 3.1MIN: 3.081. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 1.5Harness: IP Batch 1D - Data Type: f32 - Engine: CPU123246810Min: 3.17 / Avg: 3.18 / Max: 3.18Min: 3.19 / Avg: 3.31 / Max: 4.26Min: 3.17 / Avg: 3.18 / Max: 3.191. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

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 Tree1232K4K6K8K10KSE +/- 169.48, N = 15SE +/- 115.84, N = 5SE +/- 97.43, N = 510146.389866.1310202.48
OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Akka Unbalanced Cobwebbed Tree1232K4K6K8K10KMin: 9295.88 / Avg: 10146.38 / Max: 12074.41Min: 9523.31 / Avg: 9866.13 / Max: 10087Min: 9989.97 / Avg: 10202.48 / Max: 10499.16

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Apache Spark Bayes12330060090012001500SE +/- 18.43, N = 25SE +/- 18.56, N = 25SE +/- 13.83, N = 51531.731534.531526.99
OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.10.0Test: Apache Spark Bayes12330060090012001500Min: 1428.74 / Avg: 1531.73 / Max: 1819.69Min: 1442.57 / Avg: 1534.53 / Max: 1868.17Min: 1477.68 / Avg: 1526.99 / Max: 1555.52

170 Results Shown

Algebraic Multi-Grid Benchmark
eSpeak-NG Speech Engine
Java Gradle Build
oneDNN
NCNN
Rodinia
oneDNN
NeatBench
NCNN
Renaissance
Rodinia
oneDNN
Renaissance:
  Genetic Algorithm Using Jenetics + Futures
  Rand Forest
oneDNN
DaCapo Benchmark
LULESH
Renaissance
oneDNN
DaCapo Benchmark
NCNN
Renaissance
OCRMyPDF
NCNN
YafaRay
NCNN
Timed MAFFT Alignment
NCNN
Renaissance
NCNN:
  CPU - efficientnet-b0
  CPU - yolov4-tiny
oneDNN
PyPerformance
DaCapo Benchmark
oneDNN
LuxCoreRender
RNNoise
NCNN
oneDNN
Mlpack Benchmark
NCNN
Renaissance
Embree
oneDNN
G'MIC
Mobile Neural Network
Timed Linux Kernel Compilation
PyPerformance
DaCapo Benchmark
PyPerformance
NCNN
PyPerformance
Embree
LAMMPS Molecular Dynamics Simulator
oneDNN
OpenVKL
NCNN
libavif avifenc
WebP Image Encode
SVT-VP9
Renaissance
Embree
NCNN
oneDNN
KeyDB
WebP Image Encode
Intel Open Image Denoise
BRL-CAD
Hugin
NCNN
Rodinia
C-Blosc
Embree
libavif avifenc
WebP Image Encode
libavif avifenc
PyPerformance
Zstd Compression
SVT-VP9
Mobile Neural Network
G'MIC
Timed LLVM Compilation
libavif avifenc
oneDNN
Basis Universal
Zstd Compression
Rodinia
InfluxDB
Mlpack Benchmark
Mobile Neural Network
WebP Image Encode
LuxCoreRender
oneDNN
dav1d
PyPerformance
AOM AV1
Hierarchical INTegration
Mlpack Benchmark
SVT-VP9
oneDNN
AOM AV1
TensorFlow Lite
LibRaw
TensorFlow Lite
Timed Apache Compilation
SVT-AV1
NAMD
Caffe
dav1d
oneDNN
OpenVKL
Basis Universal
Embree
PyPerformance
oneDNN
AOM AV1
Rodinia
Caffe
LAMMPS Molecular Dynamics Simulator
Timed HMMer Search
Mobile Neural Network
RawTherapee
Caffe
Blender
oneDNN
Tesseract OCR
AOM AV1
dav1d
Blender:
  Fishy Cat - CPU-Only
  Classroom - CPU-Only
GPAW
PyPerformance
oneDNN
TensorFlow Lite
WebP Image Encode
TNN
Mobile Neural Network
Caffe
G'MIC
TensorFlow Lite
Blender
Embree
SVT-AV1
Caffe
Mlpack Benchmark
GNU Octave Benchmark
Basis Universal
Git
BYTE Unix Benchmark
Basis Universal
TensorFlow Lite
dav1d
oneDNN
Blender
TensorFlow Lite
TNN
Caffe
Basis Universal
Kripke
PyPerformance:
  python_startup
  regex_compile
  raytrace
  pathlib
  go
SVT-AV1
AOM AV1
OpenCV
oneDNN
Renaissance:
  Akka Unbalanced Cobwebbed Tree
  Apache Spark Bayes

1

Compiler Notes: --disable-multilib --enable-checking=release --enable-languages=c,c++,fortran
Processor Notes: Scaling Governor: intel_pstate powersave - CPU Microcode: 0x2000064
Java Notes: OpenJDK Runtime Environment (build 11.0.5+10-post-Ubuntu-0ubuntu1.119.04)
Python Notes: Python 2.7.16 + Python 3.7.3
Security Notes: 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

Testing initiated at 29 September 2020 08:43 by user phoronix.

2

Compiler Notes: --disable-multilib --enable-checking=release --enable-languages=c,c++,fortran
Processor Notes: Scaling Governor: intel_pstate powersave - CPU Microcode: 0x2000064
Java Notes: OpenJDK Runtime Environment (build 11.0.5+10-post-Ubuntu-0ubuntu1.119.04)
Python Notes: Python 2.7.16 + Python 3.7.3
Security Notes: 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

Testing initiated at 30 September 2020 12:22 by user phoronix.

3

Processor: Intel Core i7-7900X @ 4.50GHz (10 Cores / 20 Threads), Motherboard: ASRock X299 Extreme4 (P1.50 BIOS), Chipset: Intel Sky Lake-E DMI3 Registers, Memory: 16GB, Disk: 120GB Corsair Force MP500, Graphics: Zotac NVIDIA GeForce GT 610 1GB, Audio: Realtek ALC1220, Monitor: LG Ultra HD, Network: Intel I219-V

OS: Ubuntu 19.04, Kernel: 5.0.0-38-generic (x86_64), Desktop: GNOME Shell 3.32.1, Display Server: X Server 1.20.4, Display Driver: modesetting 1.20.4, Compiler: GCC 11.0.0 20200929, File-System: ext4, Screen Resolution: 1920x1080

Compiler Notes: --disable-multilib --enable-checking=release --enable-languages=c,c++,fortran
Processor Notes: Scaling Governor: intel_pstate powersave - CPU Microcode: 0x2000064
Java Notes: OpenJDK Runtime Environment (build 11.0.5+10-post-Ubuntu-0ubuntu1.119.04)
Python Notes: Python 2.7.16 + Python 3.7.3
Security Notes: 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

Testing initiated at 1 October 2020 07:13 by user phoronix.