Xeon Gold 6226R March 2021

Intel Xeon Gold 6226R testing with a Supermicro X11SPL-F v1.02 (3.1 BIOS) and ASPEED on Ubuntu 20.10 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 2103284-PTS-XEONGOLD61
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 3 Tests
Timed Code Compilation 4 Tests
C/C++ Compiler Tests 8 Tests
CPU Massive 8 Tests
Creator Workloads 11 Tests
Cryptography 2 Tests
Encoding 5 Tests
Game Development 2 Tests
HPC - High Performance Computing 4 Tests
Imaging 3 Tests
Molecular Dynamics 3 Tests
MPI Benchmarks 3 Tests
Multi-Core 14 Tests
NVIDIA GPU Compute 2 Tests
OpenMPI Tests 3 Tests
Programmer / Developer System Benchmarks 6 Tests
Python Tests 3 Tests
Scientific Computing 3 Tests
Software Defined Radio 4 Tests
Server CPU Tests 7 Tests
Single-Threaded 2 Tests
Texture Compression 2 Tests
Video Encoding 5 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
Performance Per
Dollar
Date
Run
  Test
  Duration
1
March 27 2021
  2 Hours, 47 Minutes
2
March 27 2021
  8 Hours, 11 Minutes
3
March 28 2021
  2 Hours, 47 Minutes
Invert Hiding All Results Option
  4 Hours, 35 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

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: intel_cpufreq schedutil - CPU Microcode: 0x5003003
Python Notes: Python 3.8.6
Security Notes: itlb_multihit: KVM: Mitigation of VMX disabled + l1tf: Not affected + mds: Not affected + meltdown: Not affected + 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 Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Mitigation of TSX disabled

2

3

Processor: Intel Xeon Gold 6226R @ 3.90GHz (16 Cores / 32 Threads), Motherboard: Supermicro X11SPL-F v1.02 (3.1 BIOS), Chipset: Intel Sky Lake-E DMI3 Registers, Memory: 188GB, Disk: 280GB INTEL SSDPED1D280GA, Graphics: ASPEED, Network: 2 x Intel I210

OS: Ubuntu 20.10, Kernel: 5.11.0-rc4-max-boost-inv-patch (x86_64) 20210121, Desktop: GNOME Shell 3.38.1, Display Server: X Server 1.20.9, Compiler: GCC 10.2.0, File-System: ext4, Screen Resolution: 1024x768

Xeon Gold 6226R March 2021ProcessorMotherboardChipsetMemoryDiskGraphicsNetworkOSKernelDesktopDisplay ServerCompilerFile-SystemScreen Resolution123Intel Xeon Gold 6226R @ 3.90GHz (16 Cores / 32 Threads)Supermicro X11SPL-F v1.02 (3.1 BIOS)Intel Sky Lake-E DMI3 Registers188GB280GB INTEL SSDPED1D280GAASPEED2 x Intel I210Ubuntu 20.105.11.0-rc4-max-boost-inv-patch (x86_64) 20210121GNOME Shell 3.38.1X Server 1.20.9GCC 10.2.0ext41024x768OpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler Details- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Processor Details- Scaling Governor: intel_cpufreq schedutil - CPU Microcode: 0x5003003Python Details- Python 3.8.6Security Details- itlb_multihit: KVM: Mitigation of VMX disabled + l1tf: Not affected + mds: Not affected + meltdown: Not affected + 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 Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Mitigation of TSX disabled

123Result OverviewPhoronix Test Suite100%108%117%125%toyBrot Fractal Generatordav1dStockfishJPEG XL DecodingGNU RadiooneDNNViennaCLTimed Linux Kernel CompilationNgspiceXcompact3d Incompact3dlibavif avifencGcrypt LibraryTimed Mesa CompilationSVT-HEVCSVT-VP9simdjsonAOM AV1Zstd CompressionPennantASTC EncoderJPEG XLTimed Erlang/OTP CompilationGROMACSLiquid-DSPLuaRadioBasis UniversalBotanGNU GMP GMPbenchsrsLTETimed Node.js Compilation

Xeon Gold 6226R March 2021dav1d: Summer Nature 1080pgnuradio: FIR Filterjpegxl-decode: Allviennacl: CPU BLAS - dGEMM-NTdav1d: Summer Nature 4Kgnuradio: Hilbert Transformviennacl: CPU BLAS - dGEMM-TTstockfish: Total Timeviennacl: CPU BLAS - dGEMM-TNaom-av1: Speed 9 Realtime - Bosphorus 1080paom-av1: Speed 9 Realtime - Bosphorus 4Kgnuradio: Signal Source (Cosine)toybrot: TBBonednn: Deconvolution Batch shapes_3d - f32 - CPUaom-av1: Speed 6 Realtime - Bosphorus 4Kaom-av1: Speed 6 Two-Pass - Bosphorus 4Kcompress-zstd: 19 - Compression Speedonednn: Deconvolution Batch shapes_3d - u8s8f32 - CPUviennacl: CPU BLAS - dGEMM-NNviennacl: CPU BLAS - dDOTaom-av1: Speed 8 Realtime - Bosphorus 4Kjpegxl: JPEG - 5avifenc: 10gnuradio: Five Back to Back FIR Filtersngspice: C2670svt-vp9: PSNR/SSIM Optimized - Bosphorus 1080pcompress-zstd: 3, Long Mode - Compression Speedjpegxl: JPEG - 8simdjson: Kostyaonednn: IP Shapes 3D - bf16bf16bf16 - CPUaom-av1: Speed 8 Realtime - Bosphorus 1080pcompress-zstd: 3 - Decompression Speedincompact3d: input.i3d 129 Cells Per Directioncompress-zstd: 8, Long Mode - Compression Speedcompress-zstd: 19, Long Mode - Compression Speedcompress-zstd: 19, Long Mode - Decompression Speedcompress-zstd: 19 - Decompression Speedavifenc: 6, Losslessonednn: Recurrent Neural Network Inference - bf16bf16bf16 - CPUcompress-zstd: 3 - Compression Speedonednn: Recurrent Neural Network Inference - f32 - CPUliquid-dsp: 2 - 256 - 57jpegxl-decode: 1viennacl: CPU BLAS - sDOTviennacl: CPU BLAS - sAXPYjpegxl: PNG - 8avifenc: 10, Losslesscompress-zstd: 8 - Compression Speedsvt-hevc: 10 - Bosphorus 1080pngspice: C7552avifenc: 2jpegxl: PNG - 7aom-av1: Speed 6 Realtime - Bosphorus 1080ponednn: Deconvolution Batch shapes_1d - bf16bf16bf16 - CPUbuild-linux-kernel: Time To Compiletoybrot: OpenMPavifenc: 0onednn: Recurrent Neural Network Training - bf16bf16bf16 - CPUbotan: ChaCha20Poly1305svt-hevc: 7 - Bosphorus 1080psvt-vp9: Visual Quality Optimized - Bosphorus 1080pjpegxl: JPEG - 7svt-vp9: VMAF Optimized - Bosphorus 1080psrslte: OFDM_Testonednn: IP Shapes 3D - u8s8f32 - CPUavifenc: 6viennacl: CPU BLAS - sCOPYonednn: Convolution Batch Shapes Auto - u8s8f32 - CPUviennacl: CPU BLAS - dAXPYgcrypt: basis: ETC1Sbuild-mesa: Time To Compileonednn: Matrix Multiply Batch Shapes Transformer - u8s8f32 - CPUonednn: Recurrent Neural Network Inference - u8s8f32 - CPUjpegxl: PNG - 5liquid-dsp: 4 - 256 - 57viennacl: CPU BLAS - dGEMV-Tastcenc: Mediumviennacl: CPU BLAS - dCOPYcompress-zstd: 3, Long Mode - Decompression Speedbotan: ChaCha20Poly1305 - Decryptonednn: Recurrent Neural Network Training - u8s8f32 - CPUsrslte: PHY_DL_Testluaradio: Complex Phaseluaradio: FM Deemphasis Filterluaradio: Five Back to Back FIR Filterstoybrot: C++ Tasksonednn: IP Shapes 3D - f32 - CPUonednn: Deconvolution Batch shapes_3d - bf16bf16bf16 - CPUliquid-dsp: 1 - 256 - 57onednn: Matrix Multiply Batch Shapes Transformer - f32 - CPUbotan: CAST-256viennacl: CPU BLAS - dGEMV-Npennant: sedovbigsvt-hevc: 1 - Bosphorus 1080ponednn: IP Shapes 1D - u8s8f32 - CPUonednn: Matrix Multiply Batch Shapes Transformer - bf16bf16bf16 - CPUpennant: leblancbigcompress-zstd: 8, Long Mode - Decompression Speedbotan: CAST-256 - Decryptonednn: IP Shapes 1D - bf16bf16bf16 - CPUonednn: Deconvolution Batch shapes_1d - u8s8f32 - CPUonednn: Convolution Batch Shapes Auto - f32 - CPUaom-av1: Speed 4 Two-Pass - Bosphorus 4Kliquid-dsp: 8 - 256 - 57botan: KASUMIliquid-dsp: 16 - 256 - 57botan: AES-256botan: Blowfishaom-av1: Speed 6 Two-Pass - Bosphorus 1080pastcenc: Thoroughonednn: IP Shapes 1D - f32 - CPUbasis: UASTC Level 0build-erlang: Time To Compileonednn: Recurrent Neural Network Training - f32 - CPUgromacs: water_GMX50_barebotan: AES-256 - Decryptbotan: Blowfish - Decryptsimdjson: DistinctUserIDbasis: UASTC Level 2srslte: PHY_DL_Testbotan: Twofishbasis: UASTC Level 3compress-zstd: 8 - Decompression Speedonednn: Convolution Batch Shapes Auto - bf16bf16bf16 - CPUbotan: Twofish - Decryptincompact3d: input.i3d 193 Cells Per Directionbuild-nodejs: Time To Compileluaradio: Hilbert Transformtoybrot: C++ Threadsbotan: KASUMI - Decryptastcenc: Exhaustiveliquid-dsp: 32 - 256 - 57gmpbench: Total Timeaom-av1: Speed 4 Two-Pass - Bosphorus 1080paom-av1: Speed 0 Two-Pass - Bosphorus 1080paom-av1: Speed 0 Two-Pass - Bosphorus 4Ksimdjson: PartialTweetssimdjson: LargeRandonednn: Deconvolution Batch shapes_1d - f32 - CPUgnuradio: FM Deemphasis Filtergnuradio: IIR Filter123269.99533.9198.3556.2174.26323.859.23870274158.248.8521.882257.7293763.472387.524.19590.90602257.310816.9343.464.504726.7200.497267.34367.420.752.342.8367838.543069.414.0785799429.540.72729.22670.745.573993.8182913.21008.888893400032.081241250.637.448686.5307.85171.35637.7447.1611.0912.36166.692907170.1851764.31670.062152.36201.9641.56267.031128000001.3921514.23283.74.61416105236.84828.83142.660.597683990.54440.731770200001166.715571.73303.5662.7251771.9978.3469.5357.9857.6286183.4270112.7549492190001.09663128.31599.656.2035810.410.5413642.2703427.251853269.1128.8185.881320.5810514.700482.3634371000083.9616389500003647.61407.1718.4710.63312.551489.041121.21779.671.8063659.312403.444.0325.422190.6332.66845.2133114.69.83488332.01752.4951591329.50664.92816381.21478.29817288100004757.73.360.340.133.380.835.55575645.1412.6271.02522.9179.7655.5172.64308.557.94024854957.748.1223.022152.7282873.408387.784.1360.40.93017555.710816.7442.364.612717.5205.123269.85359.220.302.322.8975638.143115.113.8286765436.741.42677.62701.145.7321005.652935.01003.289039833332.201251250.647.463676.2311.16171.17238.2897.1211.0912.423565.8502888871.0741784.29662.065152.04204.1342.04264.141138666671.4069014.30683.94.57869106235.25528.92742.8620.592321999.49040.791782100001166.713571.73291.7657.5301785.7277.8466.1359.7853.9288033.4171712.6919489783331.09014128.91799.856.4599910.380.5387772.2612627.367803254.0129.4095.880650.5818144.688142.3734262333383.6156382733333645.653408.1098.4610.66722.542749.035120.8161774.741.8093661.517404.4474.0325.465190.4333.13545.2993117.59.84755332.57452.5711772329.81665.02820181.24778.30657285566674759.53.360.340.133.380.834.50935588.5448.3317.53479.1178.4250.9187.9297.754.74122171154.846.1722.952193.4281343.33587.814.05610.90389757.111116.4942.764.619735.4202.75273.5359.320.692.372.8892438.943133.414.1164236438.241.52721.42721.446.4151011.782883991.1598979100031.681261270.637.35678.3312.5173.73538.2317.0610.9412.524566.7212869370.4431786.55663.9153.85201.7841.91266.291140000001.3926114.1784.54.62231105237.49129.09542.4760.597077995.40841.091785700001156.770872.33318.7658.1281779.4377.7468.5357.1851.8287463.4051912.6749492870001.09422129.00699.356.4781510.430.5393352.2720727.378543258.6129.0185.855890.57934.680432.3634406000083.8536363700003633.666408.6918.4910.66972.547769.011120.8111779.421.8043651.744404.0194.0425.479190.2332.44845.2113120.39.83093332.23852.5831146330.035652816481.31478.36747282400004756.23.360.340.133.380.833.98082531.6435.9OpenBenchmarking.org

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 0.8.2Video Input: Summer Nature 1080p12370140210280350SE +/- 0.83, N = 3269.99271.02317.53MIN: 246.57 / MAX: 294.05MIN: 239.91 / MAX: 295.67MIN: 291.94 / MAX: 371.151. (CC) gcc options: -pthread
OpenBenchmarking.orgFPS, More Is Betterdav1d 0.8.2Video Input: Summer Nature 1080p12360120180240300Min: 269.39 / Avg: 271.02 / Max: 272.121. (CC) gcc options: -pthread

GNU Radio

GNU Radio is a free software development toolkit providing signal processing blocks to implement software-defined radios (SDR) and signal processing systems. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterGNU RadioTest: FIR Filter123120240360480600SE +/- 12.50, N = 3533.9522.9479.11. 3.8.1.0
OpenBenchmarking.orgMiB/s, More Is BetterGNU RadioTest: FIR Filter12390180270360450Min: 498.6 / Avg: 522.87 / Max: 540.21. 3.8.1.0

JPEG XL Decoding

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is suited for JPEG XL decode performance testing to PNG output file, the pts/jpexl test is for encode performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL Decoding 0.3.3CPU Threads: All1234080120160200SE +/- 1.38, N = 3198.35179.76178.42
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL Decoding 0.3.3CPU Threads: All1234080120160200Min: 177.32 / Avg: 179.76 / Max: 182.1

ViennaCL

ViennaCL is an open-source linear algebra library written in C++ and with support for OpenCL and OpenMP. This test profile makes use of ViennaCL's built-in benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMM-NT1231326395265SE +/- 0.57, N = 356.255.550.91. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMM-NT1231122334455Min: 54.4 / Avg: 55.5 / Max: 56.31. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

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 0.8.2Video Input: Summer Nature 4K1234080120160200SE +/- 0.57, N = 3174.26172.64187.90MIN: 157.96 / MAX: 208.32MIN: 139.18 / MAX: 203.2MIN: 172.38 / MAX: 222.921. (CC) gcc options: -pthread
OpenBenchmarking.orgFPS, More Is Betterdav1d 0.8.2Video Input: Summer Nature 4K123306090120150Min: 171.51 / Avg: 172.64 / Max: 173.281. (CC) gcc options: -pthread

GNU Radio

GNU Radio is a free software development toolkit providing signal processing blocks to implement software-defined radios (SDR) and signal processing systems. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterGNU RadioTest: Hilbert Transform12370140210280350SE +/- 5.15, N = 3323.8308.5297.71. 3.8.1.0
OpenBenchmarking.orgMiB/s, More Is BetterGNU RadioTest: Hilbert Transform12360120180240300Min: 302.7 / Avg: 308.53 / Max: 318.81. 3.8.1.0

ViennaCL

ViennaCL is an open-source linear algebra library written in C++ and with support for OpenCL and OpenMP. This test profile makes use of ViennaCL's built-in benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMM-TT1231326395265SE +/- 1.65, N = 359.257.954.71. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMM-TT1231224364860Min: 54.6 / Avg: 57.87 / Max: 59.91. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

Stockfish

This is a test of Stockfish, an advanced open-source C++11 chess benchmark that can scale up to 512 CPU threads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 13Total Time1239M18M27M36M45MSE +/- 343002.27, N = 33870274140248549412217111. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fprofile-use -fno-peel-loops -fno-tracer -pedantic -O3 -msse -msse3 -mpopcnt -mavx2 -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto=jobserver
OpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 13Total Time1237M14M21M28M35MMin: 39564599 / Avg: 40248549.33 / Max: 406364661. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fprofile-use -fno-peel-loops -fno-tracer -pedantic -O3 -msse -msse3 -mpopcnt -mavx2 -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto=jobserver

ViennaCL

ViennaCL is an open-source linear algebra library written in C++ and with support for OpenCL and OpenMP. This test profile makes use of ViennaCL's built-in benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMM-TN1231326395265SE +/- 1.21, N = 358.257.754.81. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMM-TN1231224364860Min: 55.3 / Avg: 57.7 / Max: 59.21. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 9 Realtime - Input: Bosphorus 1080p1231122334455SE +/- 0.69, N = 348.8548.1246.171. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 9 Realtime - Input: Bosphorus 1080p1231020304050Min: 46.78 / Avg: 48.12 / Max: 49.081. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4K123612182430SE +/- 0.09, N = 321.8823.0222.951. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4K123510152025Min: 22.87 / Avg: 23.02 / Max: 23.181. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

GNU Radio

GNU Radio is a free software development toolkit providing signal processing blocks to implement software-defined radios (SDR) and signal processing systems. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterGNU RadioTest: Signal Source (Cosine)1235001000150020002500SE +/- 54.15, N = 32257.72152.72193.41. 3.8.1.0
OpenBenchmarking.orgMiB/s, More Is BetterGNU RadioTest: Signal Source (Cosine)123400800120016002000Min: 2051.1 / Avg: 2152.7 / Max: 22361. 3.8.1.0

toyBrot Fractal Generator

ToyBrot is a Mandelbrot fractal generator supporting C++ threads/tasks, OpenMP, Intel Threaded Building Blocks (TBB), and other targets. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BettertoyBrot Fractal Generator 2020-11-18Implementation: TBB1236K12K18K24K30KSE +/- 71.70, N = 32937628287281341. (CXX) g++ options: -O3 -lpthread
OpenBenchmarking.orgms, Fewer Is BettertoyBrot Fractal Generator 2020-11-18Implementation: TBB1235K10K15K20K25KMin: 28144 / Avg: 28287.33 / Max: 283631. (CXX) g++ options: -O3 -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 Intel oneAPI initiative. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU1230.78131.56262.34393.12523.9065SE +/- 0.03666, N = 33.472383.408383.33580MIN: 3.17MIN: 3.17MIN: 3.171. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU123246810Min: 3.34 / Avg: 3.41 / Max: 3.471. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 6 Realtime - Input: Bosphorus 4K123246810SE +/- 0.01, N = 37.527.787.811. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 6 Realtime - Input: Bosphorus 4K1233691215Min: 7.77 / Avg: 7.78 / Max: 7.791. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4K1230.94281.88562.82843.77124.714SE +/- 0.05, N = 34.194.134.051. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4K123246810Min: 4.06 / Avg: 4.13 / Max: 4.231. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

Zstd Compression

This test measures the time needed to compress/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 19 - Compression Speed1231428425670SE +/- 0.67, N = 359.060.461.01. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 19 - Compression Speed1231224364860Min: 59.1 / Avg: 60.4 / Max: 61.31. (CC) gcc options: -O3 -pthread -lz -llzma

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU1230.20930.41860.62790.83721.0465SE +/- 0.013165, N = 30.9060220.9301750.903897MIN: 0.84MIN: 0.84MIN: 0.831. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU123246810Min: 0.9 / Avg: 0.93 / Max: 0.951. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

ViennaCL

ViennaCL is an open-source linear algebra library written in C++ and with support for OpenCL and OpenMP. This test profile makes use of ViennaCL's built-in benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMM-NN1231326395265SE +/- 1.50, N = 357.355.757.11. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMM-NN1231122334455Min: 53.9 / Avg: 55.73 / Max: 58.71. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dDOT12320406080100SE +/- 0.58, N = 31081081111. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dDOT12320406080100Min: 107 / Avg: 108 / Max: 1091. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4K12348121620SE +/- 0.17, N = 316.9316.7416.491. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4K12348121620Min: 16.41 / Avg: 16.74 / Max: 16.921. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

JPEG XL

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

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL 0.3.3Input: JPEG - Encode Speed: 51231020304050SE +/- 0.23, N = 343.4642.3642.761. (CXX) g++ options: -funwind-tables -O3 -O2 -pthread -fPIE -pie -ldl
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL 0.3.3Input: JPEG - Encode Speed: 5123918273645Min: 42.02 / Avg: 42.36 / Max: 42.791. (CXX) g++ options: -funwind-tables -O3 -O2 -pthread -fPIE -pie -ldl

libavif avifenc

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

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 101231.03932.07863.11794.15725.1965SE +/- 0.015, N = 34.5044.6124.6191. (CXX) g++ options: -O3 -fPIC -lm
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 10123246810Min: 4.58 / Avg: 4.61 / Max: 4.631. (CXX) g++ options: -O3 -fPIC -lm

GNU Radio

GNU Radio is a free software development toolkit providing signal processing blocks to implement software-defined radios (SDR) and signal processing systems. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterGNU RadioTest: Five Back to Back FIR Filters123160320480640800SE +/- 12.25, N = 3726.7717.5735.41. 3.8.1.0
OpenBenchmarking.orgMiB/s, More Is BetterGNU RadioTest: Five Back to Back FIR Filters123130260390520650Min: 693.6 / Avg: 717.5 / Max: 734.11. 3.8.1.0

Ngspice

Ngspice is an open-source SPICE circuit simulator. Ngspice was originally based on the Berkeley SPICE electronic circuit simulator. Ngspice supports basic threading using OpenMP. This test profile is making use of the ISCAS 85 benchmark circuits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterNgspice 34Circuit: C267012350100150200250SE +/- 1.40, N = 3200.50205.12202.751. (CC) gcc options: -O0 -fopenmp -lm -lstdc++ -lfftw3 -lXaw -lXmu -lXt -lXext -lX11 -lSM -lICE
OpenBenchmarking.orgSeconds, Fewer Is BetterNgspice 34Circuit: C26701234080120160200Min: 202.42 / Avg: 205.12 / Max: 207.111. (CC) gcc options: -O0 -fopenmp -lm -lstdc++ -lfftw3 -lXaw -lXmu -lXt -lXext -lX11 -lSM -lICE

SVT-VP9

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

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p12360120180240300SE +/- 2.98, N = 3267.34269.85273.501. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p12350100150200250Min: 264.84 / Avg: 269.85 / Max: 275.141. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

Zstd Compression

This test measures the time needed to compress/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 3, Long Mode - Compression Speed12380160240320400SE +/- 1.95, N = 3367.4359.2359.31. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 3, Long Mode - Compression Speed12370140210280350Min: 357.1 / Avg: 359.2 / Max: 363.11. (CC) gcc options: -O3 -pthread -lz -llzma

JPEG XL

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

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL 0.3.3Input: JPEG - Encode Speed: 8123510152025SE +/- 0.19, N = 320.7520.3020.691. (CXX) g++ options: -funwind-tables -O3 -O2 -pthread -fPIE -pie -ldl
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL 0.3.3Input: JPEG - Encode Speed: 8123510152025Min: 20.08 / Avg: 20.3 / Max: 20.681. (CXX) g++ options: -funwind-tables -O3 -O2 -pthread -fPIE -pie -ldl

simdjson

This is a benchmark of SIMDJSON, a high performance JSON parser. SIMDJSON aims to be the fastest JSON parser and is used by projects like Microsoft FishStore, Yandex ClickHouse, Shopify, and others. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 0.8.2Throughput Test: Kostya1230.53331.06661.59992.13322.6665SE +/- 0.01, N = 32.342.322.371. (CXX) g++ options: -O3 -pthread
OpenBenchmarking.orgGB/s, More Is Bettersimdjson 0.8.2Throughput Test: Kostya123246810Min: 2.3 / Avg: 2.32 / Max: 2.341. (CXX) g++ options: -O3 -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 Intel oneAPI initiative. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU1230.6521.3041.9562.6083.26SE +/- 0.00361, N = 32.836782.897562.88924MIN: 2.55MIN: 2.55MIN: 2.571. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU123246810Min: 2.89 / Avg: 2.9 / Max: 2.91. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 8 Realtime - Input: Bosphorus 1080p123918273645SE +/- 0.41, N = 338.5438.1438.941. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 8 Realtime - Input: Bosphorus 1080p123816243240Min: 37.35 / Avg: 38.14 / Max: 38.711. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

Zstd Compression

This test measures the time needed to compress/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 3 - Decompression Speed1237001400210028003500SE +/- 9.25, N = 23069.43115.13133.41. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 3 - Decompression Speed1235001000150020002500Min: 3105.8 / Avg: 3115.05 / Max: 3124.31. (CC) gcc options: -O3 -pthread -lz -llzma

Xcompact3d Incompact3d

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

OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 129 Cells Per Direction12348121620SE +/- 0.20, N = 314.0813.8314.121. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent -levent_pthreads -lutil -lm -lrt -lz
OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 129 Cells Per Direction12348121620Min: 13.42 / Avg: 13.83 / Max: 14.051. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent -levent_pthreads -lutil -lm -lrt -lz

Zstd Compression

This test measures the time needed to compress/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 8, Long Mode - Compression Speed12390180270360450SE +/- 1.59, N = 3429.5436.7438.21. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 8, Long Mode - Compression Speed12380160240320400Min: 433.8 / Avg: 436.67 / Max: 439.31. (CC) gcc options: -O3 -pthread -lz -llzma

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 19, Long Mode - Compression Speed123918273645SE +/- 0.15, N = 340.741.441.51. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 19, Long Mode - Compression Speed123918273645Min: 41.2 / Avg: 41.43 / Max: 41.71. (CC) gcc options: -O3 -pthread -lz -llzma

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 19, Long Mode - Decompression Speed1236001200180024003000SE +/- 16.34, N = 32729.22677.62721.41. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 19, Long Mode - Decompression Speed1235001000150020002500Min: 2645.1 / Avg: 2677.57 / Max: 2697.11. (CC) gcc options: -O3 -pthread -lz -llzma

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 19 - Decompression Speed1236001200180024003000SE +/- 15.06, N = 32670.72701.12721.41. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 19 - Decompression Speed1235001000150020002500Min: 2672.1 / Avg: 2701.07 / Max: 2722.71. (CC) gcc options: -O3 -pthread -lz -llzma

libavif avifenc

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

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 6, Lossless1231122334455SE +/- 0.12, N = 345.5745.7346.421. (CXX) g++ options: -O3 -fPIC -lm
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 6, Lossless123918273645Min: 45.49 / Avg: 45.73 / Max: 45.851. (CXX) g++ options: -O3 -fPIC -lm

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU1232004006008001000SE +/- 0.52, N = 3993.821005.651011.78MIN: 968.8MIN: 980.08MIN: 985.371. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU1232004006008001000Min: 1005.03 / Avg: 1005.65 / Max: 1006.691. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Zstd Compression

This test measures the time needed to compress/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 3 - Compression Speed1236001200180024003000SE +/- 17.00, N = 32913.22935.02883.01. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 3 - Compression Speed1235001000150020002500Min: 2910.1 / Avg: 2935 / Max: 2967.51. (CC) gcc options: -O3 -pthread -lz -llzma

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU1232004006008001000SE +/- 0.67, N = 31008.881003.28991.16MIN: 982.03MIN: 978.75MIN: 968.711. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU1232004006008001000Min: 1002.3 / Avg: 1003.28 / Max: 1004.561. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Liquid-DSP

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

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 2 - Buffer Length: 256 - Filter Length: 5712320M40M60M80M100MSE +/- 69848.25, N = 38893400090398333897910001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 2 - Buffer Length: 256 - Filter Length: 5712316M32M48M64M80MMin: 90326000 / Avg: 90398333.33 / Max: 905380001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

JPEG XL Decoding

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is suited for JPEG XL decode performance testing to PNG output file, the pts/jpexl test is for encode performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL Decoding 0.3.3CPU Threads: 1123714212835SE +/- 0.20, N = 332.0832.2031.68
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL Decoding 0.3.3CPU Threads: 1123714212835Min: 31.83 / Avg: 32.2 / Max: 32.53

ViennaCL

ViennaCL is an open-source linear algebra library written in C++ and with support for OpenCL and OpenMP. This test profile makes use of ViennaCL's built-in benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - sDOT1233060901201501241251261. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - sAXPY123306090120150SE +/- 0.58, N = 31251251271. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - sAXPY12320406080100Min: 124 / Avg: 125 / Max: 1261. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

JPEG XL

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

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL 0.3.3Input: PNG - Encode Speed: 81230.1440.2880.4320.5760.72SE +/- 0.00, N = 30.630.640.631. (CXX) g++ options: -funwind-tables -O3 -O2 -pthread -fPIE -pie -ldl
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL 0.3.3Input: PNG - Encode Speed: 8123246810Min: 0.63 / Avg: 0.64 / Max: 0.641. (CXX) g++ options: -funwind-tables -O3 -O2 -pthread -fPIE -pie -ldl

libavif avifenc

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

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 10, Lossless123246810SE +/- 0.020, N = 37.4487.4637.3501. (CXX) g++ options: -O3 -fPIC -lm
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 10, Lossless1233691215Min: 7.44 / Avg: 7.46 / Max: 7.51. (CXX) g++ options: -O3 -fPIC -lm

Zstd Compression

This test measures the time needed to compress/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 8 - Compression Speed123150300450600750SE +/- 1.10, N = 3686.5676.2678.31. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 8 - Compression Speed123120240360480600Min: 674.4 / Avg: 676.23 / Max: 678.21. (CC) gcc options: -O3 -pthread -lz -llzma

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 10 - Input: Bosphorus 1080p12370140210280350SE +/- 1.08, N = 3307.85311.16312.501. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 10 - Input: Bosphorus 1080p12360120180240300Min: 309.92 / Avg: 311.16 / Max: 313.321. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

Ngspice

Ngspice is an open-source SPICE circuit simulator. Ngspice was originally based on the Berkeley SPICE electronic circuit simulator. Ngspice supports basic threading using OpenMP. This test profile is making use of the ISCAS 85 benchmark circuits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterNgspice 34Circuit: C75521234080120160200SE +/- 1.27, N = 3171.36171.17173.741. (CC) gcc options: -O0 -fopenmp -lm -lstdc++ -lfftw3 -lXaw -lXmu -lXt -lXext -lX11 -lSM -lICE
OpenBenchmarking.orgSeconds, Fewer Is BetterNgspice 34Circuit: C7552123306090120150Min: 169.48 / Avg: 171.17 / Max: 173.651. (CC) gcc options: -O0 -fopenmp -lm -lstdc++ -lfftw3 -lXaw -lXmu -lXt -lXext -lX11 -lSM -lICE

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 0.9.0Encoder Speed: 2123918273645SE +/- 0.21, N = 337.7438.2938.231. (CXX) g++ options: -O3 -fPIC -lm
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 2123816243240Min: 37.95 / Avg: 38.29 / Max: 38.661. (CXX) g++ options: -O3 -fPIC -lm

JPEG XL

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

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL 0.3.3Input: PNG - Encode Speed: 7123246810SE +/- 0.03, N = 37.167.127.061. (CXX) g++ options: -funwind-tables -O3 -O2 -pthread -fPIE -pie -ldl
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL 0.3.3Input: PNG - Encode Speed: 71233691215Min: 7.06 / Avg: 7.12 / Max: 7.171. (CXX) g++ options: -funwind-tables -O3 -O2 -pthread -fPIE -pie -ldl

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 6 Realtime - Input: Bosphorus 1080p1233691215SE +/- 0.01, N = 311.0911.0910.941. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 6 Realtime - Input: Bosphorus 1080p1233691215Min: 11.06 / Avg: 11.09 / Max: 11.111. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -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 Intel oneAPI initiative. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU1233691215SE +/- 0.07, N = 312.3612.4212.52MIN: 11.45MIN: 11.45MIN: 11.451. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU12348121620Min: 12.32 / Avg: 12.42 / Max: 12.551. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 5.10.20Time To Compile1231530456075SE +/- 0.38, N = 366.6965.8566.72
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 5.10.20Time To Compile1231326395265Min: 65.36 / Avg: 65.85 / Max: 66.61

toyBrot Fractal Generator

ToyBrot is a Mandelbrot fractal generator supporting C++ threads/tasks, OpenMP, Intel Threaded Building Blocks (TBB), and other targets. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BettertoyBrot Fractal Generator 2020-11-18Implementation: OpenMP1236K12K18K24K30KSE +/- 142.73, N = 32907128888286931. (CXX) g++ options: -O3 -lpthread
OpenBenchmarking.orgms, Fewer Is BettertoyBrot Fractal Generator 2020-11-18Implementation: OpenMP1235K10K15K20K25KMin: 28604 / Avg: 28888 / Max: 290551. (CXX) g++ options: -O3 -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 0.9.0Encoder Speed: 01231632486480SE +/- 0.36, N = 370.1971.0770.441. (CXX) g++ options: -O3 -fPIC -lm
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 01231428425670Min: 70.41 / Avg: 71.07 / Max: 71.631. (CXX) g++ options: -O3 -fPIC -lm

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU123400800120016002000SE +/- 4.93, N = 31764.311784.291786.55MIN: 1737.4MIN: 1739.16MIN: 1755.211. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU12330060090012001500Min: 1775.8 / Avg: 1784.29 / Max: 1792.871. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Botan

Botan is a BSD-licensed cross-platform open-source C++ crypto library "cryptography toolkit" that supports most publicly known cryptographic algorithms. The project's stated goal is to be "the best option for cryptography in C++ by offering the tools necessary to implement a range of practical systems, such as TLS protocol, X.509 certificates, modern AEAD ciphers, PKCS#11 and TPM hardware support, password hashing, and post quantum crypto schemes." Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: ChaCha20Poly1305123140280420560700SE +/- 1.77, N = 3670.06662.07663.901. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt
OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: ChaCha20Poly1305123120240360480600Min: 659.66 / Avg: 662.07 / Max: 665.521. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 7 - Input: Bosphorus 1080p123306090120150SE +/- 0.36, N = 3152.36152.04153.851. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 7 - Input: Bosphorus 1080p123306090120150Min: 151.32 / Avg: 152.04 / Max: 152.481. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

SVT-VP9

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

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: Visual Quality Optimized - Input: Bosphorus 1080p1234080120160200SE +/- 0.48, N = 3201.96204.13201.781. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: Visual Quality Optimized - Input: Bosphorus 1080p1234080120160200Min: 203.38 / Avg: 204.13 / Max: 205.021. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

JPEG XL

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

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL 0.3.3Input: JPEG - Encode Speed: 71231020304050SE +/- 0.43, N = 341.5642.0441.911. (CXX) g++ options: -funwind-tables -O3 -O2 -pthread -fPIE -pie -ldl
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL 0.3.3Input: JPEG - Encode Speed: 7123918273645Min: 41.17 / Avg: 42.04 / Max: 42.511. (CXX) g++ options: -funwind-tables -O3 -O2 -pthread -fPIE -pie -ldl

SVT-VP9

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

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 1080p12360120180240300SE +/- 1.43, N = 3267.03264.14266.291. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 1080p12350100150200250Min: 262.5 / Avg: 264.14 / Max: 266.981. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

srsLTE

srsLTE is an open-source LTE software radio suite created by Software Radio Systems (SRS). srsLTE can be used for building your own software defined (SDR) LTE mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSamples / Second, More Is BettersrsLTE 20.10.1Test: OFDM_Test12320M40M60M80M100MSE +/- 348010.22, N = 31128000001138666671140000001. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -mavx512f -mavx512cd -mavx512bw -mavx512dq -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lbladeRF -lm -lfftw3f
OpenBenchmarking.orgSamples / Second, More Is BettersrsLTE 20.10.1Test: OFDM_Test12320M40M60M80M100MMin: 113300000 / Avg: 113866666.67 / Max: 1145000001. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -mavx512f -mavx512cd -mavx512bw -mavx512dq -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lbladeRF -lm -lfftw3f

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU1230.31660.63320.94981.26641.583SE +/- 0.01012, N = 31.392151.406901.39261MIN: 1.22MIN: 1.21MIN: 1.211. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU123246810Min: 1.39 / Avg: 1.41 / Max: 1.431. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

libavif avifenc

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

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 612348121620SE +/- 0.03, N = 314.2314.3114.171. (CXX) g++ options: -O3 -fPIC -lm
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 612348121620Min: 14.25 / Avg: 14.31 / Max: 14.341. (CXX) g++ options: -O3 -fPIC -lm

ViennaCL

ViennaCL is an open-source linear algebra library written in C++ and with support for OpenCL and OpenMP. This test profile makes use of ViennaCL's built-in benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - sCOPY12320406080100SE +/- 0.09, N = 383.783.984.51. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - sCOPY1231632486480Min: 83.8 / Avg: 83.93 / Max: 84.11. (CXX) g++ options: -fopenmp -O3 -rdynamic -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 Intel oneAPI initiative. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU1231.042.083.124.165.2SE +/- 0.01741, N = 34.614164.578694.62231MIN: 4.16MIN: 4.13MIN: 4.171. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU123246810Min: 4.54 / Avg: 4.58 / Max: 4.61. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

ViennaCL

ViennaCL is an open-source linear algebra library written in C++ and with support for OpenCL and OpenMP. This test profile makes use of ViennaCL's built-in benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dAXPY123204060801001051061051. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

Gcrypt Library

Libgcrypt is a general purpose cryptographic library developed as part of the GnuPG project. This is a benchmark of libgcrypt's integrated benchmark and is measuring the time to run the benchmark command with a cipher/mac/hash repetition count set for 50 times as simple, high level look at the overall crypto performance of the system under test. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGcrypt Library 1.912350100150200250SE +/- 0.71, N = 3236.85235.26237.491. (CC) gcc options: -O2 -fvisibility=hidden -lgpg-error
OpenBenchmarking.orgSeconds, Fewer Is BetterGcrypt Library 1.91234080120160200Min: 233.85 / Avg: 235.26 / Max: 236.111. (CC) gcc options: -O2 -fvisibility=hidden -lgpg-error

Basis Universal

Basis Universal is a GPU texture codec. 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 1.13Settings: ETC1S123714212835SE +/- 0.12, N = 328.8328.9329.101. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.13Settings: ETC1S123612182430Min: 28.68 / Avg: 28.93 / Max: 29.081. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread

Timed Mesa Compilation

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

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Mesa Compilation 21.0Time To Compile1231020304050SE +/- 0.03, N = 342.6642.8642.48
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Mesa Compilation 21.0Time To Compile123918273645Min: 42.81 / Avg: 42.86 / Max: 42.91

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU1230.13450.2690.40350.5380.6725SE +/- 0.002509, N = 30.5976830.5923210.597077MIN: 0.52MIN: 0.51MIN: 0.521. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU123246810Min: 0.59 / Avg: 0.59 / Max: 0.591. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU1232004006008001000SE +/- 4.21, N = 3990.54999.49995.41MIN: 959.7MIN: 966.55MIN: 971.261. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU1232004006008001000Min: 991.11 / Avg: 999.49 / Max: 1004.411. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

JPEG XL

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

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL 0.3.3Input: PNG - Encode Speed: 5123918273645SE +/- 0.18, N = 340.7340.7941.091. (CXX) g++ options: -funwind-tables -O3 -O2 -pthread -fPIE -pie -ldl
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL 0.3.3Input: PNG - Encode Speed: 5123918273645Min: 40.45 / Avg: 40.79 / Max: 41.041. (CXX) g++ options: -funwind-tables -O3 -O2 -pthread -fPIE -pie -ldl

Liquid-DSP

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

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 4 - Buffer Length: 256 - Filter Length: 5712340M80M120M160M200MSE +/- 507181.76, N = 31770200001782100001785700001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 4 - Buffer Length: 256 - Filter Length: 5712330M60M90M120M150MMin: 177290000 / Avg: 178210000 / Max: 1790400001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

ViennaCL

ViennaCL is an open-source linear algebra library written in C++ and with support for OpenCL and OpenMP. This test profile makes use of ViennaCL's built-in benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMV-T1233060901201501161161151. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

ASTC Encoder

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

OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.4Preset: Medium123246810SE +/- 0.0030, N = 36.71556.71356.77081. (CXX) g++ options: -O3 -flto -pthread
OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.4Preset: Medium1233691215Min: 6.71 / Avg: 6.71 / Max: 6.721. (CXX) g++ options: -O3 -flto -pthread

ViennaCL

ViennaCL is an open-source linear algebra library written in C++ and with support for OpenCL and OpenMP. This test profile makes use of ViennaCL's built-in benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dCOPY1231632486480SE +/- 0.15, N = 371.771.772.31. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dCOPY1231428425670Min: 71.4 / Avg: 71.7 / Max: 71.91. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

Zstd Compression

This test measures the time needed to compress/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 3, Long Mode - Decompression Speed1237001400210028003500SE +/- 15.06, N = 33303.53291.73318.71. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 3, Long Mode - Decompression Speed1236001200180024003000Min: 3271.8 / Avg: 3291.67 / Max: 3321.21. (CC) gcc options: -O3 -pthread -lz -llzma

Botan

Botan is a BSD-licensed cross-platform open-source C++ crypto library "cryptography toolkit" that supports most publicly known cryptographic algorithms. The project's stated goal is to be "the best option for cryptography in C++ by offering the tools necessary to implement a range of practical systems, such as TLS protocol, X.509 certificates, modern AEAD ciphers, PKCS#11 and TPM hardware support, password hashing, and post quantum crypto schemes." Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: ChaCha20Poly1305 - Decrypt123140280420560700SE +/- 0.48, N = 3662.73657.53658.131. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt
OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: ChaCha20Poly1305 - Decrypt123120240360480600Min: 656.56 / Avg: 657.53 / Max: 658.051. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -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 Intel oneAPI initiative. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU123400800120016002000SE +/- 1.23, N = 31771.991785.721779.43MIN: 1740.69MIN: 1752.23MIN: 1751.551. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU12330060090012001500Min: 1783.28 / Avg: 1785.72 / Max: 1787.281. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

srsLTE

srsLTE is an open-source LTE software radio suite created by Software Radio Systems (SRS). srsLTE can be used for building your own software defined (SDR) LTE mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgUE Mb/s, More Is BettersrsLTE 20.10.1Test: PHY_DL_Test12320406080100SE +/- 0.10, N = 378.377.877.71. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -mavx512f -mavx512cd -mavx512bw -mavx512dq -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lbladeRF -lm -lfftw3f
OpenBenchmarking.orgUE Mb/s, More Is BettersrsLTE 20.10.1Test: PHY_DL_Test1231530456075Min: 77.7 / Avg: 77.8 / Max: 781. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -mavx512f -mavx512cd -mavx512bw -mavx512dq -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lbladeRF -lm -lfftw3f

LuaRadio

LuaRadio is a lightweight software-defined radio (SDR) framework built atop LuaJIT. LuaRadio provides a suite of source, sink, and processing blocks, with a simple API for defining flow graphs, running flow graphs, creating blocks, and creating data types. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterLuaRadio 0.9.1Test: Complex Phase123100200300400500SE +/- 0.97, N = 3469.5466.1468.5
OpenBenchmarking.orgMiB/s, More Is BetterLuaRadio 0.9.1Test: Complex Phase12380160240320400Min: 464.3 / Avg: 466.13 / Max: 467.6

OpenBenchmarking.orgMiB/s, More Is BetterLuaRadio 0.9.1Test: FM Deemphasis Filter12380160240320400SE +/- 0.09, N = 3357.9359.7357.1
OpenBenchmarking.orgMiB/s, More Is BetterLuaRadio 0.9.1Test: FM Deemphasis Filter12360120180240300Min: 359.5 / Avg: 359.67 / Max: 359.8

OpenBenchmarking.orgMiB/s, More Is BetterLuaRadio 0.9.1Test: Five Back to Back FIR Filters1232004006008001000SE +/- 2.65, N = 3857.6853.9851.8
OpenBenchmarking.orgMiB/s, More Is BetterLuaRadio 0.9.1Test: Five Back to Back FIR Filters123150300450600750Min: 849.7 / Avg: 853.87 / Max: 858.8

toyBrot Fractal Generator

ToyBrot is a Mandelbrot fractal generator supporting C++ threads/tasks, OpenMP, Intel Threaded Building Blocks (TBB), and other targets. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BettertoyBrot Fractal Generator 2020-11-18Implementation: C++ Tasks1236K12K18K24K30KSE +/- 160.28, N = 32861828803287461. (CXX) g++ options: -O3 -lpthread
OpenBenchmarking.orgms, Fewer Is BettertoyBrot Fractal Generator 2020-11-18Implementation: C++ Tasks1235K10K15K20K25KMin: 28593 / Avg: 28803.33 / Max: 291181. (CXX) g++ options: -O3 -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 Intel oneAPI initiative. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU1230.77111.54222.31333.08443.8555SE +/- 0.00459, N = 33.427013.417173.40519MIN: 3.11MIN: 3.11MIN: 3.121. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU123246810Min: 3.41 / Avg: 3.42 / Max: 3.431. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPU1233691215SE +/- 0.02, N = 312.7512.6912.67MIN: 12.41MIN: 12.41MIN: 12.371. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPU12348121620Min: 12.66 / Avg: 12.69 / Max: 12.721. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Liquid-DSP

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

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 1 - Buffer Length: 256 - Filter Length: 5712311M22M33M44M55MSE +/- 214667.44, N = 34921900048978333492870001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 1 - Buffer Length: 256 - Filter Length: 571239M18M27M36M45MMin: 48549000 / Avg: 48978333.33 / Max: 491940001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU1230.24670.49340.74010.98681.2335SE +/- 0.00360, N = 31.096631.090141.09422MIN: 0.95MIN: 0.94MIN: 0.951. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU123246810Min: 1.09 / Avg: 1.09 / Max: 1.11. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Botan

Botan is a BSD-licensed cross-platform open-source C++ crypto library "cryptography toolkit" that supports most publicly known cryptographic algorithms. The project's stated goal is to be "the best option for cryptography in C++ by offering the tools necessary to implement a range of practical systems, such as TLS protocol, X.509 certificates, modern AEAD ciphers, PKCS#11 and TPM hardware support, password hashing, and post quantum crypto schemes." Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: CAST-256123306090120150SE +/- 0.06, N = 3128.32128.92129.011. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt
OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: CAST-25612320406080100Min: 128.8 / Avg: 128.92 / Max: 1291. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

ViennaCL

ViennaCL is an open-source linear algebra library written in C++ and with support for OpenCL and OpenMP. This test profile makes use of ViennaCL's built-in benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMV-N12320406080100SE +/- 0.59, N = 399.699.899.31. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMV-N12320406080100Min: 99.1 / Avg: 99.83 / Max: 1011. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

Pennant

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

OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: sedovbig1231326395265SE +/- 0.02, N = 356.2056.4656.481. (CXX) g++ options: -fopenmp -pthread -lmpi_cxx -lmpi
OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: sedovbig1231122334455Min: 56.42 / Avg: 56.46 / Max: 56.491. (CXX) g++ options: -fopenmp -pthread -lmpi_cxx -lmpi

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 1 - Input: Bosphorus 1080p1233691215SE +/- 0.03, N = 310.4110.3810.431. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 1 - Input: Bosphorus 1080p1233691215Min: 10.32 / Avg: 10.38 / Max: 10.421. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -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 Intel oneAPI initiative. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU1230.12180.24360.36540.48720.609SE +/- 0.000426, N = 30.5413640.5387770.539335MIN: 0.47MIN: 0.47MIN: 0.481. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU123246810Min: 0.54 / Avg: 0.54 / Max: 0.541. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU1230.51121.02241.53362.04482.556SE +/- 0.00606, N = 32.270342.261262.27207MIN: 2.01MIN: 1.97MIN: 2.011. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU123246810Min: 2.25 / Avg: 2.26 / Max: 2.271. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Pennant

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

OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: leblancbig123612182430SE +/- 0.01, N = 327.2527.3727.381. (CXX) g++ options: -fopenmp -pthread -lmpi_cxx -lmpi
OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: leblancbig123612182430Min: 27.35 / Avg: 27.37 / Max: 27.391. (CXX) g++ options: -fopenmp -pthread -lmpi_cxx -lmpi

Zstd Compression

This test measures the time needed to compress/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 8, Long Mode - Decompression Speed1237001400210028003500SE +/- 16.76, N = 33269.13254.03258.61. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 8, Long Mode - Decompression Speed1236001200180024003000Min: 3227 / Avg: 3254.03 / Max: 3284.71. (CC) gcc options: -O3 -pthread -lz -llzma

Botan

Botan is a BSD-licensed cross-platform open-source C++ crypto library "cryptography toolkit" that supports most publicly known cryptographic algorithms. The project's stated goal is to be "the best option for cryptography in C++ by offering the tools necessary to implement a range of practical systems, such as TLS protocol, X.509 certificates, modern AEAD ciphers, PKCS#11 and TPM hardware support, password hashing, and post quantum crypto schemes." Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: CAST-256 - Decrypt123306090120150SE +/- 0.02, N = 3128.82129.41129.021. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt
OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: CAST-256 - Decrypt12320406080100Min: 129.38 / Avg: 129.41 / Max: 129.441. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -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 Intel oneAPI initiative. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU1231.32332.64663.96995.29326.6165SE +/- 0.01210, N = 35.881325.880655.85589MIN: 5.43MIN: 5.44MIN: 5.471. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU123246810Min: 5.86 / Avg: 5.88 / Max: 5.91. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU1230.13090.26180.39270.52360.6545SE +/- 0.001705, N = 30.5810510.5818140.579300MIN: 0.52MIN: 0.52MIN: 0.521. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU123246810Min: 0.58 / Avg: 0.58 / Max: 0.581. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU1231.05762.11523.17284.23045.288SE +/- 0.00914, N = 34.700484.688144.68043MIN: 4.25MIN: 4.24MIN: 4.271. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU123246810Min: 4.68 / Avg: 4.69 / Max: 4.711. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 4K1230.53331.06661.59992.13322.6665SE +/- 0.01, N = 32.362.372.361. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 4K123246810Min: 2.36 / Avg: 2.37 / Max: 2.381. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

Liquid-DSP

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

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 8 - Buffer Length: 256 - Filter Length: 5712370M140M210M280M350MSE +/- 416706.66, N = 33437100003426233333440600001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 8 - Buffer Length: 256 - Filter Length: 5712360M120M180M240M300MMin: 341790000 / Avg: 342623333.33 / Max: 3430500001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Botan

Botan is a BSD-licensed cross-platform open-source C++ crypto library "cryptography toolkit" that supports most publicly known cryptographic algorithms. The project's stated goal is to be "the best option for cryptography in C++ by offering the tools necessary to implement a range of practical systems, such as TLS protocol, X.509 certificates, modern AEAD ciphers, PKCS#11 and TPM hardware support, password hashing, and post quantum crypto schemes." Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: KASUMI12320406080100SE +/- 0.31, N = 383.9683.6283.851. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt
OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: KASUMI1231632486480Min: 83.01 / Avg: 83.62 / Max: 84.051. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

Liquid-DSP

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

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 16 - Buffer Length: 256 - Filter Length: 57123140M280M420M560M700MSE +/- 451897.24, N = 36389500006382733336363700001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 16 - Buffer Length: 256 - Filter Length: 57123110M220M330M440M550MMin: 637370000 / Avg: 638273333.33 / Max: 6387500001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Botan

Botan is a BSD-licensed cross-platform open-source C++ crypto library "cryptography toolkit" that supports most publicly known cryptographic algorithms. The project's stated goal is to be "the best option for cryptography in C++ by offering the tools necessary to implement a range of practical systems, such as TLS protocol, X.509 certificates, modern AEAD ciphers, PKCS#11 and TPM hardware support, password hashing, and post quantum crypto schemes." Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: AES-2561238001600240032004000SE +/- 0.22, N = 33647.613645.653633.671. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt
OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: AES-2561236001200180024003000Min: 3645.24 / Avg: 3645.65 / Max: 3645.991. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: Blowfish12390180270360450SE +/- 0.24, N = 3407.17408.11408.691. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt
OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: Blowfish12370140210280350Min: 407.76 / Avg: 408.11 / Max: 408.581. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 1080p123246810SE +/- 0.01, N = 38.478.468.491. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 1080p1233691215Min: 8.44 / Avg: 8.46 / Max: 8.481. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

ASTC Encoder

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

OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.4Preset: Thorough1233691215SE +/- 0.01, N = 310.6310.6710.671. (CXX) g++ options: -O3 -flto -pthread
OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.4Preset: Thorough1233691215Min: 10.65 / Avg: 10.67 / Max: 10.681. (CXX) g++ options: -O3 -flto -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 Intel oneAPI initiative. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU1230.57411.14821.72232.29642.8705SE +/- 0.00338, N = 32.551482.542742.54776MIN: 2.27MIN: 2.26MIN: 2.271. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU123246810Min: 2.54 / Avg: 2.54 / Max: 2.551. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Basis Universal

Basis Universal is a GPU texture codec. 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 1.13Settings: UASTC Level 01233691215SE +/- 0.006, N = 39.0419.0359.0111. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.13Settings: UASTC Level 01233691215Min: 9.03 / Avg: 9.03 / Max: 9.051. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread

Timed Erlang/OTP Compilation

This test times how long it takes to compile Erlang/OTP. Erlang is a programming language and run-time for massively scalable soft real-time systems with high availability requirements. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Erlang/OTP Compilation 23.2Time To Compile123306090120150SE +/- 0.32, N = 3121.20120.82120.81
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Erlang/OTP Compilation 23.2Time To Compile12320406080100Min: 120.18 / Avg: 120.82 / Max: 121.23

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU123400800120016002000SE +/- 5.17, N = 31779.671774.741779.42MIN: 1752.37MIN: 1741.55MIN: 1751.791. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU12330060090012001500Min: 1768.68 / Avg: 1774.74 / Max: 1785.031. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

GROMACS

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

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2021Input: water_GMX50_bare1230.4070.8141.2211.6282.035SE +/- 0.003, N = 31.8061.8091.8041. (CXX) g++ options: -O3 -pthread
OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2021Input: water_GMX50_bare123246810Min: 1.8 / Avg: 1.81 / Max: 1.811. (CXX) g++ options: -O3 -pthread

Botan

Botan is a BSD-licensed cross-platform open-source C++ crypto library "cryptography toolkit" that supports most publicly known cryptographic algorithms. The project's stated goal is to be "the best option for cryptography in C++ by offering the tools necessary to implement a range of practical systems, such as TLS protocol, X.509 certificates, modern AEAD ciphers, PKCS#11 and TPM hardware support, password hashing, and post quantum crypto schemes." Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: AES-256 - Decrypt1238001600240032004000SE +/- 0.85, N = 33659.313661.523651.741. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt
OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: AES-256 - Decrypt1236001200180024003000Min: 3660.63 / Avg: 3661.52 / Max: 3663.211. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: Blowfish - Decrypt12390180270360450SE +/- 0.15, N = 3403.44404.45404.021. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt
OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: Blowfish - Decrypt12370140210280350Min: 404.15 / Avg: 404.45 / Max: 404.621. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

simdjson

This is a benchmark of SIMDJSON, a high performance JSON parser. SIMDJSON aims to be the fastest JSON parser and is used by projects like Microsoft FishStore, Yandex ClickHouse, Shopify, and others. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 0.8.2Throughput Test: DistinctUserID1230.9091.8182.7273.6364.545SE +/- 0.00, N = 34.034.034.041. (CXX) g++ options: -O3 -pthread
OpenBenchmarking.orgGB/s, More Is Bettersimdjson 0.8.2Throughput Test: DistinctUserID123246810Min: 4.03 / Avg: 4.03 / Max: 4.041. (CXX) g++ options: -O3 -pthread

Basis Universal

Basis Universal is a GPU texture codec. 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 1.13Settings: UASTC Level 2123612182430SE +/- 0.04, N = 325.4225.4725.481. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.13Settings: UASTC Level 2123612182430Min: 25.4 / Avg: 25.46 / Max: 25.521. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread

srsLTE

srsLTE is an open-source LTE software radio suite created by Software Radio Systems (SRS). srsLTE can be used for building your own software defined (SDR) LTE mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgeNb Mb/s, More Is BettersrsLTE 20.10.1Test: PHY_DL_Test1234080120160200SE +/- 0.29, N = 3190.6190.4190.21. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -mavx512f -mavx512cd -mavx512bw -mavx512dq -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lbladeRF -lm -lfftw3f
OpenBenchmarking.orgeNb Mb/s, More Is BettersrsLTE 20.10.1Test: PHY_DL_Test123306090120150Min: 189.9 / Avg: 190.4 / Max: 190.91. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -mavx512f -mavx512cd -mavx512bw -mavx512dq -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lbladeRF -lm -lfftw3f

Botan

Botan is a BSD-licensed cross-platform open-source C++ crypto library "cryptography toolkit" that supports most publicly known cryptographic algorithms. The project's stated goal is to be "the best option for cryptography in C++ by offering the tools necessary to implement a range of practical systems, such as TLS protocol, X.509 certificates, modern AEAD ciphers, PKCS#11 and TPM hardware support, password hashing, and post quantum crypto schemes." Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: Twofish12370140210280350SE +/- 0.60, N = 3332.67333.14332.451. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt
OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: Twofish12360120180240300Min: 332.09 / Avg: 333.14 / Max: 334.171. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

Basis Universal

Basis Universal is a GPU texture codec. 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 1.13Settings: UASTC Level 31231020304050SE +/- 0.05, N = 345.2145.3045.211. (CXX) g++ options: -std=c++11 -fvisibility=hidden -fPIC -fno-strict-aliasing -O3 -rdynamic -lm -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterBasis Universal 1.13Settings: UASTC Level 3123918273645Min: 45.25 / Avg: 45.3 / Max: 45.391. (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/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 8 - Decompression Speed1237001400210028003500SE +/- 3.38, N = 33114.63117.53120.31. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.9Compression Level: 8 - Decompression Speed1235001000150020002500Min: 3112.1 / Avg: 3117.47 / Max: 3123.71. (CC) gcc options: -O3 -pthread -lz -llzma

oneDNN

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

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU1233691215SE +/- 0.00857, N = 39.834889.847559.83093MIN: 9.08MIN: 9.08MIN: 9.111. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU1233691215Min: 9.83 / Avg: 9.85 / Max: 9.861. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

Botan

Botan is a BSD-licensed cross-platform open-source C++ crypto library "cryptography toolkit" that supports most publicly known cryptographic algorithms. The project's stated goal is to be "the best option for cryptography in C++ by offering the tools necessary to implement a range of practical systems, such as TLS protocol, X.509 certificates, modern AEAD ciphers, PKCS#11 and TPM hardware support, password hashing, and post quantum crypto schemes." Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: Twofish - Decrypt12370140210280350SE +/- 0.24, N = 3332.02332.57332.241. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt
OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: Twofish - Decrypt12360120180240300Min: 332.1 / Avg: 332.57 / Max: 332.831. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

Xcompact3d Incompact3d

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

OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 193 Cells Per Direction1231224364860SE +/- 0.28, N = 352.5052.5752.581. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent -levent_pthreads -lutil -lm -lrt -lz
OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 193 Cells Per Direction1231122334455Min: 52.21 / Avg: 52.57 / Max: 53.111. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent -levent_pthreads -lutil -lm -lrt -lz

Timed Node.js Compilation

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

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Node.js Compilation 15.11Time To Compile12370140210280350SE +/- 0.41, N = 3329.51329.82330.04
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Node.js Compilation 15.11Time To Compile12360120180240300Min: 329.01 / Avg: 329.82 / Max: 330.27

LuaRadio

LuaRadio is a lightweight software-defined radio (SDR) framework built atop LuaJIT. LuaRadio provides a suite of source, sink, and processing blocks, with a simple API for defining flow graphs, running flow graphs, creating blocks, and creating data types. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterLuaRadio 0.9.1Test: Hilbert Transform1231530456075SE +/- 0.03, N = 364.965.065.0
OpenBenchmarking.orgMiB/s, More Is BetterLuaRadio 0.9.1Test: Hilbert Transform1231326395265Min: 64.9 / Avg: 64.97 / Max: 65

toyBrot Fractal Generator

ToyBrot is a Mandelbrot fractal generator supporting C++ threads/tasks, OpenMP, Intel Threaded Building Blocks (TBB), and other targets. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BettertoyBrot Fractal Generator 2020-11-18Implementation: C++ Threads1236K12K18K24K30KSE +/- 43.02, N = 32816328201281641. (CXX) g++ options: -O3 -lpthread
OpenBenchmarking.orgms, Fewer Is BettertoyBrot Fractal Generator 2020-11-18Implementation: C++ Threads1235K10K15K20K25KMin: 28132 / Avg: 28201 / Max: 282801. (CXX) g++ options: -O3 -lpthread

Botan

Botan is a BSD-licensed cross-platform open-source C++ crypto library "cryptography toolkit" that supports most publicly known cryptographic algorithms. The project's stated goal is to be "the best option for cryptography in C++ by offering the tools necessary to implement a range of practical systems, such as TLS protocol, X.509 certificates, modern AEAD ciphers, PKCS#11 and TPM hardware support, password hashing, and post quantum crypto schemes." Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: KASUMI - Decrypt12320406080100SE +/- 0.35, N = 381.2181.2581.311. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt
OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.17.3Test: KASUMI - Decrypt1231632486480Min: 80.55 / Avg: 81.25 / Max: 81.61. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

ASTC Encoder

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

OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.4Preset: Exhaustive12320406080100SE +/- 0.01, N = 378.3078.3178.371. (CXX) g++ options: -O3 -flto -pthread
OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.4Preset: Exhaustive1231530456075Min: 78.29 / Avg: 78.31 / Max: 78.321. (CXX) g++ options: -O3 -flto -pthread

Liquid-DSP

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

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 32 - Buffer Length: 256 - Filter Length: 57123160M320M480M640M800MSE +/- 131698.31, N = 37288100007285566677282400001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 32 - Buffer Length: 256 - Filter Length: 57123130M260M390M520M650MMin: 728420000 / Avg: 728556666.67 / Max: 7288200001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

GNU GMP GMPbench

GMPbench is a test of the GNU Multiple Precision Arithmetic (GMP) Library. GMPbench is a single-threaded integer benchmark that leverages the GMP library to stress the CPU with widening integer multiplication. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGMPbench Score, More Is BetterGNU GMP GMPbench 6.2.1Total Time123100020003000400050004757.74759.54756.21. (CC) gcc options: -O3 -fomit-frame-pointer -lm

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 1080p1230.7561.5122.2683.0243.78SE +/- 0.00, N = 33.363.363.361. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 1080p123246810Min: 3.36 / Avg: 3.36 / Max: 3.371. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 1080p1230.07650.1530.22950.3060.3825SE +/- 0.00, N = 30.340.340.341. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 1080p12312345Min: 0.34 / Avg: 0.34 / Max: 0.341. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 4K1230.02930.05860.08790.11720.1465SE +/- 0.00, N = 30.130.130.131. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.0Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 4K12312345Min: 0.13 / Avg: 0.13 / Max: 0.131. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread

simdjson

This is a benchmark of SIMDJSON, a high performance JSON parser. SIMDJSON aims to be the fastest JSON parser and is used by projects like Microsoft FishStore, Yandex ClickHouse, Shopify, and others. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 0.8.2Throughput Test: PartialTweets1230.76051.5212.28153.0423.8025SE +/- 0.00, N = 33.383.383.381. (CXX) g++ options: -O3 -pthread
OpenBenchmarking.orgGB/s, More Is Bettersimdjson 0.8.2Throughput Test: PartialTweets123246810Min: 3.38 / Avg: 3.38 / Max: 3.381. (CXX) g++ options: -O3 -pthread

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 0.8.2Throughput Test: LargeRandom1230.18680.37360.56040.74720.934SE +/- 0.00, N = 30.830.830.831. (CXX) g++ options: -O3 -pthread
OpenBenchmarking.orgGB/s, More Is Bettersimdjson 0.8.2Throughput Test: LargeRandom123246810Min: 0.83 / Avg: 0.83 / Max: 0.841. (CXX) g++ options: -O3 -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 Intel oneAPI initiative. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU1231.252.53.7556.25SE +/- 0.09201, N = 155.555754.509353.98082MIN: 3.26MIN: 3.22MIN: 3.271. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU123246810Min: 4.05 / Avg: 4.51 / Max: 5.341. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

GNU Radio

GNU Radio is a free software development toolkit providing signal processing blocks to implement software-defined radios (SDR) and signal processing systems. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterGNU RadioTest: FM Deemphasis Filter123140280420560700SE +/- 22.66, N = 3645.1588.5531.61. 3.8.1.0
OpenBenchmarking.orgMiB/s, More Is BetterGNU RadioTest: FM Deemphasis Filter123110220330440550Min: 564.6 / Avg: 588.5 / Max: 633.81. 3.8.1.0

OpenBenchmarking.orgMiB/s, More Is BetterGNU RadioTest: IIR Filter123100200300400500SE +/- 15.90, N = 3412.6448.3435.91. 3.8.1.0
OpenBenchmarking.orgMiB/s, More Is BetterGNU RadioTest: IIR Filter12380160240320400Min: 417.2 / Avg: 448.33 / Max: 469.51. 3.8.1.0

142 Results Shown

dav1d
GNU Radio
JPEG XL Decoding
ViennaCL
dav1d
GNU Radio
ViennaCL
Stockfish
ViennaCL
AOM AV1:
  Speed 9 Realtime - Bosphorus 1080p
  Speed 9 Realtime - Bosphorus 4K
GNU Radio
toyBrot Fractal Generator
oneDNN
AOM AV1:
  Speed 6 Realtime - Bosphorus 4K
  Speed 6 Two-Pass - Bosphorus 4K
Zstd Compression
oneDNN
ViennaCL:
  CPU BLAS - dGEMM-NN
  CPU BLAS - dDOT
AOM AV1
JPEG XL
libavif avifenc
GNU Radio
Ngspice
SVT-VP9
Zstd Compression
JPEG XL
simdjson
oneDNN
AOM AV1
Zstd Compression
Xcompact3d Incompact3d
Zstd Compression:
  8, Long Mode - Compression Speed
  19, Long Mode - Compression Speed
  19, Long Mode - Decompression Speed
  19 - Decompression Speed
libavif avifenc
oneDNN
Zstd Compression
oneDNN
Liquid-DSP
JPEG XL Decoding
ViennaCL:
  CPU BLAS - sDOT
  CPU BLAS - sAXPY
JPEG XL
libavif avifenc
Zstd Compression
SVT-HEVC
Ngspice
libavif avifenc
JPEG XL
AOM AV1
oneDNN
Timed Linux Kernel Compilation
toyBrot Fractal Generator
libavif avifenc
oneDNN
Botan
SVT-HEVC
SVT-VP9
JPEG XL
SVT-VP9
srsLTE
oneDNN
libavif avifenc
ViennaCL
oneDNN
ViennaCL
Gcrypt Library
Basis Universal
Timed Mesa Compilation
oneDNN:
  Matrix Multiply Batch Shapes Transformer - u8s8f32 - CPU
  Recurrent Neural Network Inference - u8s8f32 - CPU
JPEG XL
Liquid-DSP
ViennaCL
ASTC Encoder
ViennaCL
Zstd Compression
Botan
oneDNN
srsLTE
LuaRadio:
  Complex Phase
  FM Deemphasis Filter
  Five Back to Back FIR Filters
toyBrot Fractal Generator
oneDNN:
  IP Shapes 3D - f32 - CPU
  Deconvolution Batch shapes_3d - bf16bf16bf16 - CPU
Liquid-DSP
oneDNN
Botan
ViennaCL
Pennant
SVT-HEVC
oneDNN:
  IP Shapes 1D - u8s8f32 - CPU
  Matrix Multiply Batch Shapes Transformer - bf16bf16bf16 - CPU
Pennant
Zstd Compression
Botan
oneDNN:
  IP Shapes 1D - bf16bf16bf16 - CPU
  Deconvolution Batch shapes_1d - u8s8f32 - CPU
  Convolution Batch Shapes Auto - f32 - CPU
AOM AV1
Liquid-DSP
Botan
Liquid-DSP
Botan:
  AES-256
  Blowfish
AOM AV1
ASTC Encoder
oneDNN
Basis Universal
Timed Erlang/OTP Compilation
oneDNN
GROMACS
Botan:
  AES-256 - Decrypt
  Blowfish - Decrypt
simdjson
Basis Universal
srsLTE
Botan
Basis Universal
Zstd Compression
oneDNN
Botan
Xcompact3d Incompact3d
Timed Node.js Compilation
LuaRadio
toyBrot Fractal Generator
Botan
ASTC Encoder
Liquid-DSP
GNU GMP GMPbench
AOM AV1:
  Speed 4 Two-Pass - Bosphorus 1080p
  Speed 0 Two-Pass - Bosphorus 1080p
  Speed 0 Two-Pass - Bosphorus 4K
simdjson:
  PartialTweets
  LargeRand
oneDNN
GNU Radio:
  FM Deemphasis Filter
  IIR Filter

1

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: intel_cpufreq schedutil - CPU Microcode: 0x5003003
Python Notes: Python 3.8.6
Security Notes: itlb_multihit: KVM: Mitigation of VMX disabled + l1tf: Not affected + mds: Not affected + meltdown: Not affected + 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 Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Mitigation of TSX disabled

Testing initiated at 27 March 2021 18:28 by user pts.

2

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: intel_cpufreq schedutil - CPU Microcode: 0x5003003
Python Notes: Python 3.8.6
Security Notes: itlb_multihit: KVM: Mitigation of VMX disabled + l1tf: Not affected + mds: Not affected + meltdown: Not affected + 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 Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Mitigation of TSX disabled

Testing initiated at 27 March 2021 20:44 by user pts.

3

Processor: Intel Xeon Gold 6226R @ 3.90GHz (16 Cores / 32 Threads), Motherboard: Supermicro X11SPL-F v1.02 (3.1 BIOS), Chipset: Intel Sky Lake-E DMI3 Registers, Memory: 188GB, Disk: 280GB INTEL SSDPED1D280GA, Graphics: ASPEED, Network: 2 x Intel I210

OS: Ubuntu 20.10, Kernel: 5.11.0-rc4-max-boost-inv-patch (x86_64) 20210121, Desktop: GNOME Shell 3.38.1, Display Server: X Server 1.20.9, Compiler: GCC 10.2.0, File-System: ext4, Screen Resolution: 1024x768

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: intel_cpufreq schedutil - CPU Microcode: 0x5003003
Python Notes: Python 3.8.6
Security Notes: itlb_multihit: KVM: Mitigation of VMX disabled + l1tf: Not affected + mds: Not affected + meltdown: Not affected + 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 Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Mitigation of TSX disabled

Testing initiated at 28 March 2021 05:11 by user pts.