Ubuntu 21.10 EPYC Milan

AMD EPYC 74F3 24-Core testing with a ASRockRack ROME2D16-2T (P3.10 BIOS) and ASPEED on Ubuntu 21.04 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2110131-TJ-2110124TJ87
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
Chess Test Suite 2 Tests
Timed Code Compilation 4 Tests
C/C++ Compiler Tests 14 Tests
CPU Massive 30 Tests
Creator Workloads 22 Tests
Cryptocurrency Benchmarks, CPU Mining Tests 2 Tests
Cryptography 4 Tests
Database Test Suite 5 Tests
Encoding 7 Tests
Fortran Tests 3 Tests
Game Development 4 Tests
Go Language Tests 2 Tests
HPC - High Performance Computing 16 Tests
Imaging 2 Tests
Common Kernel Benchmarks 5 Tests
Machine Learning 8 Tests
Molecular Dynamics 5 Tests
MPI Benchmarks 4 Tests
Multi-Core 34 Tests
NVIDIA GPU Compute 8 Tests
Intel oneAPI 4 Tests
OpenCL 2 Tests
OpenMPI Tests 7 Tests
Programmer / Developer System Benchmarks 5 Tests
Python Tests 6 Tests
Renderers 7 Tests
Scientific Computing 5 Tests
Server 9 Tests
Server CPU Tests 20 Tests
Single-Threaded 2 Tests
Video Encoding 7 Tests
Common Workstation Benchmarks 6 Tests

Statistics

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

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Prefer Vertical Bar Graphs

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
Ubuntu 21.10
October 10 2021
  18 Hours, 7 Minutes
Ubuntu 21.10 + Linux 5.15
October 11 2021
  17 Hours, 4 Minutes
Ubuntu 21.04
October 12 2021
  17 Hours, 42 Minutes
Invert Hiding All Results Option
  17 Hours, 38 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):


Ubuntu 21.10 EPYC MilanProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerVulkanCompilerFile-SystemScreen ResolutionUbuntu 21.10Ubuntu 21.10 + Linux 5.15Ubuntu 21.04AMD EPYC 74F3 24-Core @ 3.20GHz (24 Cores / 48 Threads)ASRockRack ROME2D16-2T (P3.10 BIOS)AMD Starship/Matisse64GB1000GB Western Digital WD_BLACK SN850 1TBASPEEDAMD Starship/MatisseVE2282 x Intel 10G X550TUbuntu 21.105.13.0-16-generic (x86_64)GNOME Shell 40.5X Server1.1.182GCC 11.2.0ext41920x10805.15.0-051500rc5daily20211011-generic (x86_64) 20211010Ubuntu 21.045.11.0-37-generic (x86_64)1.0.2GCC 10.3.0OpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler Details- Ubuntu 21.10: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-ZPT0kp/gcc-11-11.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-ZPT0kp/gcc-11-11.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v - Ubuntu 21.10 + Linux 5.15: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-ZPT0kp/gcc-11-11.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-ZPT0kp/gcc-11-11.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v - Ubuntu 21.04: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --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-link-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/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-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Processor Details- Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0xa001114Java Details- Ubuntu 21.10: OpenJDK Runtime Environment (build 11.0.12+7-Ubuntu-0ubuntu3)- Ubuntu 21.10 + Linux 5.15: OpenJDK Runtime Environment (build 11.0.12+7-Ubuntu-0ubuntu3)- Ubuntu 21.04: OpenJDK Runtime Environment (build 11.0.11+9-Ubuntu-0ubuntu2)Python Details- Ubuntu 21.10: Python 3.9.7- Ubuntu 21.10 + Linux 5.15: Python 3.9.7- Ubuntu 21.04: Python 3.9.5Security Details- itlb_multihit: Not affected + 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 Full AMD retpoline IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected

Ubuntu 21.10Ubuntu 21.10 + Linux 5.15Ubuntu 21.04Result OverviewPhoronix Test Suite100%104%108%112%116%PostgreSQL pgbenchStress-NGONNX RuntimeGoogle SynthMarknginxMobile Neural NetworkMariaDBCpuminer-OptVP9 libvpx Encodinglibavif avifencApache HTTP ServerPlaidMLLeelaChessZeroNCNNJPEG XL libjxlNAS Parallel BenchmarksTNNXcompact3d Incompact3dTimed Godot Game Engine CompilationTTSIOD 3D RendererRedisTimed Linux Kernel CompilationLAMMPS Molecular Dynamics SimulatorOpenFOAMTimed Node.js Compilation7-Zip CompressionSVT-VP9SecureMarkSVT-AV1Chaos Group V-RAYx265LuxCoreRenderSVT-HEVCsimdjsonASTC EncoderRodiniaTimed LLVM CompilationIntel Open Image DenoiseGROMACSFacebook RocksDBoneDNNApache CassandraAOM AV1OSPrayAppleseedXmrigStockfishBRL-CADBlenderOpenSSLEmbreeNAMDSysbenchTensorFlow LiteNatron

Ubuntu 21.10 EPYC Milanstress-ng: IO_uringcpuminer-opt: Garlicoinstress-ng: System V Message Passingstress-ng: NUMAstress-ng: Socket Activitypgbench: 100 - 250 - Read Write - Average Latencypgbench: 100 - 250 - Read Writestress-ng: Forkingpgbench: 100 - 500 - Read Writepgbench: 100 - 500 - Read Write - Average Latencycpuminer-opt: Myriad-Groestlstress-ng: Context Switchingavifenc: 10nginx: 20stress-ng: MMAPapache: 20nginx: 100nginx: 200nginx: 500npb: IS.Dnginx: 1000stress-ng: Vector Mathrodinia: OpenMP Leukocytemnn: MobileNetV2_224synthmark: VoiceMark_100mnn: inception-v3mnn: SqueezeNetV1.0avifenc: 10, Losslessplaidml: No - Inference - VGG16 - CPUapache: 100lczero: BLASmysqlslap: 512redis: GETapache: 1000vpxenc: Speed 5 - Bosphorus 4Kx265: Bosphorus 4Kcpuminer-opt: Blake-2 Sstress-ng: MEMFDstress-ng: Glibc Qsort Data Sortingplaidml: No - Inference - VGG19 - CPUnpb: CG.Cvpxenc: Speed 0 - Bosphorus 4Kmnn: squeezenetv1.1rodinia: OpenMP LavaMDjpegxl: JPEG - 7stress-ng: Malloctnn: CPU - MobileNet v2avifenc: 6ncnn: CPU-v3-v3 - mobilenet-v3pgbench: 100 - 500 - Read Only - Average Latencypgbench: 100 - 500 - Read Onlyncnn: CPU - regnety_400mncnn: CPU - resnet50ncnn: CPU - vgg16jpegxl: JPEG - 5ncnn: CPU - resnet18apache: 500mnn: mobilenet-v1-1.0ncnn: CPU - mobilenetapache: 1x265: Bosphorus 1080papache: 200incompact3d: input.i3d 129 Cells Per Directionstress-ng: SENDFILEnginx: 1stress-ng: CPU Stressncnn: CPU - yolov4-tinysimdjson: PartialTweetssimdjson: LargeRandnpb: MG.Clammps: Rhodopsin Proteintensorflow-lite: NASNet Mobileonnx: shufflenet-v2-10 - OpenMP CPUncnn: CPU - googlenetnpb: LU.Cjpegxl: PNG - 8ospray: NASA Streamlines - SciVisjpegxl: JPEG - 8tnn: CPU - SqueezeNet v2ncnn: CPU - squeezenet_ssdjpegxl: PNG - 7jpegxl: PNG - 5openfoam: Motorbike 30Mncnn: CPU - efficientnet-b0stress-ng: CPU Cachencnn: CPU - mnasnetstress-ng: Glibc C String Functionstnn: CPU - DenseNetsimdjson: DistinctUserIDbuild-godot: Time To Compilettsiod-renderer: Phong Rendering With Soft-Shadow Mappingaom-av1: Speed 8 Realtime - Bosphorus 4Kbuild-linux-kernel: Time To Compilenpb: SP.Cnpb: SP.Bonnx: yolov4 - OpenMP CPUrodinia: OpenMP CFD Solverastcenc: Thoroughbuild-nodejs: Time To Compileavifenc: 0stress-ng: Matrix Mathavifenc: 2cpuminer-opt: LBC, LBRY Creditsstress-ng: Memory Copyingonnx: fcn-resnet101-11 - OpenMP CPUavifenc: 6, Losslesscompress-7zip: Compress Speed Testpgbench: 1000 - 500 - Read Onlypgbench: 1000 - 500 - Read Only - Average Latencyluxcorerender: LuxCore Benchmark - CPUrocksdb: Read Rand Write Randsvt-vp9: VMAF Optimized - Bosphorus 1080ppgbench: 100 - 250 - Read Onlysimdjson: Kostyaluxcorerender: Danish Mood - CPUsecuremark: SecureMark-TLSncnn: CPU-v2-v2 - mobilenet-v2pgbench: 100 - 250 - Read Only - Average Latencytensorflow-lite: Mobilenet Floatlammps: 20k Atomssvt-av1: Preset 8 - Bosphorus 4Ksvt-hevc: 10 - Bosphorus 1080pcpuminer-opt: x25xv-ray: CPUplaidml: No - Inference - ResNet 50 - CPUrocksdb: Update Randnpb: FT.Clczero: Eigenonednn: Recurrent Neural Network Inference - f32 - CPUtensorflow-lite: Mobilenet Quantstress-ng: Cryptostress-ng: Atomicaom-av1: Speed 6 Two-Pass - Bosphorus 4Kincompact3d: input.i3d 193 Cells Per Directionncnn: CPU - shufflenet-v2build-llvm: Ninjabuild-llvm: Unix Makefilespgbench: 1000 - 250 - Read Only - Average Latencysvt-hevc: 7 - Bosphorus 1080ppgbench: 1000 - 250 - Read Onlyopenfoam: Motorbike 60Monednn: Recurrent Neural Network Training - f32 - CPUrocksdb: Rand Filloidn: RT.ldr_alb_nrm.3840x2160gromacs: MPI CPU - water_GMX50_bareembree: Pathtracer ISPC - Crownappleseed: Material Testerembree: Pathtracer - Crowntensorflow-lite: Inception V4tensorflow-lite: SqueezeNetcassandra: Writesembree: Pathtracer - Asian Dragonaom-av1: Speed 9 Realtime - Bosphorus 4Kluxcorerender: Orange Juice - CPUopenssl: SHA256tensorflow-lite: Inception ResNet V2cpuminer-opt: Deepcoinxmrig: Monero - 1Mcpuminer-opt: Skeincoinluxcorerender: DLSC - CPUcpuminer-opt: Quad SHA-256, Pyriteblender: Barbershop - CPU-Onlystress-ng: Semaphoresstockfish: Total Timexmrig: Wownero - 1Mospray: NASA Streamlines - Path Tracerbrl-cad: VGR Performance Metrictnn: CPU - SqueezeNet v1.1sysbench: RAM / Memoryospray: San Miguel - Path Tracerappleseed: Emilyblender: BMW27 - CPU-Onlycpuminer-opt: Triple SHA-256, Onecoinembree: Pathtracer ISPC - Asian Dragoncpuminer-opt: Maginamd: ATPase Simulation - 327,506 Atomsastcenc: Exhaustiveopenssl: RSA4096sysbench: CPUopenssl: RSA4096natron: Spaceshipospray: Magnetic Reconnection - Path Tracerospray: Magnetic Reconnection - SciVisospray: San Miguel - SciVisonnx: super-resolution-10 - OpenMP CPUonnx: bertsquad-10 - OpenMP CPUncnn: CPU - alexnetncnn: CPU - blazefacemnn: resnet-v2-50mnn: mobilenetV3redis: SETpgbench: 1000 - 500 - Read Write - Average Latencypgbench: 1000 - 500 - Read Writepgbench: 1000 - 250 - Read Write - Average Latencypgbench: 1000 - 250 - Read Writecpuminer-opt: Ringcoinluxcorerender: Rainbow Colors and Prism - CPUrodinia: OpenMP Streamclusterrodinia: OpenMP HotSpot3Dparboil: OpenMP LBMUbuntu 21.10Ubuntu 21.10 + Linux 5.15Ubuntu 21.04116668.491977.076620370.07573.2014982.114.6835339869854.874643010.7721231511961807.373.497192582.12550.0767345.50178550.68180031.89180523.871993.83178900.55142582.1244.6013.837814.64026.6086.2495.73027.13110031.1514577032684890.00128080.8217.5126.3613351781046.21348.0922.2024114.117.464.72685.06273.99619879641.77282.1319.7225.660.67574029614.9416.4325.5174.559.26128306.912.71813.596319.9054.72133795.466.90810013532284.3252710.0369408.5921.124.171.1150332.6416.35089415.11715613.0382765.180.9650.5328.9462.60916.179.3672.0024.227.07105.975.133684070.292850.2364.9356.360969.49431.2836.73135556.4463598.144289.5416.5677153.71445.120102691.8925.023690708094.8016728.8601721176393610.7823.942687791446.627394003.213.612696885.730.33855376.016.87419.707553.02768.202782510.1072776742915.031330777.26259851.76537.91196373.377.3030.38514146.21235.750274.2640.409281.48611633243.111429.317774210.863.25928.4922147.89729530.7634122457382510.823031834.156646.807.053616427026310803202360018217.52085104.74214267229.143625859.107563107119059.610.53349508251.85210278.233.11162.67218559.7950655331.46471221.380.7772524.06086520.2114579.76425838.45.150028.5737.5150566485.562.9720.7112.3752148023.0215.886317506.349401173390.4614.116.55879.403493315.901927.239796819.42584.4619318.775.9494204673798.883859412.9581243713907201.313.197220753.31481.3576761.54202016.07202847.39203360.662013.11200086.32142690.5544.3543.604816.31026.3815.9625.39028.99120148.1113406472472940.8135459.4316.9726.5812925251130.04347.7523.9225960.117.424.44184.09979.20622292053.40280.3389.4185.310.63978287614.5015.4524.7079.048.74134586.752.69412.846116.7755.20141087.746.55719979521315.0652530.1169941.3420.154.181.1152547.8815.90586693.41770412.4884921.350.9852.1129.9961.87515.699.7474.9224.216.81105.965.013705740.262837.5324.9456.081957.38730.9636.46636599.6065350.974369.7016.5450153.04445.073105285.8524.518691638141.9817128.3581751656247940.8003.992748440444.467482633.213.692723105.610.33454956.117.00219.488556.42773.602830510.2974134343590.571306766.69058918.06539.84197807.927.3529.89442006.13234.950275.4920.403283.51620425239.681412.997724680.853.29728.2755147.98177330.9509121790381889.123155434.383546.737.093628985227010761402345018333.42094004.77213690227.943635712.187530436919138.510.53349004251.44310249.843.10163.06826859.9150804731.40441223.670.7773024.04576529.1114562.03426052.95.150028.5737.5159147514.982.7421.0212.3702195897.7924.174206889.444264743378.7814.106.69580.312257058.345759.045570866.82825.7714625.694.8315177359944.754518611.06610356.4713696558.183.677194938.96505.6269310.02181517.37182227.85180505.991790.31178392.68127352.2849.5794.018904.31029.0876.5295.90026.53112629.8414426602633469.20124932.6416.1528.5612358341054.74322.5822.4725140.106.944.77179.18976.30663350388.20263.92710.0515.450.68173469415.4316.0726.2476.028.91127031.042.85413.126466.3752.30134421.216.85141821549103.7155249.4066584.1620.784.371.0650207.1316.63785579.21792312.6281382.280.945028.7860.13116.339.7372.7825.157.02109.975.193580441.822755.1684.7857.898988.22631.9535.65935627.7163522.544409.4476.7181157.05746.226103521.2025.136708108289.3217129.0311711526392940.7824.032748890436.867319443.143.672755455.720.34154254.017.21419.318545.96782.732778110.2173824043707.661325763.62258811.96652.82199708.117.2330.29434016.23238.515278.3830.408279.49612473242.991411.477679940.863.28228.5521146.60223531.0509121353381775.222950734.081447.137.033601417116710725632353318245.22080834.76215003227.983642449.467531632619125.810.49350303250.98010283.193.11162.55411659.7350669731.38571224.030.7760824.08346530.3114490.11425938.65.150028.5737.5163667944.992.8023.0192.6172070185.3822.853225665.961424013321.2314.756.29278.192OpenBenchmarking.org

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: IO_uringUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04110K220K330K440K550KSE +/- 5001.47, N = 6SE +/- 1552.41, N = 3SE +/- 492.77, N = 3493315.90116668.49257058.341. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: IO_uringUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0490K180K270K360K450KMin: 472835.24 / Avg: 493315.9 / Max: 508855.71Min: 113812.95 / Avg: 116668.49 / Max: 119151.87Min: 256457.69 / Avg: 257058.34 / Max: 258035.341. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

Cpuminer-Opt

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

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: GarlicoinUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0412002400360048006000SE +/- 25.22, N = 3SE +/- 9.20, N = 3SE +/- 4.21, N = 31927.231977.075759.041. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: GarlicoinUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0410002000300040005000Min: 1881.68 / Avg: 1927.23 / Max: 1968.76Min: 1967.28 / Avg: 1977.07 / Max: 1995.45Min: 5753.14 / Avg: 5759.04 / Max: 5767.181. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: System V Message PassingUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042M4M6M8M10MSE +/- 6861.74, N = 3SE +/- 3955.95, N = 3SE +/- 12373.30, N = 39796819.426620370.075570866.821. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: System V Message PassingUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042M4M6M8M10MMin: 9783262.65 / Avg: 9796819.42 / Max: 9805444.7Min: 6612553.52 / Avg: 6620370.07 / Max: 6625338.92Min: 5552572.37 / Avg: 5570866.82 / Max: 5594445.971. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: NUMAUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042004006008001000SE +/- 0.96, N = 3SE +/- 2.46, N = 3SE +/- 0.76, N = 3584.46573.20825.771. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: NUMAUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04150300450600750Min: 582.57 / Avg: 584.46 / Max: 585.69Min: 570.43 / Avg: 573.2 / Max: 578.1Min: 824.31 / Avg: 825.77 / Max: 826.871. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Socket ActivityUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.044K8K12K16K20KSE +/- 66.97, N = 3SE +/- 20.77, N = 3SE +/- 0.85, N = 319318.7714982.1114625.691. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Socket ActivityUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043K6K9K12K15KMin: 19198.11 / Avg: 19318.77 / Max: 19429.44Min: 14945.11 / Avg: 14982.11 / Max: 15016.98Min: 14623.99 / Avg: 14625.69 / Max: 14626.651. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

PostgreSQL pgbench

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

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 250 - Mode: Read Write - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041.33852.6774.01555.3546.6925SE +/- 0.064, N = 5SE +/- 0.055, N = 3SE +/- 0.053, N = 45.9494.6834.831-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 250 - Mode: Read Write - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 5.8 / Avg: 5.95 / Max: 6.17Min: 4.58 / Avg: 4.68 / Max: 4.75Min: 4.67 / Avg: 4.83 / Max: 4.891. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 250 - Mode: Read WriteUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0411K22K33K44K55KSE +/- 445.65, N = 5SE +/- 632.60, N = 3SE +/- 581.80, N = 4420465339851773-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 250 - Mode: Read WriteUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.049K18K27K36K45KMin: 40502.14 / Avg: 42045.56 / Max: 43117.81Min: 52615.71 / Avg: 53397.84 / Max: 54650.15Min: 51122.75 / Avg: 51773.41 / Max: 53513.871. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: ForkingUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0416K32K48K64K80KSE +/- 284.38, N = 3SE +/- 113.04, N = 3SE +/- 497.75, N = 373798.8869854.8759944.751. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: ForkingUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0413K26K39K52K65KMin: 73230.43 / Avg: 73798.88 / Max: 74099.61Min: 69633.21 / Avg: 69854.87 / Max: 70004.2Min: 59433.46 / Avg: 59944.75 / Max: 60940.121. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

PostgreSQL pgbench

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

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 500 - Mode: Read WriteUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0410K20K30K40K50KSE +/- 429.40, N = 3SE +/- 522.25, N = 3SE +/- 214.88, N = 3385944643045186-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 500 - Mode: Read WriteUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.048K16K24K32K40KMin: 37760.43 / Avg: 38593.9 / Max: 39189.97Min: 45398.04 / Avg: 46430 / Max: 47085.74Min: 44757.74 / Avg: 45186.36 / Max: 45427.771. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 500 - Mode: Read Write - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215SE +/- 0.15, N = 3SE +/- 0.12, N = 3SE +/- 0.05, N = 312.9610.7711.07-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 500 - Mode: Read Write - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620Min: 12.76 / Avg: 12.96 / Max: 13.24Min: 10.62 / Avg: 10.77 / Max: 11.01Min: 11.01 / Avg: 11.07 / Max: 11.171. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

Cpuminer-Opt

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

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: Myriad-GroestlUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043K6K9K12K15KSE +/- 98.67, N = 15SE +/- 100.41, N = 15SE +/- 86.06, N = 1512437.0012315.0010356.471. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: Myriad-GroestlUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042K4K6K8K10KMin: 11910 / Avg: 12437.33 / Max: 12860Min: 11900 / Avg: 12314.67 / Max: 12820Min: 9849.3 / Avg: 10356.47 / Max: 107401. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Context SwitchingUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043M6M9M12M15MSE +/- 24088.83, N = 3SE +/- 106207.99, N = 3SE +/- 99682.48, N = 313907201.3111961807.3713696558.181. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Context SwitchingUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042M4M6M8M10MMin: 13863305.48 / Avg: 13907201.31 / Max: 13946344.64Min: 11749811.47 / Avg: 11961807.37 / Max: 12079368.68Min: 13535800.82 / Avg: 13696558.18 / Max: 13879052.461. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

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: 10Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.82731.65462.48193.30924.1365SE +/- 0.008, N = 3SE +/- 0.004, N = 3SE +/- 0.010, N = 33.1973.4973.6771. (CXX) g++ options: -O3 -fPIC -lm
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 10Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 3.19 / Avg: 3.2 / Max: 3.21Min: 3.49 / Avg: 3.5 / Max: 3.5Min: 3.67 / Avg: 3.68 / Max: 3.71. (CXX) g++ options: -O3 -fPIC -lm

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the Golang "Bombardier" program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 20Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0450K100K150K200K250KSE +/- 512.02, N = 3SE +/- 250.97, N = 3SE +/- 278.11, N = 3220753.31192582.12194938.96-ldl -lpthread1. (CC) gcc options: -lcrypt -lz -O3 -march=native
OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 20Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KMin: 219874.61 / Avg: 220753.31 / Max: 221648.09Min: 192093.56 / Avg: 192582.12 / Max: 192926.11Min: 194651.88 / Avg: 194938.96 / Max: 195495.081. (CC) gcc options: -lcrypt -lz -O3 -march=native

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: MMAPUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04120240360480600SE +/- 1.10, N = 3SE +/- 0.60, N = 3SE +/- 1.43, N = 3481.35550.07505.621. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: MMAPUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04100200300400500Min: 479.15 / Avg: 481.35 / Max: 482.51Min: 549.41 / Avg: 550.07 / Max: 551.27Min: 503.89 / Avg: 505.62 / Max: 508.451. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

Apache HTTP Server

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

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 20Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0416K32K48K64K80KSE +/- 580.65, N = 3SE +/- 99.66, N = 3SE +/- 159.41, N = 376761.5467345.5069310.02-pthread1. (CC) gcc options: -shared -fPIC -O2
OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 20Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0413K26K39K52K65KMin: 75693.94 / Avg: 76761.54 / Max: 77691.1Min: 67223.24 / Avg: 67345.5 / Max: 67542.97Min: 69007.39 / Avg: 69310.02 / Max: 69548.221. (CC) gcc options: -shared -fPIC -O2

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the Golang "Bombardier" program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 100Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KSE +/- 409.77, N = 3SE +/- 366.10, N = 3SE +/- 219.93, N = 3202016.07178550.68181517.37-ldl -lpthread1. (CC) gcc options: -lcrypt -lz -O3 -march=native
OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 100Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KMin: 201196.54 / Avg: 202016.07 / Max: 202427.67Min: 177944.09 / Avg: 178550.68 / Max: 179209.12Min: 181226.56 / Avg: 181517.37 / Max: 181948.581. (CC) gcc options: -lcrypt -lz -O3 -march=native

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 200Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KSE +/- 239.94, N = 3SE +/- 140.36, N = 3SE +/- 184.91, N = 3202847.39180031.89182227.85-ldl -lpthread1. (CC) gcc options: -lcrypt -lz -O3 -march=native
OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 200Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KMin: 202499.38 / Avg: 202847.39 / Max: 203307.55Min: 179828.14 / Avg: 180031.89 / Max: 180300.99Min: 181947.58 / Avg: 182227.85 / Max: 182576.941. (CC) gcc options: -lcrypt -lz -O3 -march=native

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 500Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KSE +/- 271.45, N = 3SE +/- 236.63, N = 3SE +/- 1811.59, N = 3203360.66180523.87180505.99-ldl -lpthread1. (CC) gcc options: -lcrypt -lz -O3 -march=native
OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 500Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KMin: 202831.58 / Avg: 203360.66 / Max: 203730.64Min: 180273.16 / Avg: 180523.87 / Max: 180996.84Min: 176886.89 / Avg: 180505.99 / Max: 182464.381. (CC) gcc options: -lcrypt -lz -O3 -march=native

NAS Parallel Benchmarks

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

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: IS.DUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04400800120016002000SE +/- 3.59, N = 3SE +/- 6.66, N = 3SE +/- 6.56, N = 32013.111993.831790.31-pthread -ldl -lutil -lrt1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0
OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: IS.DUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04400800120016002000Min: 2007.87 / Avg: 2013.11 / Max: 2019.99Min: 1980.69 / Avg: 1993.83 / Max: 2002.34Min: 1777.38 / Avg: 1790.31 / Max: 1798.691. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the Golang "Bombardier" program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 1000Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KSE +/- 177.77, N = 3SE +/- 175.84, N = 3SE +/- 904.31, N = 3200086.32178900.55178392.68-ldl -lpthread1. (CC) gcc options: -lcrypt -lz -O3 -march=native
OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 1000Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430K60K90K120K150KMin: 199901.6 / Avg: 200086.32 / Max: 200441.78Min: 178556.18 / Avg: 178900.55 / Max: 179134.48Min: 177018.25 / Avg: 178392.68 / Max: 180098.011. (CC) gcc options: -lcrypt -lz -O3 -march=native

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Vector MathUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430K60K90K120K150KSE +/- 16.18, N = 3SE +/- 12.72, N = 3SE +/- 13.25, N = 3142690.55142582.12127352.281. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Vector MathUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420K40K60K80K100KMin: 142662.17 / Avg: 142690.55 / Max: 142718.22Min: 142558.52 / Avg: 142582.12 / Max: 142602.15Min: 127331.59 / Avg: 127352.28 / Max: 127376.961. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

Rodinia

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

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP LeukocyteUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041122334455SE +/- 0.22, N = 3SE +/- 0.11, N = 3SE +/- 0.30, N = 344.3544.6049.581. (CXX) g++ options: -O2 -lOpenCL
OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP LeukocyteUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041020304050Min: 43.98 / Avg: 44.35 / Max: 44.74Min: 44.48 / Avg: 44.6 / Max: 44.82Min: 49.01 / Avg: 49.58 / Max: 50.011. (CXX) g++ options: -O2 -lOpenCL

Mobile Neural Network

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

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: MobileNetV2_224Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.90411.80822.71233.61644.5205SE +/- 0.032, N = 14SE +/- 0.032, N = 15SE +/- 0.022, N = 153.6043.8374.018MIN: 3.38 / MAX: 7.24MIN: 3.58 / MAX: 4.5MIN: 3.82 / MAX: 4.551. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl
OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: MobileNetV2_224Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 3.42 / Avg: 3.6 / Max: 3.86Min: 3.63 / Avg: 3.84 / Max: 4.09Min: 3.85 / Avg: 4.02 / Max: 4.191. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Google SynthMark

SynthMark is a cross platform tool for benchmarking CPU performance under a variety of real-time audio workloads. It uses a polyphonic synthesizer model to provide standardized tests for latency, jitter and computational throughput. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgVoices, More Is BetterGoogle SynthMark 20201109Test: VoiceMark_100Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042004006008001000SE +/- 0.47, N = 3SE +/- 1.66, N = 3SE +/- 1.42, N = 3816.31814.64904.311. (CXX) g++ options: -lm -lpthread -std=c++11 -Ofast
OpenBenchmarking.orgVoices, More Is BetterGoogle SynthMark 20201109Test: VoiceMark_100Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04160320480640800Min: 815.39 / Avg: 816.31 / Max: 816.9Min: 811.33 / Avg: 814.64 / Max: 816.43Min: 901.54 / Avg: 904.31 / Max: 906.251. (CXX) g++ options: -lm -lpthread -std=c++11 -Ofast

Mobile Neural Network

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

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: inception-v3Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835SE +/- 0.30, N = 14SE +/- 0.33, N = 15SE +/- 0.26, N = 1526.3826.6129.09MIN: 23.92 / MAX: 31.4MIN: 24.29 / MAX: 34.57MIN: 26.58 / MAX: 30.751. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl
OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: inception-v3Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430Min: 24.11 / Avg: 26.38 / Max: 28.28Min: 24.57 / Avg: 26.61 / Max: 28.7Min: 26.78 / Avg: 29.09 / Max: 30.281. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: SqueezeNetV1.0Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810SE +/- 0.044, N = 14SE +/- 0.054, N = 15SE +/- 0.048, N = 155.9626.2496.529MIN: 5.6 / MAX: 9.22MIN: 5.67 / MAX: 11.64MIN: 6.09 / MAX: 7.561. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl
OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: SqueezeNetV1.0Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 5.68 / Avg: 5.96 / Max: 6.29Min: 5.81 / Avg: 6.25 / Max: 6.65Min: 6.18 / Avg: 6.53 / Max: 6.931. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

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, LosslessUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041.32752.6553.98255.316.6375SE +/- 0.014, N = 3SE +/- 0.018, N = 3SE +/- 0.003, N = 35.3905.7305.9001. (CXX) g++ options: -O3 -fPIC -lm
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 10, LosslessUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 5.37 / Avg: 5.39 / Max: 5.42Min: 5.7 / Avg: 5.73 / Max: 5.75Min: 5.89 / Avg: 5.9 / Max: 5.91. (CXX) g++ options: -O3 -fPIC -lm

PlaidML

This test profile uses PlaidML deep learning framework developed by Intel for offering up various benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: VGG16 - Device: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835SE +/- 0.23, N = 3SE +/- 0.09, N = 3SE +/- 0.30, N = 328.9927.1326.53
OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: VGG16 - Device: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430Min: 28.52 / Avg: 28.99 / Max: 29.25Min: 26.95 / Avg: 27.13 / Max: 27.23Min: 25.94 / Avg: 26.53 / Max: 26.87

Apache HTTP Server

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

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 100Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430K60K90K120K150KSE +/- 464.12, N = 3SE +/- 282.17, N = 3SE +/- 919.33, N = 3120148.11110031.15112629.84-pthread1. (CC) gcc options: -shared -fPIC -O2
OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 100Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420K40K60K80K100KMin: 119237.13 / Avg: 120148.11 / Max: 120757.94Min: 109503.43 / Avg: 110031.15 / Max: 110468.18Min: 111449.57 / Avg: 112629.84 / Max: 114440.931. (CC) gcc options: -shared -fPIC -O2

LeelaChessZero

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

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: BLASUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430060090012001500SE +/- 14.18, N = 9SE +/- 12.03, N = 3SE +/- 8.33, N = 31340145714421. (CXX) g++ options: -flto -pthread
OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: BLASUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430060090012001500Min: 1290 / Avg: 1339.56 / Max: 1404Min: 1434 / Avg: 1456.67 / Max: 1475Min: 1426 / Avg: 1442 / Max: 14541. (CXX) g++ options: -flto -pthread

MariaDB

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

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 10.6.4Clients: 512Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04150300450600750SE +/- 6.28, N = 3SE +/- 1.91, N = 3SE +/- 2.61, N = 3647703660-pthread1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -llzma -lbz2 -lsnappy -lnuma -lpcre2-8 -lcrypt -laio -lz -lm -lssl -lcrypto -lpthread -ldl
OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 10.6.4Clients: 512Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04120240360480600Min: 638.16 / Avg: 647.09 / Max: 659.2Min: 701.26 / Avg: 703.41 / Max: 707.21Min: 655.31 / Avg: 660.38 / Max: 664.011. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -llzma -lbz2 -lsnappy -lnuma -lpcre2-8 -lcrypt -laio -lz -lm -lssl -lcrypto -lpthread -ldl

Redis

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

OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 6.0.9Test: GETUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04600K1200K1800K2400K3000KSE +/- 18296.40, N = 3SE +/- 8561.06, N = 3SE +/- 24084.60, N = 152472940.802684890.002633469.201. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3
OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 6.0.9Test: GETUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04500K1000K1500K2000K2500KMin: 2445585.5 / Avg: 2472940.83 / Max: 2507667Min: 2672419 / Avg: 2684890 / Max: 2701285.75Min: 2505018 / Avg: 2633469.2 / Max: 2824079.251. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3

Apache HTTP Server

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

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 1000Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430K60K90K120K150KSE +/- 224.44, N = 3SE +/- 23.56, N = 3SE +/- 1392.94, N = 4135459.43128080.82124932.64-pthread1. (CC) gcc options: -shared -fPIC -O2
OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 1000Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420K40K60K80K100KMin: 135020.6 / Avg: 135459.43 / Max: 135760.65Min: 128033.71 / Avg: 128080.82 / Max: 128105.36Min: 121816.03 / Avg: 124932.64 / Max: 128303.471. (CC) gcc options: -shared -fPIC -O2

VP9 libvpx Encoding

This is a standard video encoding performance test of Google's libvpx library and the vpxenc command for the VP9 video format. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVP9 libvpx Encoding 1.10.0Speed: Speed 5 - Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620SE +/- 0.07, N = 3SE +/- 0.15, N = 3SE +/- 0.11, N = 316.9717.5116.151. (CXX) g++ options: -m64 -lm -lpthread -O3 -fPIC -U_FORTIFY_SOURCE -std=gnu++11
OpenBenchmarking.orgFrames Per Second, More Is BetterVP9 libvpx Encoding 1.10.0Speed: Speed 5 - Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620Min: 16.85 / Avg: 16.97 / Max: 17.08Min: 17.34 / Avg: 17.51 / Max: 17.8Min: 15.99 / Avg: 16.15 / Max: 16.361. (CXX) g++ options: -m64 -lm -lpthread -O3 -fPIC -U_FORTIFY_SOURCE -std=gnu++11

x265

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

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835SE +/- 0.21, N = 3SE +/- 0.11, N = 3SE +/- 0.27, N = 326.5826.3628.561. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430Min: 26.17 / Avg: 26.58 / Max: 26.86Min: 26.15 / Avg: 26.36 / Max: 26.49Min: 28.03 / Avg: 28.56 / Max: 28.941. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

Cpuminer-Opt

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

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: Blake-2 SUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04300K600K900K1200K1500KSE +/- 12480.44, N = 15SE +/- 14384.10, N = 5SE +/- 14341.65, N = 151292525133517812358341. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: Blake-2 SUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04200K400K600K800K1000KMin: 1218370 / Avg: 1292525.33 / Max: 1381440Min: 1281300 / Avg: 1335178 / Max: 1363620Min: 1123600 / Avg: 1235834 / Max: 13425101. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: MEMFDUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042004006008001000SE +/- 0.46, N = 3SE +/- 2.37, N = 3SE +/- 0.45, N = 31130.041046.211054.741. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: MEMFDUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042004006008001000Min: 1129.14 / Avg: 1130.04 / Max: 1130.66Min: 1042.77 / Avg: 1046.21 / Max: 1050.75Min: 1053.86 / Avg: 1054.74 / Max: 1055.321. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Glibc Qsort Data SortingUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0480160240320400SE +/- 0.23, N = 3SE +/- 0.08, N = 3SE +/- 0.75, N = 3347.75348.09322.581. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Glibc Qsort Data SortingUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0460120180240300Min: 347.43 / Avg: 347.75 / Max: 348.2Min: 347.97 / Avg: 348.09 / Max: 348.23Min: 321.1 / Avg: 322.58 / Max: 323.571. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

PlaidML

This test profile uses PlaidML deep learning framework developed by Intel for offering up various benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: VGG19 - Device: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430SE +/- 0.30, N = 3SE +/- 0.12, N = 3SE +/- 0.12, N = 323.9222.2022.47
OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: VGG19 - Device: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430Min: 23.49 / Avg: 23.92 / Max: 24.51Min: 21.96 / Avg: 22.2 / Max: 22.33Min: 22.33 / Avg: 22.47 / Max: 22.72

NAS Parallel Benchmarks

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

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: CG.CUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.046K12K18K24K30KSE +/- 235.98, N = 3SE +/- 158.46, N = 3SE +/- 24.29, N = 325960.1124114.1125140.10-pthread -ldl -lutil -lrt1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0
OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: CG.CUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.045K10K15K20K25KMin: 25518.89 / Avg: 25960.11 / Max: 26325.8Min: 23799 / Avg: 24114.11 / Max: 24301.01Min: 25098.44 / Avg: 25140.1 / Max: 25182.561. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0

VP9 libvpx Encoding

This is a standard video encoding performance test of Google's libvpx library and the vpxenc command for the VP9 video format. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVP9 libvpx Encoding 1.10.0Speed: Speed 0 - Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810SE +/- 0.08, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 37.427.466.941. (CXX) g++ options: -m64 -lm -lpthread -O3 -fPIC -U_FORTIFY_SOURCE -std=gnu++11
OpenBenchmarking.orgFrames Per Second, More Is BetterVP9 libvpx Encoding 1.10.0Speed: Speed 0 - Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 7.27 / Avg: 7.42 / Max: 7.51Min: 7.44 / Avg: 7.46 / Max: 7.5Min: 6.89 / Avg: 6.94 / Max: 6.971. (CXX) g++ options: -m64 -lm -lpthread -O3 -fPIC -U_FORTIFY_SOURCE -std=gnu++11

Mobile Neural Network

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

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: squeezenetv1.1Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041.07352.1473.22054.2945.3675SE +/- 0.054, N = 14SE +/- 0.054, N = 15SE +/- 0.050, N = 154.4414.7264.771MIN: 3.94 / MAX: 7.88MIN: 4.13 / MAX: 8.49MIN: 4.19 / MAX: 5.51. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl
OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: squeezenetv1.1Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 4.04 / Avg: 4.44 / Max: 4.79Min: 4.33 / Avg: 4.73 / Max: 5.14Min: 4.26 / Avg: 4.77 / Max: 51. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Rodinia

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

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP LavaMDUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420406080100SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 384.1085.0679.191. (CXX) g++ options: -O2 -lOpenCL
OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP LavaMDUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041632486480Min: 84.07 / Avg: 84.1 / Max: 84.16Min: 85.03 / Avg: 85.06 / Max: 85.11Min: 79.16 / Avg: 79.19 / Max: 79.211. (CXX) g++ options: -O2 -lOpenCL

JPEG XL libjxl

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

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.5Input: JPEG - Encode Speed: 7Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420406080100SE +/- 0.21, N = 3SE +/- 0.18, N = 3SE +/- 0.26, N = 379.2073.9976.30-pthread1. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.5Input: JPEG - Encode Speed: 7Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041530456075Min: 78.8 / Avg: 79.2 / Max: 79.51Min: 73.68 / Avg: 73.99 / Max: 74.3Min: 75.77 / Avg: 76.3 / Max: 76.581. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: MallocUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04140M280M420M560M700MSE +/- 407272.05, N = 3SE +/- 310818.96, N = 3SE +/- 2413197.53, N = 3622292053.40619879641.77663350388.201. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: MallocUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04110M220M330M440M550MMin: 621807260.51 / Avg: 622292053.4 / Max: 623101321.49Min: 619262996.12 / Avg: 619879641.77 / Max: 620256055.58Min: 658539400.71 / Avg: 663350388.2 / Max: 666089598.081. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

TNN

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

OpenBenchmarking.orgms, Fewer Is BetterTNN 0.3Target: CPU - Model: MobileNet v2Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0460120180240300SE +/- 0.96, N = 3SE +/- 1.74, N = 3SE +/- 0.24, N = 3280.34282.13263.93MIN: 275.48 / MAX: 311.76MIN: 275.4 / MAX: 316.89MIN: 259.63 / MAX: 295.761. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -fvisibility=default -O3 -rdynamic -ldl
OpenBenchmarking.orgms, Fewer Is BetterTNN 0.3Target: CPU - Model: MobileNet v2Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0450100150200250Min: 278.52 / Avg: 280.34 / Max: 281.79Min: 278.9 / Avg: 282.13 / Max: 284.84Min: 263.48 / Avg: 263.93 / Max: 264.31. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -fvisibility=default -O3 -rdynamic -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: 6Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215SE +/- 0.024, N = 3SE +/- 0.018, N = 3SE +/- 0.063, N = 39.4189.72210.0511. (CXX) g++ options: -O3 -fPIC -lm
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 6Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 9.37 / Avg: 9.42 / Max: 9.44Min: 9.69 / Avg: 9.72 / Max: 9.75Min: 9.93 / Avg: 10.05 / Max: 10.141. (CXX) g++ options: -O3 -fPIC -lm

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU-v3-v3 - Model: mobilenet-v3Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041.27352.5473.82055.0946.3675SE +/- 0.02, N = 3SE +/- 0.16, N = 3SE +/- 0.03, N = 35.315.665.45MIN: 5.21 / MAX: 15.93MIN: 5.23 / MAX: 133.29-pthread - MIN: 5.34 / MAX: 5.971. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU-v3-v3 - Model: mobilenet-v3Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 5.28 / Avg: 5.31 / Max: 5.34Min: 5.4 / Avg: 5.66 / Max: 5.96Min: 5.41 / Avg: 5.45 / Max: 5.51. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

PostgreSQL pgbench

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

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 500 - Mode: Read Only - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.15320.30640.45960.61280.766SE +/- 0.003, N = 3SE +/- 0.007, N = 3SE +/- 0.003, N = 30.6390.6750.681-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 500 - Mode: Read Only - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 0.64 / Avg: 0.64 / Max: 0.65Min: 0.67 / Avg: 0.68 / Max: 0.69Min: 0.68 / Avg: 0.68 / Max: 0.691. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 500 - Mode: Read OnlyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04200K400K600K800K1000KSE +/- 3621.04, N = 3SE +/- 7536.29, N = 3SE +/- 2943.65, N = 3782876740296734694-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 500 - Mode: Read OnlyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04140K280K420K560K700KMin: 775698.86 / Avg: 782876.44 / Max: 787300.4Min: 725247.83 / Avg: 740295.57 / Max: 748568.62Min: 730391.21 / Avg: 734693.94 / Max: 740325.251. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: regnety_400mUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620SE +/- 0.04, N = 3SE +/- 0.18, N = 3SE +/- 0.06, N = 314.5014.9415.43MIN: 14.35 / MAX: 17.97MIN: 14.42 / MAX: 114.45-pthread - MIN: 15.28 / MAX: 17.871. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: regnety_400mUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620Min: 14.43 / Avg: 14.5 / Max: 14.56Min: 14.67 / Avg: 14.94 / Max: 15.28Min: 15.36 / Avg: 15.43 / Max: 15.541. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: resnet50Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620SE +/- 0.24, N = 3SE +/- 0.31, N = 3SE +/- 0.06, N = 315.4516.4316.07MIN: 14.99 / MAX: 19.35MIN: 15.63 / MAX: 176.4-pthread - MIN: 15.82 / MAX: 18.831. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: resnet50Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620Min: 15.17 / Avg: 15.45 / Max: 15.94Min: 15.94 / Avg: 16.43 / Max: 17.01Min: 15.96 / Avg: 16.07 / Max: 16.171. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: vgg16Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430SE +/- 0.57, N = 3SE +/- 0.16, N = 3SE +/- 0.09, N = 324.7025.5126.24MIN: 23.74 / MAX: 37.55MIN: 24.52 / MAX: 94.79-pthread - MIN: 25.85 / MAX: 59.931. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: vgg16Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430Min: 24.01 / Avg: 24.7 / Max: 25.84Min: 25.23 / Avg: 25.51 / Max: 25.78Min: 26.1 / Avg: 26.24 / Max: 26.421. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

JPEG XL libjxl

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

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.5Input: JPEG - Encode Speed: 5Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420406080100SE +/- 0.48, N = 3SE +/- 0.17, N = 3SE +/- 0.44, N = 379.0474.5576.02-pthread1. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.5Input: JPEG - Encode Speed: 5Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041530456075Min: 78.3 / Avg: 79.04 / Max: 79.93Min: 74.34 / Avg: 74.55 / Max: 74.9Min: 75.15 / Avg: 76.02 / Max: 76.591. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: resnet18Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215SE +/- 0.22, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 38.749.268.91MIN: 8.38 / MAX: 10.87MIN: 8.7 / MAX: 58.92-pthread - MIN: 8.62 / MAX: 10.751. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: resnet18Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 8.52 / Avg: 8.74 / Max: 9.17Min: 9.17 / Avg: 9.26 / Max: 9.34Min: 8.86 / Avg: 8.91 / Max: 8.941. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Apache HTTP Server

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

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 500Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430K60K90K120K150KSE +/- 303.65, N = 3SE +/- 293.41, N = 3SE +/- 633.46, N = 3134586.75128306.91127031.04-pthread1. (CC) gcc options: -shared -fPIC -O2
OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 500Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420K40K60K80K100KMin: 134278.47 / Avg: 134586.75 / Max: 135194.03Min: 127727.65 / Avg: 128306.91 / Max: 128677.86Min: 126229.82 / Avg: 127031.04 / Max: 128281.561. (CC) gcc options: -shared -fPIC -O2

Mobile Neural Network

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

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: mobilenet-v1-1.0Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.64221.28441.92662.56883.211SE +/- 0.021, N = 14SE +/- 0.024, N = 15SE +/- 0.019, N = 152.6942.7182.854MIN: 2.47 / MAX: 6.09MIN: 2.45 / MAX: 5.04MIN: 2.71 / MAX: 3.091. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl
OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: mobilenet-v1-1.0Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 2.57 / Avg: 2.69 / Max: 2.77Min: 2.51 / Avg: 2.72 / Max: 2.86Min: 2.73 / Avg: 2.85 / Max: 3.011. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: mobilenetUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215SE +/- 0.18, N = 3SE +/- 0.12, N = 3SE +/- 0.04, N = 312.8413.5913.12MIN: 12.4 / MAX: 14.89MIN: 12.98 / MAX: 71.45-pthread - MIN: 12.84 / MAX: 15.231. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: mobilenetUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620Min: 12.64 / Avg: 12.84 / Max: 13.19Min: 13.36 / Avg: 13.59 / Max: 13.75Min: 13.05 / Avg: 13.12 / Max: 13.21. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Apache HTTP Server

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

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 1Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0414002800420056007000SE +/- 56.37, N = 3SE +/- 55.13, N = 3SE +/- 50.44, N = 36116.776319.906466.37-pthread1. (CC) gcc options: -shared -fPIC -O2
OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 1Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0411002200330044005500Min: 6051.47 / Avg: 6116.77 / Max: 6229Min: 6228.76 / Avg: 6319.9 / Max: 6419.21Min: 6365.81 / Avg: 6466.37 / Max: 6523.511. (CC) gcc options: -shared -fPIC -O2

x265

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

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 1080pUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041224364860SE +/- 0.67, N = 3SE +/- 0.62, N = 4SE +/- 0.49, N = 355.2054.7252.301. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 1080pUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041122334455Min: 53.95 / Avg: 55.2 / Max: 56.25Min: 53.44 / Avg: 54.72 / Max: 56.42Min: 51.58 / Avg: 52.3 / Max: 53.241. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

Apache HTTP Server

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

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 200Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430K60K90K120K150KSE +/- 1124.68, N = 3SE +/- 657.97, N = 3SE +/- 315.40, N = 3141087.74133795.46134421.21-pthread1. (CC) gcc options: -shared -fPIC -O2
OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 200Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420K40K60K80K100KMin: 138839.76 / Avg: 141087.74 / Max: 142279.97Min: 132481.36 / Avg: 133795.46 / Max: 134512.73Min: 133831.09 / Avg: 134421.21 / Max: 134909.281. (CC) gcc options: -shared -fPIC -O2

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 DirectionUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810SE +/- 0.05267586, N = 3SE +/- 0.02220467, N = 3SE +/- 0.07087910, N = 56.557199796.908100136.85141821-pthread -ldl -lutil -lrt1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 129 Cells Per DirectionUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 6.48 / Avg: 6.56 / Max: 6.66Min: 6.86 / Avg: 6.91 / Max: 6.93Min: 6.57 / Avg: 6.85 / Max: 6.951. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: SENDFILEUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04120K240K360K480K600KSE +/- 645.89, N = 3SE +/- 68.97, N = 3SE +/- 78.83, N = 3521315.06532284.32549103.711. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: SENDFILEUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04100K200K300K400K500KMin: 520029.33 / Avg: 521315.06 / Max: 522066.05Min: 532192.27 / Avg: 532284.32 / Max: 532419.31Min: 549018.54 / Avg: 549103.71 / Max: 549261.191. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the Golang "Bombardier" program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 1Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0412K24K36K48K60KSE +/- 385.67, N = 15SE +/- 432.80, N = 9SE +/- 721.16, N = 352530.1152710.0355249.40-ldl -lpthread1. (CC) gcc options: -lcrypt -lz -O3 -march=native
OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 1Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0410K20K30K40K50KMin: 49862.03 / Avg: 52530.11 / Max: 54928.34Min: 51190.48 / Avg: 52710.03 / Max: 54529.09Min: 53912.84 / Avg: 55249.4 / Max: 56387.151. (CC) gcc options: -lcrypt -lz -O3 -march=native

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: CPU StressUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0415K30K45K60K75KSE +/- 15.37, N = 3SE +/- 64.21, N = 3SE +/- 73.09, N = 369941.3469408.5966584.161. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: CPU StressUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0412K24K36K48K60KMin: 69913.32 / Avg: 69941.34 / Max: 69966.3Min: 69331.69 / Avg: 69408.59 / Max: 69536.1Min: 66438.74 / Avg: 66584.16 / Max: 66669.831. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: yolov4-tinyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04510152025SE +/- 0.62, N = 3SE +/- 0.35, N = 3SE +/- 0.06, N = 320.1521.1220.78MIN: 19.23 / MAX: 24.44MIN: 19.68 / MAX: 123.27-pthread - MIN: 20.48 / MAX: 23.041. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: yolov4-tinyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04510152025Min: 19.45 / Avg: 20.15 / Max: 21.38Min: 20.64 / Avg: 21.12 / Max: 21.8Min: 20.67 / Avg: 20.78 / Max: 20.871. (CXX) g++ options: -O3 -rdynamic -lgomp -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 1.0Throughput Test: PartialTweetsUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.98331.96662.94993.93324.9165SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 34.184.174.37-pthread1. (CXX) g++ options: -O3
OpenBenchmarking.orgGB/s, More Is Bettersimdjson 1.0Throughput Test: PartialTweetsUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 4.18 / Avg: 4.18 / Max: 4.18Min: 4.16 / Avg: 4.17 / Max: 4.18Min: 4.37 / Avg: 4.37 / Max: 4.371. (CXX) g++ options: -O3

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 1.0Throughput Test: LargeRandomUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.24980.49960.74940.99921.249SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 31.111.111.06-pthread1. (CXX) g++ options: -O3
OpenBenchmarking.orgGB/s, More Is Bettersimdjson 1.0Throughput Test: LargeRandomUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 1.1 / Avg: 1.11 / Max: 1.11Min: 1.11 / Avg: 1.11 / Max: 1.11Min: 1.06 / Avg: 1.06 / Max: 1.061. (CXX) g++ options: -O3

NAS Parallel Benchmarks

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

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: MG.CUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0411K22K33K44K55KSE +/- 55.76, N = 3SE +/- 39.41, N = 3SE +/- 67.20, N = 352547.8850332.6450207.13-pthread -ldl -lutil -lrt1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0
OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: MG.CUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.049K18K27K36K45KMin: 52444.44 / Avg: 52547.88 / Max: 52635.7Min: 50264.28 / Avg: 50332.64 / Max: 50400.8Min: 50078.86 / Avg: 50207.13 / Max: 503061. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0

LAMMPS Molecular Dynamics Simulator

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

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 29Oct2020Model: Rhodopsin ProteinUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620SE +/- 0.05, N = 3SE +/- 0.16, N = 15SE +/- 0.14, N = 1515.9116.3516.64-pthread1. (CXX) g++ options: -O3 -lm
OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 29Oct2020Model: Rhodopsin ProteinUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620Min: 15.81 / Avg: 15.91 / Max: 15.98Min: 15.44 / Avg: 16.35 / Max: 17.09Min: 15.58 / Avg: 16.64 / Max: 17.361. (CXX) g++ options: -O3 -lm

TensorFlow Lite

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

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: NASNet MobileUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420K40K60K80K100KSE +/- 291.70, N = 3SE +/- 285.21, N = 3SE +/- 227.05, N = 386693.489415.185579.2
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: NASNet MobileUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0416K32K48K64K80KMin: 86119.2 / Avg: 86693.43 / Max: 87069.8Min: 89124.9 / Avg: 89415.1 / Max: 89985.5Min: 85321.9 / Avg: 85579.2 / Max: 86031.9

ONNX Runtime

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

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.8.2Model: shufflenet-v2-10 - Device: OpenMP CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.044K8K12K16K20KSE +/- 26.51, N = 3SE +/- 67.67, N = 3SE +/- 21.40, N = 31770417156179231. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt
OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.8.2Model: shufflenet-v2-10 - Device: OpenMP CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043K6K9K12K15KMin: 17674 / Avg: 17704.17 / Max: 17757Min: 17072 / Avg: 17155.5 / Max: 17289.5Min: 17887.5 / Avg: 17923.17 / Max: 17961.51. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: googlenetUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215SE +/- 0.34, N = 3SE +/- 0.35, N = 3SE +/- 0.06, N = 312.4813.0312.62MIN: 11.83 / MAX: 15.88MIN: 12.37 / MAX: 104.76-pthread - MIN: 12.35 / MAX: 13.351. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: googlenetUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620Min: 12.13 / Avg: 12.48 / Max: 13.17Min: 12.67 / Avg: 13.03 / Max: 13.73Min: 12.53 / Avg: 12.62 / Max: 12.721. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NAS Parallel Benchmarks

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

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: LU.CUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420K40K60K80K100KSE +/- 115.55, N = 3SE +/- 302.50, N = 3SE +/- 271.58, N = 384921.3582765.1881382.28-pthread -ldl -lutil -lrt1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0
OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: LU.CUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0415K30K45K60K75KMin: 84703.11 / Avg: 84921.35 / Max: 85096.31Min: 82160.38 / Avg: 82765.18 / Max: 83080.74Min: 80939.26 / Avg: 81382.28 / Max: 81875.941. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0

JPEG XL libjxl

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

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.5Input: PNG - Encode Speed: 8Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.22050.4410.66150.8821.1025SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.980.960.94-pthread1. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.5Input: PNG - Encode Speed: 8Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 0.98 / Avg: 0.98 / Max: 0.98Min: 0.96 / Avg: 0.96 / Max: 0.96Min: 0.94 / Avg: 0.94 / Max: 0.941. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie

OSPray

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

OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: NASA Streamlines - Renderer: SciVisUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041224364860SE +/- 0.53, N = 5SE +/- 0.53, N = 552.1150.5350.00MIN: 47.62 / MAX: 52.63MIN: 35.71 / MAX: 52.63MIN: 47.62 / MAX: 52.63
OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: NASA Streamlines - Renderer: SciVisUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041020304050Min: 50 / Avg: 52.11 / Max: 52.63Min: 50 / Avg: 50.53 / Max: 52.63

JPEG XL libjxl

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

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.5Input: JPEG - Encode Speed: 8Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835SE +/- 0.16, N = 3SE +/- 0.19, N = 3SE +/- 0.08, N = 329.9928.9428.78-pthread1. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.5Input: JPEG - Encode Speed: 8Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835Min: 29.68 / Avg: 29.99 / Max: 30.21Min: 28.64 / Avg: 28.94 / Max: 29.3Min: 28.67 / Avg: 28.78 / Max: 28.941. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie

TNN

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

OpenBenchmarking.orgms, Fewer Is BetterTNN 0.3Target: CPU - Model: SqueezeNet v2Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041428425670SE +/- 0.12, N = 3SE +/- 0.65, N = 3SE +/- 0.34, N = 361.8862.6160.13MIN: 61.5 / MAX: 62.54MIN: 61.65 / MAX: 64.18MIN: 59.55 / MAX: 60.971. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -fvisibility=default -O3 -rdynamic -ldl
OpenBenchmarking.orgms, Fewer Is BetterTNN 0.3Target: CPU - Model: SqueezeNet v2Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041224364860Min: 61.63 / Avg: 61.88 / Max: 62.02Min: 61.74 / Avg: 62.61 / Max: 63.88Min: 59.67 / Avg: 60.13 / Max: 60.81. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -fvisibility=default -O3 -rdynamic -ldl

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: squeezenet_ssdUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620SE +/- 0.11, N = 3SE +/- 0.11, N = 3SE +/- 0.01, N = 315.6916.1716.33MIN: 15.28 / MAX: 19.14MIN: 15.43 / MAX: 158.11-pthread - MIN: 16.02 / MAX: 17.241. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: squeezenet_ssdUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620Min: 15.55 / Avg: 15.69 / Max: 15.92Min: 16.01 / Avg: 16.17 / Max: 16.39Min: 16.32 / Avg: 16.33 / Max: 16.341. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

JPEG XL libjxl

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

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.5Input: PNG - Encode Speed: 7Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 39.749.369.73-pthread1. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.5Input: PNG - Encode Speed: 7Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 9.73 / Avg: 9.74 / Max: 9.74Min: 9.35 / Avg: 9.36 / Max: 9.36Min: 9.72 / Avg: 9.73 / Max: 9.731. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.5Input: PNG - Encode Speed: 5Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420406080100SE +/- 0.06, N = 3SE +/- 0.17, N = 3SE +/- 0.15, N = 374.9272.0072.78-pthread1. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie
OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.5Input: PNG - Encode Speed: 5Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041428425670Min: 74.81 / Avg: 74.92 / Max: 74.98Min: 71.76 / Avg: 72 / Max: 72.32Min: 72.5 / Avg: 72.78 / Max: 73.021. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie

OpenFOAM

OpenFOAM is the leading free, open source software for computational fluid dynamics (CFD). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 8Input: Motorbike 30MUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430SE +/- 0.06, N = 3SE +/- 0.07, N = 3SE +/- 0.15, N = 324.2124.2225.151. (CXX) g++ options: -std=c++11 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm
OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 8Input: Motorbike 30MUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430Min: 24.11 / Avg: 24.21 / Max: 24.31Min: 24.12 / Avg: 24.22 / Max: 24.36Min: 24.91 / Avg: 25.15 / Max: 25.421. (CXX) g++ options: -std=c++11 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: efficientnet-b0Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810SE +/- 0.00, N = 3SE +/- 0.06, N = 3SE +/- 0.01, N = 36.817.077.02MIN: 6.74 / MAX: 10.45MIN: 6.84 / MAX: 61.76-pthread - MIN: 6.94 / MAX: 7.291. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: efficientnet-b0Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 6.81 / Avg: 6.81 / Max: 6.82Min: 6.98 / Avg: 7.07 / Max: 7.19Min: 7 / Avg: 7.02 / Max: 7.041. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: CPU CacheUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420406080100SE +/- 1.59, N = 15SE +/- 1.52, N = 15SE +/- 1.03, N = 15105.96105.97109.971. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: CPU CacheUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420406080100Min: 95.91 / Avg: 105.96 / Max: 113.94Min: 98.08 / Avg: 105.97 / Max: 116.14Min: 101.14 / Avg: 109.97 / Max: 115.711. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: mnasnetUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041.16782.33563.50344.67125.839SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 35.015.135.19MIN: 4.91 / MAX: 15.1MIN: 5 / MAX: 20.79-pthread - MIN: 5.1 / MAX: 7.631. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: mnasnetUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 4.97 / Avg: 5.01 / Max: 5.05Min: 5.08 / Avg: 5.13 / Max: 5.17Min: 5.16 / Avg: 5.19 / Max: 5.231. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Glibc C String FunctionsUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04800K1600K2400K3200K4000KSE +/- 2603.56, N = 3SE +/- 23669.22, N = 3SE +/- 9931.73, N = 33705740.263684070.293580441.821. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Glibc C String FunctionsUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04600K1200K1800K2400K3000KMin: 3700555.76 / Avg: 3705740.26 / Max: 3708752.48Min: 3637466.11 / Avg: 3684070.29 / Max: 3714564.99Min: 3563231.5 / Avg: 3580441.82 / Max: 3597636.021. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

TNN

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

OpenBenchmarking.orgms, Fewer Is BetterTNN 0.3Target: CPU - Model: DenseNetUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.046001200180024003000SE +/- 0.69, N = 3SE +/- 0.80, N = 3SE +/- 1.49, N = 32837.532850.242755.17MIN: 2800.78 / MAX: 2894.04MIN: 2802.46 / MAX: 2916.75MIN: 2709.88 / MAX: 2820.441. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -fvisibility=default -O3 -rdynamic -ldl
OpenBenchmarking.orgms, Fewer Is BetterTNN 0.3Target: CPU - Model: DenseNetUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.045001000150020002500Min: 2836.47 / Avg: 2837.53 / Max: 2838.82Min: 2848.63 / Avg: 2850.24 / Max: 2851.12Min: 2752.26 / Avg: 2755.17 / Max: 2757.171. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -fvisibility=default -O3 -rdynamic -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 1.0Throughput Test: DistinctUserIDUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041.11152.2233.33454.4465.5575SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 34.944.934.78-pthread1. (CXX) g++ options: -O3
OpenBenchmarking.orgGB/s, More Is Bettersimdjson 1.0Throughput Test: DistinctUserIDUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 4.93 / Avg: 4.94 / Max: 4.95Min: 4.92 / Avg: 4.93 / Max: 4.93Min: 4.77 / Avg: 4.78 / Max: 4.791. (CXX) g++ options: -O3

Timed Godot Game Engine Compilation

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

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Godot Game Engine Compilation 3.2.3Time To CompileUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041326395265SE +/- 0.12, N = 3SE +/- 0.15, N = 3SE +/- 0.03, N = 356.0856.3657.90
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Godot Game Engine Compilation 3.2.3Time To CompileUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041122334455Min: 55.86 / Avg: 56.08 / Max: 56.3Min: 56.07 / Avg: 56.36 / Max: 56.55Min: 57.85 / Avg: 57.9 / Max: 57.95

TTSIOD 3D Renderer

A portable GPL 3D software renderer that supports OpenMP and Intel Threading Building Blocks with many different rendering modes. This version does not use OpenGL but is entirely CPU/software based. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterTTSIOD 3D Renderer 2.3bPhong Rendering With Soft-Shadow MappingUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042004006008001000SE +/- 13.57, N = 3SE +/- 0.68, N = 3SE +/- 3.03, N = 3957.39969.49988.231. (CXX) g++ options: -O3 -fomit-frame-pointer -ffast-math -mtune=native -flto -msse -mrecip -mfpmath=sse -msse2 -mssse3 -lSDL -fopenmp -fwhole-program -lstdc++
OpenBenchmarking.orgFPS, More Is BetterTTSIOD 3D Renderer 2.3bPhong Rendering With Soft-Shadow MappingUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042004006008001000Min: 930.93 / Avg: 957.39 / Max: 975.8Min: 968.24 / Avg: 969.49 / Max: 970.59Min: 982.22 / Avg: 988.23 / Max: 991.971. (CXX) g++ options: -O3 -fomit-frame-pointer -ffast-math -mtune=native -flto -msse -mrecip -mfpmath=sse -msse2 -mssse3 -lSDL -fopenmp -fwhole-program -lstdc++

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 3.1Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835SE +/- 0.13, N = 3SE +/- 0.03, N = 3SE +/- 0.20, N = 330.9631.2831.95-lpthread1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.1Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835Min: 30.79 / Avg: 30.96 / Max: 31.22Min: 31.24 / Avg: 31.28 / Max: 31.34Min: 31.57 / Avg: 31.95 / Max: 32.231. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

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 5.14Time To CompileUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04816243240SE +/- 0.44, N = 3SE +/- 0.38, N = 3SE +/- 0.40, N = 336.4736.7335.66
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 5.14Time To CompileUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04816243240Min: 35.98 / Avg: 36.47 / Max: 37.35Min: 36.34 / Avg: 36.73 / Max: 37.49Min: 35.23 / Avg: 35.66 / Max: 36.46

NAS Parallel Benchmarks

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

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.CUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.048K16K24K32K40KSE +/- 24.90, N = 3SE +/- 27.26, N = 3SE +/- 62.61, N = 336599.6035556.4435627.71-pthread -ldl -lutil -lrt1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0
OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.CUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.046K12K18K24K30KMin: 36551.79 / Avg: 36599.6 / Max: 36635.58Min: 35516.13 / Avg: 35556.44 / Max: 35608.39Min: 35507.95 / Avg: 35627.71 / Max: 35719.231. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.BUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0414K28K42K56K70KSE +/- 150.74, N = 3SE +/- 123.88, N = 3SE +/- 199.94, N = 365350.9763598.1463522.54-pthread -ldl -lutil -lrt1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0
OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.BUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0411K22K33K44K55KMin: 65155.58 / Avg: 65350.97 / Max: 65647.5Min: 63443.55 / Avg: 63598.14 / Max: 63843.11Min: 63252.93 / Avg: 63522.54 / Max: 63913.11. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0

ONNX Runtime

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

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.8.2Model: yolov4 - Device: OpenMP CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04100200300400500SE +/- 1.89, N = 3SE +/- 0.50, N = 3SE +/- 1.76, N = 34364284401. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt
OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.8.2Model: yolov4 - Device: OpenMP CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0480160240320400Min: 433 / Avg: 436 / Max: 439.5Min: 427.5 / Avg: 428 / Max: 429Min: 437.5 / Avg: 440.17 / Max: 443.51. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt

Rodinia

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

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP CFD SolverUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215SE +/- 0.035, N = 3SE +/- 0.033, N = 3SE +/- 0.002, N = 39.7019.5419.4471. (CXX) g++ options: -O2 -lOpenCL
OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP CFD SolverUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 9.65 / Avg: 9.7 / Max: 9.77Min: 9.48 / Avg: 9.54 / Max: 9.58Min: 9.44 / Avg: 9.45 / Max: 9.451. (CXX) g++ options: -O2 -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 3.2Preset: ThoroughUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810SE +/- 0.0028, N = 3SE +/- 0.0094, N = 3SE +/- 0.0094, N = 36.54506.56776.71811. (CXX) g++ options: -O3 -flto -pthread
OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 3.2Preset: ThoroughUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 6.54 / Avg: 6.55 / Max: 6.55Min: 6.55 / Avg: 6.57 / Max: 6.58Min: 6.7 / Avg: 6.72 / Max: 6.731. (CXX) g++ options: -O3 -flto -pthread

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 CompileUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04306090120150SE +/- 0.20, N = 3SE +/- 0.14, N = 3SE +/- 0.16, N = 3153.04153.71157.06
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Node.js Compilation 15.11Time To CompileUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04306090120150Min: 152.67 / Avg: 153.04 / Max: 153.37Min: 153.45 / Avg: 153.71 / Max: 153.94Min: 156.79 / Avg: 157.06 / Max: 157.35

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: 0Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041020304050SE +/- 0.14, N = 3SE +/- 0.04, N = 3SE +/- 0.14, N = 345.0745.1246.231. (CXX) g++ options: -O3 -fPIC -lm
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 0Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04918273645Min: 44.8 / Avg: 45.07 / Max: 45.27Min: 45.05 / Avg: 45.12 / Max: 45.18Min: 46.06 / Avg: 46.23 / Max: 46.51. (CXX) g++ options: -O3 -fPIC -lm

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Matrix MathUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420K40K60K80K100KSE +/- 87.63, N = 3SE +/- 452.67, N = 3SE +/- 17.49, N = 3105285.85102691.89103521.201. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Matrix MathUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420K40K60K80K100KMin: 105122.52 / Avg: 105285.85 / Max: 105422.55Min: 102077.43 / Avg: 102691.89 / Max: 103574.94Min: 103499.49 / Avg: 103521.2 / Max: 103555.811. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

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: 2Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430SE +/- 0.03, N = 3SE +/- 0.14, N = 3SE +/- 0.09, N = 324.5225.0225.141. (CXX) g++ options: -O3 -fPIC -lm
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 2Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430Min: 24.48 / Avg: 24.52 / Max: 24.57Min: 24.88 / Avg: 25.02 / Max: 25.3Min: 25.01 / Avg: 25.14 / Max: 25.311. (CXX) g++ options: -O3 -fPIC -lm

Cpuminer-Opt

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

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: LBC, LBRY CreditsUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0415K30K45K60K75KSE +/- 288.46, N = 3SE +/- 158.75, N = 3SE +/- 105.99, N = 36916369070708101. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: LBC, LBRY CreditsUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0412K24K36K48K60KMin: 68860 / Avg: 69163.33 / Max: 69740Min: 68770 / Avg: 69070 / Max: 69310Min: 70600 / Avg: 70810 / Max: 709401. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Memory CopyingUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042K4K6K8K10KSE +/- 4.32, N = 3SE +/- 3.56, N = 3SE +/- 4.75, N = 38141.988094.808289.321. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Memory CopyingUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0414002800420056007000Min: 8134.42 / Avg: 8141.98 / Max: 8149.37Min: 8090.06 / Avg: 8094.8 / Max: 8101.78Min: 8279.91 / Avg: 8289.32 / Max: 8295.081. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

ONNX Runtime

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

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.8.2Model: fcn-resnet101-11 - Device: OpenMP CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.044080120160200SE +/- 1.00, N = 3SE +/- 0.60, N = 3SE +/- 2.02, N = 31711671711. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt
OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.8.2Model: fcn-resnet101-11 - Device: OpenMP CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04306090120150Min: 169.5 / Avg: 170.5 / Max: 172.5Min: 166 / Avg: 167.17 / Max: 168Min: 168 / Avg: 170.5 / Max: 174.51. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt

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, LosslessUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835SE +/- 0.12, N = 3SE +/- 0.18, N = 3SE +/- 0.05, N = 328.3628.8629.031. (CXX) g++ options: -O3 -fPIC -lm
OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.9.0Encoder Speed: 6, LosslessUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430Min: 28.24 / Avg: 28.36 / Max: 28.59Min: 28.68 / Avg: 28.86 / Max: 29.22Min: 28.94 / Avg: 29.03 / Max: 29.131. (CXX) g++ options: -O3 -fPIC -lm

7-Zip Compression

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

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 16.02Compress Speed TestUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KSE +/- 240.00, N = 3SE +/- 363.33, N = 3SE +/- 539.21, N = 31751651721171711521. (CXX) g++ options: -pipe -lpthread
OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 16.02Compress Speed TestUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430K60K90K120K150KMin: 174924 / Avg: 175165 / Max: 175645Min: 171591 / Avg: 172116.67 / Max: 172814Min: 170404 / Avg: 171152.33 / Max: 1721991. (CXX) g++ options: -pipe -lpthread

PostgreSQL pgbench

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

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 500 - Mode: Read OnlyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04140K280K420K560K700KSE +/- 1517.26, N = 3SE +/- 1182.41, N = 3SE +/- 1225.85, N = 3624794639361639294-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 500 - Mode: Read OnlyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04110K220K330K440K550KMin: 621902.78 / Avg: 624793.85 / Max: 627037.82Min: 637849.71 / Avg: 639361.16 / Max: 641691.98Min: 637040.9 / Avg: 639294.02 / Max: 641257.661. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 500 - Mode: Read Only - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.180.360.540.720.9SE +/- 0.002, N = 3SE +/- 0.002, N = 3SE +/- 0.001, N = 30.8000.7820.782-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 500 - Mode: Read Only - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 0.8 / Avg: 0.8 / Max: 0.8Min: 0.78 / Avg: 0.78 / Max: 0.78Min: 0.78 / Avg: 0.78 / Max: 0.791. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

LuxCoreRender

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

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: LuxCore Benchmark - Acceleration: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.90681.81362.72043.62724.534SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 33.993.944.03MIN: 1.52 / MAX: 4.63MIN: 1.48 / MAX: 4.61MIN: 1.51 / MAX: 4.68
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: LuxCore Benchmark - Acceleration: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 3.97 / Avg: 3.99 / Max: 4Min: 3.93 / Avg: 3.94 / Max: 3.96Min: 4 / Avg: 4.03 / Max: 4.04

Facebook RocksDB

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

OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.22.1Test: Read Random Write RandomUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04600K1200K1800K2400K3000KSE +/- 16007.37, N = 3SE +/- 15779.88, N = 3SE +/- 6660.96, N = 32748440268779127488901. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.22.1Test: Read Random Write RandomUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04500K1000K1500K2000K2500KMin: 2716964 / Avg: 2748440 / Max: 2769243Min: 2664456 / Avg: 2687790.67 / Max: 2717860Min: 2737323 / Avg: 2748889.67 / Max: 27603971. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

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 1080pUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04100200300400500SE +/- 5.09, N = 14SE +/- 3.38, N = 13SE +/- 3.59, N = 14444.46446.62436.861. (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 1080pUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0480160240320400Min: 383.41 / Avg: 444.46 / Max: 456.07Min: 408.39 / Avg: 446.62 / Max: 454.84Min: 394.39 / Avg: 436.86 / Max: 448.981. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

PostgreSQL pgbench

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

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 250 - Mode: Read OnlyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04160K320K480K640K800KSE +/- 1092.30, N = 3SE +/- 2610.98, N = 3SE +/- 2711.78, N = 3748263739400731944-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 250 - Mode: Read OnlyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04130K260K390K520K650KMin: 746292.63 / Avg: 748263.01 / Max: 750065.22Min: 734178.96 / Avg: 739400.43 / Max: 742073.12Min: 727028.21 / Avg: 731943.97 / Max: 736386.231. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

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 1.0Throughput Test: KostyaUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.72231.44462.16692.88923.6115SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 33.213.213.14-pthread1. (CXX) g++ options: -O3
OpenBenchmarking.orgGB/s, More Is Bettersimdjson 1.0Throughput Test: KostyaUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 3.21 / Avg: 3.21 / Max: 3.21Min: 3.2 / Avg: 3.21 / Max: 3.21Min: 3.14 / Avg: 3.14 / Max: 3.141. (CXX) g++ options: -O3

LuxCoreRender

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

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Danish Mood - Acceleration: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.83031.66062.49093.32124.1515SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 33.693.613.67MIN: 1.43 / MAX: 4.3MIN: 1.38 / MAX: 4.2MIN: 1.38 / MAX: 4.29
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Danish Mood - Acceleration: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 3.63 / Avg: 3.69 / Max: 3.77Min: 3.6 / Avg: 3.61 / Max: 3.63Min: 3.63 / Avg: 3.67 / Max: 3.71

SecureMark

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

OpenBenchmarking.orgmarks, More Is BetterSecureMark 1.0.4Benchmark: SecureMark-TLSUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0460K120K180K240K300KSE +/- 277.92, N = 3SE +/- 1689.23, N = 3SE +/- 2193.40, N = 32723102696882755451. (CC) gcc options: -pedantic -O3
OpenBenchmarking.orgmarks, More Is BetterSecureMark 1.0.4Benchmark: SecureMark-TLSUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0450K100K150K200K250KMin: 271817.94 / Avg: 272309.97 / Max: 272779.94Min: 267373.03 / Avg: 269688.34 / Max: 272976.72Min: 271170.59 / Avg: 275544.57 / Max: 278021.721. (CC) gcc options: -pedantic -O3

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU-v2-v2 - Model: mobilenet-v2Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041.28932.57863.86795.15726.4465SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 35.615.735.72MIN: 5.46 / MAX: 9.47MIN: 5.52 / MAX: 15.08-pthread - MIN: 5.58 / MAX: 6.051. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU-v2-v2 - Model: mobilenet-v2Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 5.57 / Avg: 5.61 / Max: 5.65Min: 5.69 / Avg: 5.73 / Max: 5.75Min: 5.69 / Avg: 5.72 / Max: 5.751. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

PostgreSQL pgbench

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

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 250 - Mode: Read Only - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.07670.15340.23010.30680.3835SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.001, N = 30.3340.3380.341-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 100 - Clients: 250 - Mode: Read Only - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0412345Min: 0.33 / Avg: 0.33 / Max: 0.34Min: 0.34 / Avg: 0.34 / Max: 0.34Min: 0.34 / Avg: 0.34 / Max: 0.341. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

TensorFlow Lite

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

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet FloatUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0412K24K36K48K60KSE +/- 104.96, N = 3SE +/- 98.36, N = 3SE +/- 116.71, N = 354956.155376.054254.0
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet FloatUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0410K20K30K40K50KMin: 54781.7 / Avg: 54956.13 / Max: 55144.5Min: 55264.8 / Avg: 55375.97 / Max: 55572.1Min: 54125.8 / Avg: 54253.97 / Max: 54487

LAMMPS Molecular Dynamics Simulator

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

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 29Oct2020Model: 20k AtomsUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 317.0016.8717.21-pthread1. (CXX) g++ options: -O3 -lm
OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 29Oct2020Model: 20k AtomsUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620Min: 16.95 / Avg: 17 / Max: 17.03Min: 16.79 / Avg: 16.87 / Max: 16.92Min: 17.15 / Avg: 17.21 / Max: 17.251. (CXX) g++ options: -O3 -lm

SVT-AV1

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

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8.7Encoder Mode: Preset 8 - Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04510152025SE +/- 0.10, N = 3SE +/- 0.12, N = 3SE +/- 0.12, N = 319.4919.7119.321. (CXX) g++ options: -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq -pie
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8.7Encoder Mode: Preset 8 - Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04510152025Min: 19.29 / Avg: 19.49 / Max: 19.59Min: 19.46 / Avg: 19.71 / Max: 19.87Min: 19.15 / Avg: 19.32 / Max: 19.551. (CXX) g++ options: -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq -pie

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 1080pUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04120240360480600SE +/- 0.62, N = 3SE +/- 2.56, N = 3SE +/- 1.52, N = 3556.42553.02545.961. (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 1080pUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04100200300400500Min: 555.56 / Avg: 556.42 / Max: 557.62Min: 550.46 / Avg: 553.02 / Max: 558.14Min: 543.97 / Avg: 545.96 / Max: 548.951. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

Cpuminer-Opt

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

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: x25xUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042004006008001000SE +/- 3.73, N = 3SE +/- 4.78, N = 3SE +/- 2.57, N = 3773.60768.20782.731. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: x25xUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04140280420560700Min: 766.14 / Avg: 773.6 / Max: 777.45Min: 760.62 / Avg: 768.2 / Max: 777.03Min: 777.59 / Avg: 782.73 / Max: 785.411. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Chaos Group V-RAY

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

OpenBenchmarking.orgvsamples, More Is BetterChaos Group V-RAY 5Mode: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.046K12K18K24K30KSE +/- 262.43, N = 3SE +/- 357.24, N = 3SE +/- 317.32, N = 4283052782527781
OpenBenchmarking.orgvsamples, More Is BetterChaos Group V-RAY 5Mode: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.045K10K15K20K25KMin: 27794 / Avg: 28304.67 / Max: 28665Min: 27111 / Avg: 27825.33 / Max: 28195Min: 26956 / Avg: 27780.75 / Max: 28501

PlaidML

This test profile uses PlaidML deep learning framework developed by Intel for offering up various benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: ResNet 50 - Device: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215SE +/- 0.02, N = 3SE +/- 0.09, N = 3SE +/- 0.02, N = 310.2910.1010.21
OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: ResNet 50 - Device: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 10.24 / Avg: 10.29 / Max: 10.31Min: 9.92 / Avg: 10.1 / Max: 10.2Min: 10.17 / Avg: 10.21 / Max: 10.23

Facebook RocksDB

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

OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.22.1Test: Update RandomUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04160K320K480K640K800KSE +/- 2476.72, N = 3SE +/- 2766.16, N = 3SE +/- 1654.98, N = 37413437277677382401. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.22.1Test: Update RandomUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04130K260K390K520K650KMin: 738270 / Avg: 741343 / Max: 746244Min: 724582 / Avg: 727767 / Max: 733277Min: 736063 / Avg: 738240.33 / Max: 7414881. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

NAS Parallel Benchmarks

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

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: FT.CUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.049K18K27K36K45KSE +/- 29.98, N = 3SE +/- 131.14, N = 3SE +/- 94.87, N = 343590.5742915.0343707.66-pthread -ldl -lutil -lrt1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0
OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: FT.CUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.048K16K24K32K40KMin: 43549.37 / Avg: 43590.57 / Max: 43648.91Min: 42654.89 / Avg: 42915.03 / Max: 43074.01Min: 43573.23 / Avg: 43707.66 / Max: 43890.851. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.0

LeelaChessZero

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

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: EigenUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430060090012001500SE +/- 6.17, N = 3SE +/- 12.42, N = 3SE +/- 9.77, N = 31306133013251. (CXX) g++ options: -flto -pthread
OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: EigenUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042004006008001000Min: 1299 / Avg: 1305.67 / Max: 1318Min: 1308 / Avg: 1330 / Max: 1351Min: 1311 / Avg: 1325.33 / Max: 13441. (CXX) g++ options: -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: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042004006008001000SE +/- 1.21, N = 3SE +/- 2.29, N = 3SE +/- 0.35, N = 3766.69777.26763.62MIN: 756.56MIN: 757.43MIN: 755.241. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04140280420560700Min: 764.47 / Avg: 766.69 / Max: 768.65Min: 772.93 / Avg: 777.26 / Max: 780.73Min: 763.17 / Avg: 763.62 / Max: 764.321. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread

TensorFlow Lite

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

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet QuantUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0413K26K39K52K65KSE +/- 189.93, N = 3SE +/- 4.08, N = 3SE +/- 48.17, N = 358918.059851.758811.9
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet QuantUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0410K20K30K40K50KMin: 58540.8 / Avg: 58918 / Max: 59145.4Min: 59843.8 / Avg: 59851.67 / Max: 59857.5Min: 58717.4 / Avg: 58811.93 / Max: 58875.3

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: CryptoUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0414002800420056007000SE +/- 1.02, N = 3SE +/- 0.93, N = 3SE +/- 2.39, N = 36539.846537.916652.821. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: CryptoUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0412002400360048006000Min: 6538.81 / Avg: 6539.84 / Max: 6541.87Min: 6536.21 / Avg: 6537.91 / Max: 6539.42Min: 6648.1 / Avg: 6652.82 / Max: 6655.841. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: AtomicUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KSE +/- 158.29, N = 3SE +/- 198.79, N = 3SE +/- 1480.39, N = 3197807.92196373.37199708.111. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: AtomicUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430K60K90K120K150KMin: 197647.34 / Avg: 197807.92 / Max: 198124.49Min: 196023.47 / Avg: 196373.37 / Max: 196711.83Min: 196748.42 / Avg: 199708.11 / Max: 201257.241. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

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 3.1Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810SE +/- 0.02, N = 3SE +/- 0.08, N = 4SE +/- 0.00, N = 37.357.307.23-lpthread1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.1Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 7.31 / Avg: 7.35 / Max: 7.39Min: 7.08 / Avg: 7.3 / Max: 7.42Min: 7.23 / Avg: 7.23 / Max: 7.241. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

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 DirectionUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835SE +/- 0.18, N = 3SE +/- 0.15, N = 3SE +/- 0.03, N = 329.8930.3930.29-pthread -ldl -lutil -lrt1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 193 Cells Per DirectionUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835Min: 29.53 / Avg: 29.89 / Max: 30.08Min: 30.08 / Avg: 30.39 / Max: 30.57Min: 30.25 / Avg: 30.29 / Max: 30.351. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: shufflenet-v2Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 36.136.216.23MIN: 5.99 / MAX: 8.3MIN: 6.02 / MAX: 21.42-pthread - MIN: 6.08 / MAX: 6.81. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: shufflenet-v2Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 6.11 / Avg: 6.13 / Max: 6.16Min: 6.17 / Avg: 6.21 / Max: 6.25Min: 6.14 / Avg: 6.23 / Max: 6.31. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Timed LLVM Compilation

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

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 13.0Build System: NinjaUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0450100150200250SE +/- 0.32, N = 3SE +/- 0.59, N = 3SE +/- 0.21, N = 3234.95235.75238.52
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 13.0Build System: NinjaUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.044080120160200Min: 234.63 / Avg: 234.95 / Max: 235.58Min: 235.09 / Avg: 235.75 / Max: 236.93Min: 238.1 / Avg: 238.52 / Max: 238.78

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 13.0Build System: Unix MakefilesUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0460120180240300SE +/- 2.61, N = 6SE +/- 3.77, N = 3SE +/- 1.43, N = 3275.49274.26278.38
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 13.0Build System: Unix MakefilesUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0450100150200250Min: 267.31 / Avg: 275.49 / Max: 285.19Min: 266.75 / Avg: 274.26 / Max: 278.45Min: 275.53 / Avg: 278.38 / Max: 280.1

PostgreSQL pgbench

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

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 250 - Mode: Read Only - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.0920.1840.2760.3680.46SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.003, N = 30.4030.4090.408-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 250 - Mode: Read Only - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0412345Min: 0.4 / Avg: 0.4 / Max: 0.4Min: 0.41 / Avg: 0.41 / Max: 0.41Min: 0.41 / Avg: 0.41 / Max: 0.411. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample 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 1080pUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0460120180240300SE +/- 2.41, N = 8SE +/- 3.36, N = 4SE +/- 2.36, N = 8283.51281.48279.491. (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 1080pUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0450100150200250Min: 266.79 / Avg: 283.51 / Max: 287.63Min: 271.49 / Avg: 281.48 / Max: 285.71Min: 263.16 / Avg: 279.49 / Max: 283.551. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

PostgreSQL pgbench

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

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 250 - Mode: Read OnlyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04130K260K390K520K650KSE +/- 682.74, N = 3SE +/- 1809.68, N = 3SE +/- 3814.98, N = 3620425611633612473-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 250 - Mode: Read OnlyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04110K220K330K440K550KMin: 619079.55 / Avg: 620425.39 / Max: 621298.21Min: 608117.11 / Avg: 611633.03 / Max: 614135.01Min: 604860.73 / Avg: 612473.44 / Max: 616723.821. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenFOAM

OpenFOAM is the leading free, open source software for computational fluid dynamics (CFD). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 8Input: Motorbike 60MUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0450100150200250SE +/- 0.24, N = 3SE +/- 0.49, N = 3SE +/- 0.38, N = 3239.68243.11242.991. (CXX) g++ options: -std=c++11 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm
OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 8Input: Motorbike 60MUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.044080120160200Min: 239.34 / Avg: 239.68 / Max: 240.14Min: 242.61 / Avg: 243.11 / Max: 244.08Min: 242.3 / Avg: 242.99 / Max: 243.611. (CXX) g++ options: -std=c++11 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -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: f32 - Engine: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430060090012001500SE +/- 0.45, N = 3SE +/- 1.18, N = 3SE +/- 2.26, N = 31412.991429.311411.47MIN: 1396.9MIN: 1397.53MIN: 1395.151. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042004006008001000Min: 1412.25 / Avg: 1412.99 / Max: 1413.79Min: 1427.21 / Avg: 1429.31 / Max: 1431.31Min: 1408.72 / Avg: 1411.47 / Max: 1415.951. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread

Facebook RocksDB

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

OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.22.1Test: Random FillUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04170K340K510K680K850KSE +/- 6927.89, N = 3SE +/- 4583.67, N = 3SE +/- 8363.02, N = 57724687774217679941. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.22.1Test: Random FillUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04130K260K390K520K650KMin: 760409 / Avg: 772468 / Max: 784407Min: 768342 / Avg: 777421 / Max: 783060Min: 754562 / Avg: 767993.6 / Max: 8001161. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

Intel Open Image Denoise

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

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.4.0Run: RT.ldr_alb_nrm.3840x2160Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.19350.3870.58050.7740.9675SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.850.860.86
OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.4.0Run: RT.ldr_alb_nrm.3840x2160Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 0.85 / Avg: 0.85 / Max: 0.86Min: 0.86 / Avg: 0.86 / Max: 0.86Min: 0.86 / Avg: 0.86 / Max: 0.86

GROMACS

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

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2021.2Implementation: MPI CPU - Input: water_GMX50_bareUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.74181.48362.22542.96723.709SE +/- 0.000, N = 3SE +/- 0.011, N = 3SE +/- 0.004, N = 33.2973.2593.282-pthread1. (CXX) g++ options: -O3
OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2021.2Implementation: MPI CPU - Input: water_GMX50_bareUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 3.3 / Avg: 3.3 / Max: 3.3Min: 3.24 / Avg: 3.26 / Max: 3.27Min: 3.27 / Avg: 3.28 / Max: 3.291. (CXX) g++ options: -O3

Embree

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

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer ISPC - Model: CrownUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 328.2828.4928.55MIN: 27.86 / MAX: 28.93MIN: 28.01 / MAX: 29.3MIN: 28.05 / MAX: 29.17
OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer ISPC - Model: CrownUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430Min: 28.23 / Avg: 28.28 / Max: 28.34Min: 28.43 / Avg: 28.49 / Max: 28.59Min: 28.46 / Avg: 28.55 / Max: 28.63

Appleseed

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

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Material TesterUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04306090120150147.98147.90146.60

Embree

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

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer - Model: CrownUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835SE +/- 0.14, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 330.9530.7631.05MIN: 30.32 / MAX: 32.01MIN: 30.24 / MAX: 31.73MIN: 30.63 / MAX: 31.84
OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer - Model: CrownUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835Min: 30.76 / Avg: 30.95 / Max: 31.21Min: 30.65 / Avg: 30.76 / Max: 30.84Min: 31.02 / Avg: 31.05 / Max: 31.07

TensorFlow Lite

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

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Inception V4Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04300K600K900K1200K1500KSE +/- 1952.13, N = 3SE +/- 1392.75, N = 3SE +/- 1058.34, N = 3121790312245731213533
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Inception V4Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04200K400K600K800K1000KMin: 1214120 / Avg: 1217903.33 / Max: 1220630Min: 1222910 / Avg: 1224573.33 / Max: 1227340Min: 1211570 / Avg: 1213533.33 / Max: 1215200

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: SqueezeNetUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420K40K60K80K100KSE +/- 134.59, N = 3SE +/- 42.73, N = 3SE +/- 18.94, N = 381889.182510.881775.2
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: SqueezeNetUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0414K28K42K56K70KMin: 81629.8 / Avg: 81889.07 / Max: 82081.4Min: 82433.5 / Avg: 82510.8 / Max: 82581Min: 81742.8 / Avg: 81775.23 / Max: 81808.4

Apache Cassandra

This is a benchmark of the Apache Cassandra NoSQL database management system making use of cassandra-stress. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterApache Cassandra 4.0Test: WritesUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0450K100K150K200K250KSE +/- 1062.06, N = 3SE +/- 927.98, N = 3SE +/- 1033.60, N = 3231554230318229507
OpenBenchmarking.orgOp/s, More Is BetterApache Cassandra 4.0Test: WritesUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KMin: 229741 / Avg: 231554.33 / Max: 233419Min: 229017 / Avg: 230318.33 / Max: 232115Min: 227646 / Avg: 229506.67 / Max: 231217

Embree

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

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer - Model: Asian DragonUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04816243240SE +/- 0.02, N = 3SE +/- 0.16, N = 3SE +/- 0.31, N = 334.3834.1634.08MIN: 33.98 / MAX: 35.18MIN: 33.63 / MAX: 34.9MIN: 33.43 / MAX: 35.23
OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer - Model: Asian DragonUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835Min: 34.36 / Avg: 34.38 / Max: 34.43Min: 33.85 / Avg: 34.16 / Max: 34.37Min: 33.63 / Avg: 34.08 / Max: 34.67

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 3.1Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041122334455SE +/- 0.19, N = 3SE +/- 0.15, N = 3SE +/- 0.47, N = 346.7346.8047.13-lpthread1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.1Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4KUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041020304050Min: 46.42 / Avg: 46.73 / Max: 47.06Min: 46.58 / Avg: 46.8 / Max: 47.08Min: 46.55 / Avg: 47.13 / Max: 48.051. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

LuxCoreRender

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

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Orange Juice - Acceleration: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 37.097.057.03MIN: 6.35 / MAX: 7.47MIN: 6.32 / MAX: 7.44MIN: 6.25 / MAX: 7.39
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Orange Juice - Acceleration: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 7.06 / Avg: 7.09 / Max: 7.11Min: 7.04 / Avg: 7.05 / Max: 7.07Min: 7.01 / Avg: 7.03 / Max: 7.05

OpenSSL

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

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.0Algorithm: SHA256Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.048000M16000M24000M32000M40000MSE +/- 63835094.69, N = 3SE +/- 85432560.60, N = 3SE +/- 15039336.54, N = 33628985227036164270263360141711671. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.0Algorithm: SHA256Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.046000M12000M18000M24000M30000MMin: 36162249750 / Avg: 36289852270 / Max: 36357252920Min: 36021457220 / Avg: 36164270263.33 / Max: 36316914050Min: 35988548410 / Avg: 36014171166.67 / Max: 360406262501. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

TensorFlow Lite

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

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Inception ResNet V2Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04200K400K600K800K1000KSE +/- 529.75, N = 3SE +/- 399.50, N = 3SE +/- 863.35, N = 3107614010803201072563
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Inception ResNet V2Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04200K400K600K800K1000KMin: 1075410 / Avg: 1076140 / Max: 1077170Min: 1079660 / Avg: 1080320 / Max: 1081040Min: 1071290 / Avg: 1072563.33 / Max: 1074210

Cpuminer-Opt

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

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: DeepcoinUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.045K10K15K20K25KSE +/- 28.87, N = 3SE +/- 170.39, N = 3SE +/- 114.07, N = 32345023600235331. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: DeepcoinUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.044K8K12K16K20KMin: 23400 / Avg: 23450 / Max: 23500Min: 23410 / Avg: 23600 / Max: 23940Min: 23320 / Avg: 23533.33 / Max: 237101. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Xmrig

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

OpenBenchmarking.orgH/s, More Is BetterXmrig 6.12.1Variant: Monero - Hash Count: 1MUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.044K8K12K16K20KSE +/- 44.27, N = 3SE +/- 22.84, N = 3SE +/- 22.27, N = 318333.418217.518245.21. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
OpenBenchmarking.orgH/s, More Is BetterXmrig 6.12.1Variant: Monero - Hash Count: 1MUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043K6K9K12K15KMin: 18282.9 / Avg: 18333.37 / Max: 18421.6Min: 18188.1 / Avg: 18217.53 / Max: 18262.5Min: 18205 / Avg: 18245.23 / Max: 18281.91. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

Cpuminer-Opt

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

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: SkeincoinUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KSE +/- 880.02, N = 3SE +/- 1356.26, N = 3SE +/- 581.44, N = 32094002085102080831. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: SkeincoinUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KMin: 208010 / Avg: 209400 / Max: 211030Min: 205850 / Avg: 208510 / Max: 210300Min: 207200 / Avg: 208083.33 / Max: 2091801. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

LuxCoreRender

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

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: DLSC - Acceleration: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041.07332.14663.21994.29325.3665SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 34.774.744.76MIN: 4.56 / MAX: 5.2MIN: 4.6 / MAX: 5.11MIN: 4.64 / MAX: 5.03
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: DLSC - Acceleration: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 4.74 / Avg: 4.77 / Max: 4.79Min: 4.72 / Avg: 4.74 / Max: 4.78Min: 4.74 / Avg: 4.76 / Max: 4.78

Cpuminer-Opt

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

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: Quad SHA-256, PyriteUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0450K100K150K200K250KSE +/- 257.75, N = 3SE +/- 44.85, N = 3SE +/- 447.38, N = 32136902142672150031. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: Quad SHA-256, PyriteUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0440K80K120K160K200KMin: 213180 / Avg: 213690 / Max: 214010Min: 214180 / Avg: 214266.67 / Max: 214330Min: 214410 / Avg: 215003.33 / Max: 2158801. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Blender

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

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Barbershop - Compute: CPU-OnlyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0450100150200250SE +/- 0.53, N = 3SE +/- 0.51, N = 3SE +/- 0.23, N = 3227.94229.14227.98
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Barbershop - Compute: CPU-OnlyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.044080120160200Min: 227.22 / Avg: 227.94 / Max: 228.97Min: 228.16 / Avg: 229.14 / Max: 229.89Min: 227.52 / Avg: 227.98 / Max: 228.21

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: SemaphoresUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04800K1600K2400K3200K4000KSE +/- 4849.41, N = 3SE +/- 15414.76, N = 3SE +/- 21731.06, N = 33635712.183625859.103642449.461. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: SemaphoresUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04600K1200K1800K2400K3000KMin: 3627187.26 / Avg: 3635712.18 / Max: 3643980.2Min: 3610062.78 / Avg: 3625859.1 / Max: 3656685.44Min: 3608365.52 / Avg: 3642449.46 / Max: 3682845.891. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lcrypt -lrt -lz -ldl -pthread -lc -latomic

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 TimeUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0416M32M48M64M80MSE +/- 552677.79, N = 15SE +/- 716479.64, N = 3SE +/- 461786.73, N = 157530436975631071753163261. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fprofile-use -fno-peel-loops -fno-tracer -pedantic -O3 -msse -msse3 -mpopcnt -mavx2 -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto=jobserver
OpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 13Total TimeUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0413M26M39M52M65MMin: 70173742 / Avg: 75304369.07 / Max: 79163766Min: 74755871 / Avg: 75631071 / Max: 77051294Min: 73832111 / Avg: 75316325.67 / Max: 807368441. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fprofile-use -fno-peel-loops -fno-tracer -pedantic -O3 -msse -msse3 -mpopcnt -mavx2 -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto=jobserver

Xmrig

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

OpenBenchmarking.orgH/s, More Is BetterXmrig 6.12.1Variant: Wownero - Hash Count: 1MUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.044K8K12K16K20KSE +/- 6.03, N = 3SE +/- 14.16, N = 3SE +/- 11.01, N = 319138.519059.619125.81. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
OpenBenchmarking.orgH/s, More Is BetterXmrig 6.12.1Variant: Wownero - Hash Count: 1MUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043K6K9K12K15KMin: 19128.1 / Avg: 19138.5 / Max: 19149Min: 19035.7 / Avg: 19059.63 / Max: 19084.7Min: 19106.2 / Avg: 19125.83 / Max: 19144.31. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

OSPray

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

OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: NASA Streamlines - Renderer: Path TracerUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.04, N = 310.5310.5310.49MIN: 9.35 / MAX: 10.64MIN: 10.31 / MAX: 10.75MIN: 10.2 / MAX: 10.75
OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: NASA Streamlines - Renderer: Path TracerUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 10.53 / Avg: 10.53 / Max: 10.53Min: 10.53 / Avg: 10.53 / Max: 10.53Min: 10.42 / Avg: 10.49 / Max: 10.53

BRL-CAD

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

OpenBenchmarking.orgVGR Performance Metric, More Is BetterBRL-CAD 7.32.2VGR Performance MetricUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0480K160K240K320K400K349004349508350303-pthread1. (CXX) g++ options: -std=c++11 -pipe -fvisibility=hidden -fno-strict-aliasing -fno-common -fexceptions -ftemplate-depth-128 -m64 -ggdb3 -O3 -fipa-pta -fstrength-reduce -finline-functions -flto -pedantic -ldl -lm

TNN

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

OpenBenchmarking.orgms, Fewer Is BetterTNN 0.3Target: CPU - Model: SqueezeNet v1.1Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0460120180240300SE +/- 0.02, N = 3SE +/- 0.07, N = 3SE +/- 0.23, N = 3251.44251.85250.98MIN: 251.22 / MAX: 253.43MIN: 251.62 / MAX: 252.41MIN: 250.4 / MAX: 251.411. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -fvisibility=default -O3 -rdynamic -ldl
OpenBenchmarking.orgms, Fewer Is BetterTNN 0.3Target: CPU - Model: SqueezeNet v1.1Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0450100150200250Min: 251.4 / Avg: 251.44 / Max: 251.48Min: 251.76 / Avg: 251.85 / Max: 252Min: 250.51 / Avg: 250.98 / Max: 251.231. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -fvisibility=default -O3 -rdynamic -ldl

Sysbench

This is a benchmark of Sysbench with the built-in CPU and memory sub-tests. Sysbench is a scriptable multi-threaded benchmark tool based on LuaJIT. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/sec, More Is BetterSysbench 1.0.20Test: RAM / MemoryUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042K4K6K8K10KSE +/- 29.24, N = 3SE +/- 12.34, N = 3SE +/- 19.16, N = 310249.8410278.2310283.19-pthread1. (CC) gcc options: -O2 -funroll-loops -rdynamic -ldl -laio -lm
OpenBenchmarking.orgMiB/sec, More Is BetterSysbench 1.0.20Test: RAM / MemoryUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042K4K6K8K10KMin: 10193.1 / Avg: 10249.84 / Max: 10290.46Min: 10257.61 / Avg: 10278.23 / Max: 10300.28Min: 10250.19 / Avg: 10283.19 / Max: 10316.551. (CC) gcc options: -O2 -funroll-loops -rdynamic -ldl -laio -lm

OSPray

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

OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: San Miguel - Renderer: Path TracerUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.69981.39962.09942.79923.499SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 33.103.113.11MIN: 3.01 / MAX: 3.12MIN: 3.05 / MAX: 3.13MIN: 3.08 / MAX: 3.13
OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: San Miguel - Renderer: Path TracerUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 3.1 / Avg: 3.1 / Max: 3.11Min: 3.11 / Avg: 3.11 / Max: 3.11Min: 3.1 / Avg: 3.11 / Max: 3.12

Appleseed

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

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: EmilyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.044080120160200163.07162.67162.55

Blender

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

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: BMW27 - Compute: CPU-OnlyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041326395265SE +/- 0.11, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 359.9159.7959.73
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: BMW27 - Compute: CPU-OnlyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041224364860Min: 59.71 / Avg: 59.91 / Max: 60.08Min: 59.74 / Avg: 59.79 / Max: 59.87Min: 59.71 / Avg: 59.73 / Max: 59.78

Cpuminer-Opt

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

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: Triple SHA-256, OnecoinUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04110K220K330K440K550KSE +/- 1780.96, N = 3SE +/- 526.67, N = 3SE +/- 2697.29, N = 35080475065535066971. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: Triple SHA-256, OnecoinUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0490K180K270K360K450KMin: 505890 / Avg: 508046.67 / Max: 511580Min: 505680 / Avg: 506553.33 / Max: 507500Min: 501550 / Avg: 506696.67 / Max: 5106701. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Embree

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

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer ISPC - Model: Asian DragonUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835SE +/- 0.20, N = 3SE +/- 0.06, N = 3SE +/- 0.09, N = 331.4031.4631.39MIN: 30.66 / MAX: 32.29MIN: 30.9 / MAX: 32.08MIN: 30.74 / MAX: 32.14
OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer ISPC - Model: Asian DragonUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835Min: 31.13 / Avg: 31.4 / Max: 31.79Min: 31.37 / Avg: 31.46 / Max: 31.57Min: 31.22 / Avg: 31.39 / Max: 31.5

Cpuminer-Opt

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

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: MagiUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0430060090012001500SE +/- 1.98, N = 3SE +/- 1.27, N = 3SE +/- 5.50, N = 31223.671221.381224.031. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: MagiUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042004006008001000Min: 1220.03 / Avg: 1223.67 / Max: 1226.82Min: 1218.84 / Avg: 1221.38 / Max: 1222.66Min: 1213.43 / Avg: 1224.03 / Max: 1231.891. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

NAMD

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

OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.14ATPase Simulation - 327,506 AtomsUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.17490.34980.52470.69960.8745SE +/- 0.00218, N = 3SE +/- 0.00109, N = 3SE +/- 0.00019, N = 30.777300.777250.77608
OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.14ATPase Simulation - 327,506 AtomsUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 0.77 / Avg: 0.78 / Max: 0.78Min: 0.78 / Avg: 0.78 / Max: 0.78Min: 0.78 / Avg: 0.78 / Max: 0.78

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 3.2Preset: ExhaustiveUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 324.0524.0624.081. (CXX) g++ options: -O3 -flto -pthread
OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 3.2Preset: ExhaustiveUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430Min: 24.04 / Avg: 24.05 / Max: 24.05Min: 24.05 / Avg: 24.06 / Max: 24.07Min: 24.07 / Avg: 24.08 / Max: 24.091. (CXX) g++ options: -O3 -flto -pthread

OpenSSL

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

OpenBenchmarking.orgsign/s, More Is BetterOpenSSL 3.0Algorithm: RSA4096Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0414002800420056007000SE +/- 0.35, N = 3SE +/- 4.34, N = 3SE +/- 0.67, N = 36529.16520.26530.31. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenBenchmarking.orgsign/s, More Is BetterOpenSSL 3.0Algorithm: RSA4096Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0411002200330044005500Min: 6528.5 / Avg: 6529.07 / Max: 6529.7Min: 6515.1 / Avg: 6520.17 / Max: 6528.8Min: 6529 / Avg: 6530.3 / Max: 6531.21. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

Sysbench

This is a benchmark of Sysbench with the built-in CPU and memory sub-tests. Sysbench is a scriptable multi-threaded benchmark tool based on LuaJIT. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgEvents Per Second, More Is BetterSysbench 1.0.20Test: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420K40K60K80K100KSE +/- 6.02, N = 3SE +/- 5.29, N = 3SE +/- 20.32, N = 3114562.03114579.76114490.11-pthread1. (CC) gcc options: -O2 -funroll-loops -rdynamic -ldl -laio -lm
OpenBenchmarking.orgEvents Per Second, More Is BetterSysbench 1.0.20Test: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420K40K60K80K100KMin: 114550.57 / Avg: 114562.03 / Max: 114570.95Min: 114571.75 / Avg: 114579.76 / Max: 114589.75Min: 114458.64 / Avg: 114490.11 / Max: 114528.111. (CC) gcc options: -O2 -funroll-loops -rdynamic -ldl -laio -lm

OpenSSL

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

OpenBenchmarking.orgverify/s, More Is BetterOpenSSL 3.0Algorithm: RSA4096Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0490K180K270K360K450KSE +/- 30.40, N = 3SE +/- 52.16, N = 3SE +/- 74.56, N = 3426052.9425838.4425938.61. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenBenchmarking.orgverify/s, More Is BetterOpenSSL 3.0Algorithm: RSA4096Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0470K140K210K280K350KMin: 426007.5 / Avg: 426052.87 / Max: 426110.6Min: 425774.2 / Avg: 425838.4 / Max: 425941.7Min: 425853.3 / Avg: 425938.63 / Max: 426087.21. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

Natron

Natron is an open-source, cross-platform compositing software for visual effects (VFX) and motion graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterNatron 2.4Input: SpaceshipUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041.14752.2953.44254.595.7375SE +/- 0.00, N = 3SE +/- 0.07, N = 3SE +/- 0.05, N = 55.15.15.1
OpenBenchmarking.orgFPS, More Is BetterNatron 2.4Input: SpaceshipUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 5.1 / Avg: 5.1 / Max: 5.1Min: 5 / Avg: 5.07 / Max: 5.2Min: 4.9 / Avg: 5.06 / Max: 5.2

OSPray

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

OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: Magnetic Reconnection - Renderer: Path TracerUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04110220330440550SE +/- 0.00, N = 3SE +/- 0.00, N = 3500500500MIN: 250MIN: 250MIN: 250
OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: Magnetic Reconnection - Renderer: Path TracerUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0490180270360450Min: 500 / Avg: 500 / Max: 500Min: 500 / Avg: 500 / Max: 500

OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: Magnetic Reconnection - Renderer: SciVisUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04714212835SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 328.5728.5728.57MIN: 27.03MIN: 22.73 / MAX: 29.41MIN: 27.78 / MAX: 29.41
OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: Magnetic Reconnection - Renderer: SciVisUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430Min: 28.57 / Avg: 28.57 / Max: 28.57Min: 28.57 / Avg: 28.57 / Max: 28.57Min: 28.57 / Avg: 28.57 / Max: 28.57

OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: San Miguel - Renderer: SciVisUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04918273645SE +/- 0.47, N = 3SE +/- 0.47, N = 3SE +/- 0.47, N = 337.5137.5137.51MIN: 34.48 / MAX: 40MIN: 35.71 / MAX: 40MIN: 35.71 / MAX: 40
OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: San Miguel - Renderer: SciVisUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04816243240Min: 37.04 / Avg: 37.51 / Max: 38.46Min: 37.04 / Avg: 37.51 / Max: 38.46Min: 37.04 / Avg: 37.51 / Max: 38.46

ONNX Runtime

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

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.8.2Model: super-resolution-10 - Device: OpenMP CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0414002800420056007000SE +/- 171.09, N = 12SE +/- 48.77, N = 12SE +/- 0.83, N = 35914505663661. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt
OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.8.2Model: super-resolution-10 - Device: OpenMP CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0411002200330044005500Min: 5111.5 / Avg: 5913.75 / Max: 6442.5Min: 4878 / Avg: 5055.63 / Max: 5341.5Min: 6365 / Avg: 6365.83 / Max: 6367.51. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.8.2Model: bertsquad-10 - Device: OpenMP CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.042004006008001000SE +/- 20.02, N = 12SE +/- 8.70, N = 12SE +/- 8.53, N = 57516487941. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt
OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.8.2Model: bertsquad-10 - Device: OpenMP CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04140280420560700Min: 638 / Avg: 750.83 / Max: 807.5Min: 622 / Avg: 648.17 / Max: 720.5Min: 761 / Avg: 794.4 / Max: 8071. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt

PlaidML

This test profile uses PlaidML deep learning framework developed by Intel for offering up various benchmarks. Learn more via the OpenBenchmarking.org test page.

FP16: No - Mode: Inference - Network: NASNer Large - Device: CPU

Ubuntu 21.10: Test failed to run.

Ubuntu 21.10 + Linux 5.15: Test failed to run.

Ubuntu 21.04: Test failed to run.

FP16: No - Mode: Inference - Network: Inception V3 - Device: CPU

Ubuntu 21.10: Test failed to run.

Ubuntu 21.10 + Linux 5.15: Test failed to run.

Ubuntu 21.04: Test failed to run.

FP16: No - Mode: Inference - Network: DenseNet 201 - Device: CPU

Ubuntu 21.10: Test failed to run.

Ubuntu 21.10 + Linux 5.15: Test failed to run.

Ubuntu 21.04: Test failed to run.

FP16: No - Mode: Inference - Network: Mobilenet - Device: CPU

Ubuntu 21.10: Test failed to run.

Ubuntu 21.10 + Linux 5.15: Test failed to run.

Ubuntu 21.04: Test failed to run.

FP16: No - Mode: Inference - Network: IMDB LSTM - Device: CPU

Ubuntu 21.10: Test failed to run.

Ubuntu 21.10 + Linux 5.15: Test failed to run.

Ubuntu 21.04: Test failed to run.

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: alexnetUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041.2512.5023.7535.0046.255SE +/- 0.25, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 34.985.564.99MIN: 4.57 / MAX: 8.33MIN: 5.35 / MAX: 49.86-pthread - MIN: 4.87 / MAX: 5.61. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: alexnetUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 4.73 / Avg: 4.98 / Max: 5.48Min: 5.5 / Avg: 5.56 / Max: 5.65Min: 4.98 / Avg: 4.99 / Max: 51. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: blazefaceUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.66831.33662.00492.67323.3415SE +/- 0.02, N = 3SE +/- 0.12, N = 3SE +/- 0.01, N = 32.742.972.80MIN: 2.67 / MAX: 3.22MIN: 2.7 / MAX: 104.82-pthread - MIN: 2.73 / MAX: 3.371. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: blazefaceUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 2.72 / Avg: 2.74 / Max: 2.78Min: 2.77 / Avg: 2.97 / Max: 3.18Min: 2.78 / Avg: 2.8 / Max: 2.821. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Mobile Neural Network

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

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: resnet-v2-50Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430SE +/- 0.39, N = 14SE +/- 0.42, N = 15SE +/- 0.28, N = 1521.0220.7123.02MIN: 18.32 / MAX: 25.57MIN: 18.2 / MAX: 28.32MIN: 20.47 / MAX: 24.671. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl
OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: resnet-v2-50Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04510152025Min: 18.47 / Avg: 21.02 / Max: 22.31Min: 18.4 / Avg: 20.71 / Max: 22.98Min: 20.64 / Avg: 23.02 / Max: 24.261. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: mobilenetV3Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.040.58881.17761.76642.35522.944SE +/- 0.034, N = 14SE +/- 0.037, N = 15SE +/- 0.045, N = 152.3702.3752.617MIN: 2.11 / MAX: 5.91MIN: 2.15 / MAX: 5.63MIN: 2.36 / MAX: 3.221. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl
OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: mobilenetV3Ubuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810Min: 2.15 / Avg: 2.37 / Max: 2.54Min: 2.19 / Avg: 2.38 / Max: 2.58Min: 2.39 / Avg: 2.62 / Max: 2.811. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Stress-NG

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

Test: RdRand

Ubuntu 21.10: stress-ng: error: [1436129] No stress workers invoked (one or more were unsupported)

Ubuntu 21.10 + Linux 5.15: stress-ng: error: [711647] No stress workers invoked (one or more were unsupported)

Ubuntu 21.04: stress-ng: error: [1403499] No stress workers invoked (one or more were unsupported)

Redis

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

OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 6.0.9Test: SETUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04500K1000K1500K2000K2500KSE +/- 39012.54, N = 15SE +/- 45756.31, N = 15SE +/- 27973.62, N = 152195897.792148023.022070185.381. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3
OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 6.0.9Test: SETUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04400K800K1200K1600K2000KMin: 2006433.25 / Avg: 2195897.79 / Max: 2508175.5Min: 1968154.38 / Avg: 2148023.02 / Max: 2545193.25Min: 1831837.25 / Avg: 2070185.38 / Max: 2269639.51. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3

PostgreSQL pgbench

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

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 500 - Mode: Read Write - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430SE +/- 0.24, N = 3SE +/- 0.44, N = 12SE +/- 1.15, N = 1224.1715.8922.85-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 500 - Mode: Read Write - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04612182430Min: 23.84 / Avg: 24.17 / Max: 24.64Min: 13.02 / Avg: 15.89 / Max: 18.41Min: 16.11 / Avg: 22.85 / Max: 29.81. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 500 - Mode: Read WriteUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.047K14K21K28K35KSE +/- 204.70, N = 3SE +/- 916.28, N = 12SE +/- 1259.88, N = 12206883175022566-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 500 - Mode: Read WriteUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.046K12K18K24K30KMin: 20292.24 / Avg: 20687.79 / Max: 20976.99Min: 27153.61 / Avg: 31750.42 / Max: 38405.57Min: 16780.15 / Avg: 22565.71 / Max: 31039.331. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 250 - Mode: Read Write - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215SE +/- 0.060, N = 3SE +/- 0.255, N = 12SE +/- 0.229, N = 99.4446.3495.961-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 250 - Mode: Read Write - Average LatencyUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 9.36 / Avg: 9.44 / Max: 9.56Min: 5.02 / Avg: 6.35 / Max: 7.45Min: 5.2 / Avg: 5.96 / Max: 7.331. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 250 - Mode: Read WriteUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.049K18K27K36K45KSE +/- 167.14, N = 3SE +/- 1677.10, N = 12SE +/- 1518.58, N = 9264744011742401-lpthread -lrt -ldl1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 250 - Mode: Read WriteUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.047K14K21K28K35KMin: 26156.56 / Avg: 26473.65 / Max: 26723.84Min: 33568.71 / Avg: 40116.74 / Max: 49793.6Min: 34115.76 / Avg: 42401.33 / Max: 48099.451. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

Cpuminer-Opt

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

OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: RingcoinUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.047001400210028003500SE +/- 38.34, N = 15SE +/- 46.58, N = 15SE +/- 52.00, N = 153378.783390.463321.231. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.18Algorithm: RingcoinUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.046001200180024003000Min: 3044.75 / Avg: 3378.78 / Max: 3525.2Min: 3030.71 / Avg: 3390.46 / Max: 3513.48Min: 3024.73 / Avg: 3321.23 / Max: 3559.481. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

LuxCoreRender

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

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Rainbow Colors and Prism - Acceleration: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620SE +/- 0.43, N = 15SE +/- 0.13, N = 15SE +/- 0.36, N = 1514.1014.1114.75MIN: 10.13 / MAX: 16.63MIN: 11.99 / MAX: 14.83MIN: 11.75 / MAX: 16.5
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Rainbow Colors and Prism - Acceleration: CPUUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0448121620Min: 10.21 / Avg: 14.1 / Max: 16.55Min: 12.82 / Avg: 14.11 / Max: 14.75Min: 12.63 / Avg: 14.75 / Max: 16.32

Rodinia

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

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP StreamclusterUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.04246810SE +/- 0.032, N = 3SE +/- 0.086, N = 15SE +/- 0.110, N = 156.6956.5586.2921. (CXX) g++ options: -O2 -lOpenCL
OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP StreamclusterUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.043691215Min: 6.64 / Avg: 6.7 / Max: 6.75Min: 5.68 / Avg: 6.56 / Max: 6.85Min: 5.68 / Avg: 6.29 / Max: 6.621. (CXX) g++ options: -O2 -lOpenCL

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP HotSpot3DUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.0420406080100SE +/- 1.39, N = 15SE +/- 1.29, N = 15SE +/- 1.28, N = 1580.3179.4078.191. (CXX) g++ options: -O2 -lOpenCL
OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenMP HotSpot3DUbuntu 21.10 + Linux 5.15Ubuntu 21.10Ubuntu 21.041530456075Min: 73.45 / Avg: 80.31 / Max: 90.76Min: 73.44 / Avg: 79.4 / Max: 84.32Min: 73.5 / Avg: 78.19 / Max: 84.51. (CXX) g++ options: -O2 -lOpenCL

Parboil

The Parboil Benchmarks from the IMPACT Research Group at University of Illinois are a set of throughput computing applications for looking at computing architecture and compilers. Parboil test-cases support OpenMP, OpenCL, and CUDA multi-processing environments. However, at this time the test profile is just making use of the OpenMP and OpenCL test workloads. Learn more via the OpenBenchmarking.org test page.

Test: OpenMP MRI Gridding

Ubuntu 21.10: /usr/bin/env: ‘python’: No such file or directory

Ubuntu 21.10 + Linux 5.15: /usr/bin/env: ‘python’: No such file or directory

Ubuntu 21.04: /usr/bin/env: ‘python’: No such file or directory

Test: OpenMP Stencil

Ubuntu 21.10: /usr/bin/env: ‘python’: No such file or directory

Ubuntu 21.10 + Linux 5.15: /usr/bin/env: ‘python’: No such file or directory

Ubuntu 21.04: /usr/bin/env: ‘python’: No such file or directory

Test: OpenMP MRI-Q

Ubuntu 21.10: /usr/bin/env: ‘python’: No such file or directory

Ubuntu 21.10 + Linux 5.15: /usr/bin/env: ‘python’: No such file or directory

Ubuntu 21.04: /usr/bin/env: ‘python’: No such file or directory

Test: OpenMP CUTCP

Ubuntu 21.10: /usr/bin/env: ‘python’: No such file or directory

Ubuntu 21.10 + Linux 5.15: /usr/bin/env: ‘python’: No such file or directory

Ubuntu 21.04: /usr/bin/env: ‘python’: No such file or directory

Test: OpenMP LBM

Ubuntu 21.10: /usr/bin/env: ‘python’: No such file or directory

Ubuntu 21.10 + Linux 5.15: /usr/bin/env: ‘python’: No such file or directory

Ubuntu 21.04: /usr/bin/env: ‘python’: No such file or directory

199 Results Shown

Stress-NG
Cpuminer-Opt
Stress-NG:
  System V Message Passing
  NUMA
  Socket Activity
PostgreSQL pgbench:
  100 - 250 - Read Write - Average Latency
  100 - 250 - Read Write
Stress-NG
PostgreSQL pgbench:
  100 - 500 - Read Write
  100 - 500 - Read Write - Average Latency
Cpuminer-Opt
Stress-NG
libavif avifenc
nginx
Stress-NG
Apache HTTP Server
nginx:
  100
  200
  500
NAS Parallel Benchmarks
nginx
Stress-NG
Rodinia
Mobile Neural Network
Google SynthMark
Mobile Neural Network:
  inception-v3
  SqueezeNetV1.0
libavif avifenc
PlaidML
Apache HTTP Server
LeelaChessZero
MariaDB
Redis
Apache HTTP Server
VP9 libvpx Encoding
x265
Cpuminer-Opt
Stress-NG:
  MEMFD
  Glibc Qsort Data Sorting
PlaidML
NAS Parallel Benchmarks
VP9 libvpx Encoding
Mobile Neural Network
Rodinia
JPEG XL libjxl
Stress-NG
TNN
libavif avifenc
NCNN
PostgreSQL pgbench:
  100 - 500 - Read Only - Average Latency
  100 - 500 - Read Only
NCNN:
  CPU - regnety_400m
  CPU - resnet50
  CPU - vgg16
JPEG XL libjxl
NCNN
Apache HTTP Server
Mobile Neural Network
NCNN
Apache HTTP Server
x265
Apache HTTP Server
Xcompact3d Incompact3d
Stress-NG
nginx
Stress-NG
NCNN
simdjson:
  PartialTweets
  LargeRand
NAS Parallel Benchmarks
LAMMPS Molecular Dynamics Simulator
TensorFlow Lite
ONNX Runtime
NCNN
NAS Parallel Benchmarks
JPEG XL libjxl
OSPray
JPEG XL libjxl
TNN
NCNN
JPEG XL libjxl:
  PNG - 7
  PNG - 5
OpenFOAM
NCNN
Stress-NG
NCNN
Stress-NG
TNN
simdjson
Timed Godot Game Engine Compilation
TTSIOD 3D Renderer
AOM AV1
Timed Linux Kernel Compilation
NAS Parallel Benchmarks:
  SP.C
  SP.B
ONNX Runtime
Rodinia
ASTC Encoder
Timed Node.js Compilation
libavif avifenc
Stress-NG
libavif avifenc
Cpuminer-Opt
Stress-NG
ONNX Runtime
libavif avifenc
7-Zip Compression
PostgreSQL pgbench:
  1000 - 500 - Read Only
  1000 - 500 - Read Only - Average Latency
LuxCoreRender
Facebook RocksDB
SVT-VP9
PostgreSQL pgbench
simdjson
LuxCoreRender
SecureMark
NCNN
PostgreSQL pgbench
TensorFlow Lite
LAMMPS Molecular Dynamics Simulator
SVT-AV1
SVT-HEVC
Cpuminer-Opt
Chaos Group V-RAY
PlaidML
Facebook RocksDB
NAS Parallel Benchmarks
LeelaChessZero
oneDNN
TensorFlow Lite
Stress-NG:
  Crypto
  Atomic
AOM AV1
Xcompact3d Incompact3d
NCNN
Timed LLVM Compilation:
  Ninja
  Unix Makefiles
PostgreSQL pgbench
SVT-HEVC
PostgreSQL pgbench
OpenFOAM
oneDNN
Facebook RocksDB
Intel Open Image Denoise
GROMACS
Embree
Appleseed
Embree
TensorFlow Lite:
  Inception V4
  SqueezeNet
Apache Cassandra
Embree
AOM AV1
LuxCoreRender
OpenSSL
TensorFlow Lite
Cpuminer-Opt
Xmrig
Cpuminer-Opt
LuxCoreRender
Cpuminer-Opt
Blender
Stress-NG
Stockfish
Xmrig
OSPray
BRL-CAD
TNN
Sysbench
OSPray
Appleseed
Blender
Cpuminer-Opt
Embree
Cpuminer-Opt
NAMD
ASTC Encoder
OpenSSL
Sysbench
OpenSSL
Natron
OSPray:
  Magnetic Reconnection - Path Tracer
  Magnetic Reconnection - SciVis
  San Miguel - SciVis
ONNX Runtime:
  super-resolution-10 - OpenMP CPU
  bertsquad-10 - OpenMP CPU
NCNN:
  CPU - alexnet
  CPU - blazeface
Mobile Neural Network:
  resnet-v2-50
  mobilenetV3
Redis
PostgreSQL pgbench:
  1000 - 500 - Read Write - Average Latency
  1000 - 500 - Read Write
  1000 - 250 - Read Write - Average Latency
  1000 - 250 - Read Write
Cpuminer-Opt
LuxCoreRender
Rodinia:
  OpenMP Streamcluster
  OpenMP HotSpot3D