Linux 5.10 Ryzen 7 3800XT

AMD Ryzen 7 3800XT 8-Core testing with a ASRock X570 Pro4 (P3.00 BIOS) and Sapphire AMD Radeon RX 470/480/570/570X/580/580X/590 8GB on Ubuntu 20.04 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2010245-FI-LINUX510R37
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 2 Tests
BLAS (Basic Linear Algebra Sub-Routine) Tests 3 Tests
Web Browsers 1 Tests
C/C++ Compiler Tests 10 Tests
CPU Massive 14 Tests
Creator Workloads 9 Tests
Database Test Suite 4 Tests
Desktop Graphics 2 Tests
Encoding 4 Tests
Fortran Tests 2 Tests
HPC - High Performance Computing 12 Tests
Imaging 2 Tests
Common Kernel Benchmarks 5 Tests
Machine Learning 6 Tests
Molecular Dynamics 4 Tests
MPI Benchmarks 3 Tests
Multi-Core 10 Tests
NVIDIA GPU Compute 8 Tests
OpenMPI Tests 3 Tests
Python Tests 4 Tests
Scientific Computing 6 Tests
Server 5 Tests
Server CPU Tests 9 Tests
Video Encoding 4 Tests
Vulkan Compute 4 Tests
Common Workstation Benchmarks 2 Tests

Statistics

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

Graph Settings

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

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
Linux 5.9.1
October 23 2020
  12 Hours, 35 Minutes
Linux 5.10 Git 23 Oct
October 24 2020
  11 Hours, 26 Minutes
Invert Hiding All Results Option
  12 Hours, 1 Minute
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):

Linux 5.9.1

Processor: AMD Ryzen 7 3800XT 8-Core @ 3.90GHz (8 Cores / 16 Threads), Motherboard: ASRock X570 Pro4 (P3.00 BIOS), Chipset: AMD Starship/Matisse, Memory: 16GB, Disk: 500GB CT500P2SSD8, Graphics: Sapphire AMD Radeon RX 470/480/570/570X/580/580X/590 8GB (1560/2100MHz), Audio: AMD Ellesmere HDMI Audio, Monitor: G237HL, Network: Intel I211

OS: Ubuntu 20.04, Kernel: 5.9.1-050901-generic (x86_64), Desktop: GNOME Shell 3.36.4, Display Server: X Server 1.20.8, Display Driver: modesetting 1.20.8, OpenGL: 4.6 Mesa 20.0.4 (LLVM 9.0.1), Vulkan: 1.2.128, Compiler: GCC 9.3.0, File-System: ext4, Screen Resolution: 1920x1080

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,gm2 --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-9-HskZEa/gcc-9-9.3.0/debian/tmp-nvptx/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Disk Notes: NONE / errors=remount-ro,relatime,rw
Processor Notes: Scaling Governor: acpi-cpufreq ondemand - CPU Microcode: 0x8701021
Graphics Notes: GLAMOR
Python Notes: Python 3.8.5
Security Notes: 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 STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected

Linux 5.10 Git 23 Oct

OS: Ubuntu 20.04, Kernel: 5.9.0-050900daily20201023-generic (x86_64), Desktop: GNOME Shell 3.36.4, Display Server: X Server 1.20.8, Display Driver: modesetting 1.20.8, OpenGL: 4.6 Mesa 20.0.4 (LLVM 9.0.1), Vulkan: 1.2.128, Compiler: GCC 9.3.0, File-System: ext4, Screen Resolution: 1920x1080

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,gm2 --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-9-HskZEa/gcc-9-9.3.0/debian/tmp-nvptx/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Disk Notes: NONE / errors=remount-ro,no_fc,relatime,rw
Processor Notes: Scaling Governor: acpi-cpufreq ondemand - CPU Microcode: 0x8701021
Graphics Notes: GLAMOR
Python Notes: Python 3.8.5
Security Notes: 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 STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected

Linux 5.10 Ryzen 7 3800XTOpenBenchmarking.orgPhoronix Test SuiteAMD Ryzen 7 3800XT 8-Core @ 3.90GHz (8 Cores / 16 Threads)ASRock X570 Pro4 (P3.00 BIOS)AMD Starship/Matisse16GB500GB CT500P2SSD8Sapphire AMD Radeon RX 470/480/570/570X/580/580X/590 8GB (1560/2100MHz)AMD Ellesmere HDMI AudioG237HLIntel I211Ubuntu 20.045.9.1-050901-generic (x86_64)5.9.0-050900daily20201023-generic (x86_64)GNOME Shell 3.36.4X Server 1.20.8modesetting 1.20.84.6 Mesa 20.0.4 (LLVM 9.0.1)1.2.128GCC 9.3.0ext41920x1080ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelsDesktopDisplay ServerDisplay DriverOpenGLVulkanCompilerFile-SystemScreen ResolutionLinux 5.10 Ryzen 7 3800XT BenchmarksSystem Logs- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,gm2 --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-9-HskZEa/gcc-9-9.3.0/debian/tmp-nvptx/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v - Linux 5.9.1: NONE / errors=remount-ro,relatime,rw- Linux 5.10 Git 23 Oct: NONE / errors=remount-ro,no_fc,relatime,rw- Scaling Governor: acpi-cpufreq ondemand - CPU Microcode: 0x8701021- GLAMOR- Python 3.8.5- 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 STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected

Linux 5.9.1 vs. Linux 5.10 Git 23 Oct ComparisonPhoronix Test SuiteBaseline+11.6%+11.6%+23.2%+23.2%+34.8%+34.8%+46.4%+46.4%22.2%17.5%13.4%8.7%4.2%3.9%3.1%2.7%2.5%2.5%2.2%Latency Under Load46.5%Vulkan GPU-v2-v2 - mobilenet-v225.5%Vulkan GPU - resnet1823.6%Vulkan GPU - efficientnet-b023.3%Vulkan GPU-v3-v3 - mobilenet-v322.7%Write TestVulkan GPU - mnasnet19.2%Vulkan GPU - googlenet18%Vulkan GPU - squeezenet17.6%S.V.M.PVulkan GPU - resnet5015.8%8 - Process14.8%4 - Process14.6%8 - Thread14.3%4 - Thread14.2%16 - Process13.7%Read Test16 - Thread13.2%2 - Process13%2 - Thread12.3%Chimera 1080p12.3%Vulkan GPU - shufflenet-v211.8%32 - Process11.4%20k Atoms10.4%Summer Nature 4K10.3%S.N.19.1%Fill SyncRhodopsin Protein8.3%Memory Copying7.2%1 - Process6.3%1 - Thread5.6%Socket Activity4.8%MotionMark - Firefox4.5%D.T.P4.2%Create Threads504.1%Vulkan GPU - mobilenet3.9%EigenForking3.4%103.2%MMAPCreate Files2.8%Carbon Nanotube2.7%CanvasMark - Firefox2.7%G.C.S.FCPU Cache2.6%2502.6%ARES-6 - FirefoxQ.1.LBasemark - Firefox2.4%MEMFDP.P.B2.2%2x - 3 - No2.1%SockperfNCNNNCNNNCNNNCNNIORNCNNNCNNNCNNStress-NGNCNNHackbenchHackbenchHackbenchHackbenchHackbenchIORHackbenchHackbenchHackbenchdav1dNCNNHackbenchLAMMPS Molecular Dynamics Simulatordav1ddav1dLevelDBLAMMPS Molecular Dynamics SimulatorStress-NGHackbenchHackbenchStress-NGSeleniumPostMarkOSBenchApache SiegeNCNNLeelaChessZeroStress-NGApache SiegeStress-NGOSBenchGPAWSeleniumStress-NGStress-NGApache SiegeSeleniumWebP Image EncodeSeleniumStress-NGLibRawWaifu2x-NCNN VulkanLinux 5.9.1Linux 5.10 Git 23 Oct

Linux 5.10 Ryzen 7 3800XTncnn: Vulkan GPU-v2-v2 - mobilenet-v2ncnn: Vulkan GPU - resnet18ncnn: Vulkan GPU - efficientnet-b0ncnn: Vulkan GPU-v3-v3 - mobilenet-v3ncnn: Vulkan GPU - mnasnetncnn: Vulkan GPU - googlenetncnn: Vulkan GPU - squeezenetstress-ng: System V Message Passingncnn: Vulkan GPU - resnet50hackbench: 8 - Processhackbench: 4 - Processhackbench: 8 - Threadhackbench: 4 - Threadhackbench: 16 - Processhackbench: 16 - Threadhackbench: 2 - Processhackbench: 2 - Threaddav1d: Chimera 1080pncnn: Vulkan GPU - shufflenet-v2hackbench: 32 - Processlammps: 20k Atomsdav1d: Summer Nature 4Kdav1d: Summer Nature 1080plammps: Rhodopsin Proteinstress-ng: Memory Copyinghackbench: 1 - Processhackbench: 1 - Threadstress-ng: Socket Activitypostmark: Disk Transaction Performanceosbench: Create Threadsapache-siege: 50ncnn: Vulkan GPU - mobilenetlczero: Eigenstress-ng: Forkingapache-siege: 10stress-ng: MMAPosbench: Create Filesgpaw: Carbon Nanotubeselenium: CanvasMark - Firefoxstress-ng: Glibc C String Functionsstress-ng: CPU Cacheapache-siege: 250selenium: ARES-6 - Firefoxwebp: Quality 100, Losslessselenium: Basemark - Firefoxstress-ng: MEMFDlibraw: Post-Processing Benchmarkwaifu2x-ncnn: 2x - 3 - Nowebp: Quality 100, Highest Compressionkeydb: openvino: Age Gender Recognition Retail 0013 FP32 - CPUsockperf: Latency Ping Pongncnn: Vulkan GPU - yolov4-tinystress-ng: SENDFILEleveldb: Hot Readapache-siege: 100dolfyn: Computational Fluid Dynamicscouchdb: 100 - 1000 - 24leveldb: Rand Readselenium: Kraken - Firefoxlczero: BLASespeak: Text-To-Speech Synthesisapache-siege: 200waifu2x-ncnn: 2x - 3 - Yeswebp: Quality 100stress-ng: Context Switchingtnn: CPU - SqueezeNet v1.1vkfft: osbench: Launch Programscompress-zstd: 19stress-ng: Mallocncnn: Vulkan GPU - vgg16x265: Bosphorus 4Krealsr-ncnn: 4x - Noselenium: Jetstream - Firefoxrealsr-ncnn: 4x - Yesdav1d: Chimera 1080p 10-bittensorflow-lite: NASNet Mobilekvazaar: Bosphorus 1080p - Very Fastsvt-av1: Enc Mode 0 - 1080pstress-ng: Semaphoresstress-ng: NUMAkvazaar: Bosphorus 1080p - Ultra Fastncnn: CPU - resnet18tnn: CPU - MobileNet v2ncnn: CPU-v3-v3 - mobilenet-v3compress-zstd: 3openvino: Person Detection 0106 FP32 - CPUncnn: CPU - vgg16ncnn: CPU - efficientnet-b0leveldb: Overwritesvt-av1: Enc Mode 4 - 1080pncnn: CPU - mnasnetosbench: Memory Allocationssvt-av1: Enc Mode 8 - 1080popenvino: Face Detection 0106 FP16 - CPUncnn: CPU - blazefacekvazaar: Bosphorus 4K - Slowx265: Bosphorus 1080pselenium: Octane - Firefoxleveldb: Overwritewebp: Defaultmafft: Multiple Sequence Alignment - LSU RNAopenvino: Face Detection 0106 FP16 - CPUkvazaar: Bosphorus 1080p - Slowstress-ng: Glibc Qsort Data Sortingkvazaar: Bosphorus 1080p - Mediumkvazaar: Bosphorus 4K - Mediumleveldb: Rand Fillblender: Pabellon Barcelona - CPU-Onlyncnn: CPU - resnet50leveldb: Rand Fillleveldb: Seq Fillsockperf: Throughputleveldb: Seq Fillblender: Classroom - CPU-Onlystress-ng: Matrix Mathblender: Barbershop - CPU-Onlyblender: Fishy Cat - CPU-Onlyopenvino: Age Gender Recognition Retail 0013 FP16 - CPUncnn: CPU - shufflenet-v2kvazaar: Bosphorus 4K - Ultra Fastcaffe: GoogleNet - CPU - 100ncnn: CPU - alexnetleveldb: Seek Randblender: BMW27 - CPU-Onlycaffe: AlexNet - CPU - 100stress-ng: Cryptotensorflow-lite: Mobilenet Quantnamd: ATPase Simulation - 327,506 Atomsopenvino: Face Detection 0106 FP32 - CPUgromacs: Water Benchmarkselenium: StyleBench - Firefoxopenvino: Age Gender Recognition Retail 0013 FP32 - CPUncnn: CPU - mobilenettensorflow-lite: Inception ResNet V2tensorflow-lite: SqueezeNetstress-ng: Atomicopenvino: Person Detection 0106 FP32 - CPUkvazaar: Bosphorus 4K - Very Faststress-ng: CPU Stressncnn: CPU - squeezenettensorflow-lite: Inception V4openvino: Person Detection 0106 FP16 - CPUncnn: CPU - yolov4-tinyleveldb: Rand Deletetensorflow-lite: Mobilenet Floatwebp: Quality 100, Lossless, Highest Compressionstress-ng: Vector Mathbasemark: Vulkan - 1920 x 1080 - Mediumbasemark: Vulkan - 1920 x 1080 - Highxonotic: 1920 x 1080 - Ultimateunigine-super: 1920 x 1080 - Fullscreen - Ultra - OpenGLunigine-super: 1920 x 1080 - Fullscreen - High - OpenGLtesseract: 1920 x 1080influxdb: 1024 - 10000 - 2,5000,1 - 10000influxdb: 64 - 10000 - 2,5000,1 - 10000influxdb: 4 - 10000 - 2,5000,1 - 10000selenium: WASM collisionDetection - Firefoxselenium: WASM imageConvolute - Firefoxselenium: Speedometer - Firefoxselenium: Maze Solver - Firefoxselenium: Jetstream 2 - Firefoxselenium: WebXPRT - Firefoxopenvino: Age Gender Recognition Retail 0013 FP16 - CPUopenvino: Person Detection 0106 FP16 - CPUopenvino: Face Detection 0106 FP32 - CPUncnn: Vulkan GPU - alexnetncnn: Vulkan GPU - blazefacencnn: CPU - googlenetncnn: CPU-v2-v2 - mobilenet-v2selenium: MotionMark - Firefoxleveldb: Fill Syncleveldb: Fill Syncosbench: Create Processessockperf: Latency Under Loadior: Read Testior: Write TestLinux 5.9.1Linux 5.10 Git 23 Oct6.159.7615.147.356.8416.0914.8211459628.0223.4320.54410.75021.07811.12542.75643.6675.7746.003490.583.9994.6826.458159.70455.556.2143443.504.4914.7146765.8178969.76991788406.2311.3059351191.5634254.15184.2912.063223379.34614235583277.5921.3776561.9440.0515.426906.01589.6840.163.0356.723623244.730.573.56618.76151380.147.35092711.9415.41694.6707.316743.260028.25282538.8619.9442.1364257616.04237.7231085138.18353030.883874667.8353.6411.5447.531228.28363.352102.4116065059.410.1151161533.53219.61110.2919.15247.2575.603749.41.4976.307.9632.43.2465.0975.08134827.2801.871.895.7950.613782354.6451.39811.4932137.2824.96141.1725.595.8754.775498.4232.1032.333.469648653.027447.0440473.69648.28208.006806.094.7129.1712860717.2811.143155.68529172406.521267742.193582141.330.77585.66759.0719.502415433185313764049.462669.2316.393567.5318.6726728432665.5831.3550.60412401532.16376006.77606.02119.63315.018853718.041.2529.03091451332.51420949.51225807.0364.727.198.34.4107.9722730.571.491.879.981.5318.246.38116.886684.9440.320.60871967.366347.81159.917.7212.0618.679.028.1518.9917.4313469379.6127.1423.58512.31624.08912.70748.60349.4156.5246.744436.854.46105.4735.848144.76417.415.7363212.424.7744.9806454.5475759.38018284898.9711.7461649488.9933189.87189.9612.404537389.70613858598962.9520.8374632.4139.0615.054884.84602.8939.303.0996.851612175.880.583.62619.06148999.837.46791299.4215.65393.2417.428732.060927.84281351.0920.2352.1674198734.92234.4951070638.69374630.482826646.8654.3011.4048.114225.69367.415101.3616218558.890.1141151702.00217.83109.4219.00245.4375.563723.21.4875.807.9132.23.2265.0674.65497627.1261.881.885.7650.353764254.8961.40411.4442128.9024.87141.6825.505.8554.953500.0032.0032.233.369851353.170448.2240575.62646.68208.506822.194.7229.1112885817.2511.162155.43528332402.771269482.196432138.560.77485.76751.8619.482417670185481764678.982667.2916.383565.4218.6626741932666.6031.3650.61912405132.16776005.992730.571.491.879.981.5318.246.38111.816147.4870.320.92758898.686394.46195.45OpenBenchmarking.org

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2Linux 5.9.1Linux 5.10 Git 23 Oct246810SE +/- 0.03, N = 3SE +/- 0.03, N = 36.157.72MIN: 5.77 / MAX: 10.07MIN: 7.45 / MAX: 16.741. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2Linux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 6.1 / Avg: 6.15 / Max: 6.18Min: 7.67 / Avg: 7.72 / Max: 7.761. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: resnet18Linux 5.9.1Linux 5.10 Git 23 Oct3691215SE +/- 0.04, N = 3SE +/- 0.03, N = 39.7612.06MIN: 8.39 / MAX: 20.03MIN: 11.58 / MAX: 20.21. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: resnet18Linux 5.9.1Linux 5.10 Git 23 Oct48121620Min: 9.7 / Avg: 9.76 / Max: 9.83Min: 12.01 / Avg: 12.06 / Max: 12.091. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: efficientnet-b0Linux 5.9.1Linux 5.10 Git 23 Oct510152025SE +/- 0.04, N = 3SE +/- 0.04, N = 315.1418.67MIN: 14.22 / MAX: 21.44MIN: 18.2 / MAX: 26.481. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: efficientnet-b0Linux 5.9.1Linux 5.10 Git 23 Oct510152025Min: 15.07 / Avg: 15.14 / Max: 15.21Min: 18.59 / Avg: 18.67 / Max: 18.711. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3Linux 5.9.1Linux 5.10 Git 23 Oct3691215SE +/- 0.03, N = 3SE +/- 0.01, N = 37.359.02MIN: 6.95 / MAX: 16.06MIN: 8.64 / MAX: 18.111. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3Linux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 7.3 / Avg: 7.35 / Max: 7.4Min: 9.01 / Avg: 9.02 / Max: 9.051. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: mnasnetLinux 5.9.1Linux 5.10 Git 23 Oct246810SE +/- 0.04, N = 3SE +/- 0.06, N = 36.848.15MIN: 6.52 / MAX: 14.49MIN: 7.8 / MAX: 16.021. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: mnasnetLinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 6.8 / Avg: 6.84 / Max: 6.91Min: 8.08 / Avg: 8.15 / Max: 8.271. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: googlenetLinux 5.9.1Linux 5.10 Git 23 Oct510152025SE +/- 0.08, N = 3SE +/- 0.05, N = 316.0918.99MIN: 15.17 / MAX: 27.56MIN: 18.54 / MAX: 30.651. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: googlenetLinux 5.9.1Linux 5.10 Git 23 Oct510152025Min: 15.97 / Avg: 16.09 / Max: 16.25Min: 18.92 / Avg: 18.99 / Max: 19.11. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: squeezenetLinux 5.9.1Linux 5.10 Git 23 Oct48121620SE +/- 0.06, N = 3SE +/- 0.07, N = 314.8217.43MIN: 13.9 / MAX: 28.87MIN: 17.04 / MAX: 27.831. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: squeezenetLinux 5.9.1Linux 5.10 Git 23 Oct48121620Min: 14.74 / Avg: 14.82 / Max: 14.94Min: 17.3 / Avg: 17.43 / Max: 17.531. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: System V Message PassingLinux 5.9.1Linux 5.10 Git 23 Oct3M6M9M12M15MSE +/- 25794.08, N = 3SE +/- 17928.61, N = 311459628.0213469379.611. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: System V Message PassingLinux 5.9.1Linux 5.10 Git 23 Oct2M4M6M8M10MMin: 11411601.86 / Avg: 11459628.02 / Max: 11499954.28Min: 13434678.4 / Avg: 13469379.61 / Max: 13494551.641. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: resnet50Linux 5.9.1Linux 5.10 Git 23 Oct612182430SE +/- 0.12, N = 3SE +/- 0.12, N = 323.4327.14MIN: 21.98 / MAX: 37.49MIN: 22.09 / MAX: 36.291. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: resnet50Linux 5.9.1Linux 5.10 Git 23 Oct612182430Min: 23.19 / Avg: 23.43 / Max: 23.56Min: 26.96 / Avg: 27.14 / Max: 27.371. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Hackbench

This is a benchmark of Hackbench, a test of the Linux kernel scheduler. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 8 - Type: ProcessLinux 5.9.1Linux 5.10 Git 23 Oct612182430SE +/- 0.07, N = 3SE +/- 0.06, N = 320.5423.591. (CC) gcc options: -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 8 - Type: ProcessLinux 5.9.1Linux 5.10 Git 23 Oct612182430Min: 20.42 / Avg: 20.54 / Max: 20.66Min: 23.51 / Avg: 23.59 / Max: 23.71. (CC) gcc options: -lpthread

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 4 - Type: ProcessLinux 5.9.1Linux 5.10 Git 23 Oct3691215SE +/- 0.07, N = 3SE +/- 0.01, N = 310.7512.321. (CC) gcc options: -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 4 - Type: ProcessLinux 5.9.1Linux 5.10 Git 23 Oct48121620Min: 10.62 / Avg: 10.75 / Max: 10.85Min: 12.29 / Avg: 12.32 / Max: 12.331. (CC) gcc options: -lpthread

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 8 - Type: ThreadLinux 5.9.1Linux 5.10 Git 23 Oct612182430SE +/- 0.04, N = 3SE +/- 0.07, N = 321.0824.091. (CC) gcc options: -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 8 - Type: ThreadLinux 5.9.1Linux 5.10 Git 23 Oct612182430Min: 21.01 / Avg: 21.08 / Max: 21.14Min: 23.99 / Avg: 24.09 / Max: 24.221. (CC) gcc options: -lpthread

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 4 - Type: ThreadLinux 5.9.1Linux 5.10 Git 23 Oct3691215SE +/- 0.07, N = 3SE +/- 0.07, N = 311.1312.711. (CC) gcc options: -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 4 - Type: ThreadLinux 5.9.1Linux 5.10 Git 23 Oct48121620Min: 10.99 / Avg: 11.13 / Max: 11.23Min: 12.61 / Avg: 12.71 / Max: 12.851. (CC) gcc options: -lpthread

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 16 - Type: ProcessLinux 5.9.1Linux 5.10 Git 23 Oct1122334455SE +/- 0.01, N = 3SE +/- 0.08, N = 342.7648.601. (CC) gcc options: -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 16 - Type: ProcessLinux 5.9.1Linux 5.10 Git 23 Oct1020304050Min: 42.75 / Avg: 42.76 / Max: 42.77Min: 48.5 / Avg: 48.6 / Max: 48.761. (CC) gcc options: -lpthread

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 16 - Type: ThreadLinux 5.9.1Linux 5.10 Git 23 Oct1122334455SE +/- 0.10, N = 3SE +/- 0.07, N = 343.6749.421. (CC) gcc options: -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 16 - Type: ThreadLinux 5.9.1Linux 5.10 Git 23 Oct1020304050Min: 43.56 / Avg: 43.67 / Max: 43.86Min: 49.27 / Avg: 49.42 / Max: 49.511. (CC) gcc options: -lpthread

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 2 - Type: ProcessLinux 5.9.1Linux 5.10 Git 23 Oct246810SE +/- 0.033, N = 3SE +/- 0.029, N = 35.7746.5241. (CC) gcc options: -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 2 - Type: ProcessLinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 5.72 / Avg: 5.77 / Max: 5.83Min: 6.48 / Avg: 6.52 / Max: 6.581. (CC) gcc options: -lpthread

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 2 - Type: ThreadLinux 5.9.1Linux 5.10 Git 23 Oct246810SE +/- 0.031, N = 3SE +/- 0.055, N = 36.0036.7441. (CC) gcc options: -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 2 - Type: ThreadLinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 5.95 / Avg: 6 / Max: 6.06Min: 6.64 / Avg: 6.74 / Max: 6.831. (CC) gcc options: -lpthread

dav1d

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

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Chimera 1080pLinux 5.9.1Linux 5.10 Git 23 Oct110220330440550SE +/- 1.87, N = 3SE +/- 4.03, N = 10490.58436.85MIN: 377.92 / MAX: 704.59MIN: 320.22 / MAX: 628.691. (CC) gcc options: -pthread
OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Chimera 1080pLinux 5.9.1Linux 5.10 Git 23 Oct90180270360450Min: 486.9 / Avg: 490.58 / Max: 493.02Min: 402.53 / Avg: 436.85 / Max: 446.371. (CC) gcc options: -pthread

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: shufflenet-v2Linux 5.9.1Linux 5.10 Git 23 Oct1.00352.0073.01054.0145.0175SE +/- 0.01, N = 3SE +/- 0.01, N = 33.994.46MIN: 3.83 / MAX: 4.63MIN: 4.39 / MAX: 5.131. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: shufflenet-v2Linux 5.9.1Linux 5.10 Git 23 Oct246810Min: 3.98 / Avg: 3.99 / Max: 4Min: 4.45 / Avg: 4.46 / Max: 4.471. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Hackbench

This is a benchmark of Hackbench, a test of the Linux kernel scheduler. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 32 - Type: ProcessLinux 5.9.1Linux 5.10 Git 23 Oct20406080100SE +/- 0.87, N = 3SE +/- 1.42, N = 494.68105.471. (CC) gcc options: -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 32 - Type: ProcessLinux 5.9.1Linux 5.10 Git 23 Oct20406080100Min: 93.71 / Avg: 94.68 / Max: 96.42Min: 103.54 / Avg: 105.47 / Max: 109.691. (CC) gcc options: -lpthread

LAMMPS Molecular Dynamics Simulator

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

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 24Aug2020Model: 20k AtomsLinux 5.9.1Linux 5.10 Git 23 Oct246810SE +/- 0.028, N = 3SE +/- 0.017, N = 36.4585.8481. (CXX) g++ options: -O3 -pthread -lm
OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 24Aug2020Model: 20k AtomsLinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 6.43 / Avg: 6.46 / Max: 6.52Min: 5.82 / Avg: 5.85 / Max: 5.871. (CXX) g++ options: -O3 -pthread -lm

dav1d

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

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Summer Nature 4KLinux 5.9.1Linux 5.10 Git 23 Oct4080120160200SE +/- 0.07, N = 3SE +/- 1.51, N = 7159.70144.76MIN: 140.75 / MAX: 170.03MIN: 116.22 / MAX: 154.791. (CC) gcc options: -pthread
OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Summer Nature 4KLinux 5.9.1Linux 5.10 Git 23 Oct306090120150Min: 159.58 / Avg: 159.7 / Max: 159.82Min: 135.91 / Avg: 144.76 / Max: 147.141. (CC) gcc options: -pthread

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Summer Nature 1080pLinux 5.9.1Linux 5.10 Git 23 Oct100200300400500SE +/- 1.52, N = 3SE +/- 3.47, N = 3455.55417.41MIN: 381.07 / MAX: 495.69MIN: 343.92 / MAX: 453.561. (CC) gcc options: -pthread
OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Summer Nature 1080pLinux 5.9.1Linux 5.10 Git 23 Oct80160240320400Min: 453.9 / Avg: 455.55 / Max: 458.6Min: 410.49 / Avg: 417.41 / Max: 421.311. (CC) gcc options: -pthread

LAMMPS Molecular Dynamics Simulator

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

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 24Aug2020Model: Rhodopsin ProteinLinux 5.9.1Linux 5.10 Git 23 Oct246810SE +/- 0.064, N = 15SE +/- 0.067, N = 36.2145.7361. (CXX) g++ options: -O3 -pthread -lm
OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 24Aug2020Model: Rhodopsin ProteinLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 5.86 / Avg: 6.21 / Max: 6.64Min: 5.62 / Avg: 5.74 / Max: 5.851. (CXX) g++ options: -O3 -pthread -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 0.11.07Test: Memory CopyingLinux 5.9.1Linux 5.10 Git 23 Oct7001400210028003500SE +/- 16.57, N = 3SE +/- 33.77, N = 33443.503212.421. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Memory CopyingLinux 5.9.1Linux 5.10 Git 23 Oct6001200180024003000Min: 3419.57 / Avg: 3443.5 / Max: 3475.31Min: 3175.1 / Avg: 3212.42 / Max: 3279.831. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

Hackbench

This is a benchmark of Hackbench, a test of the Linux kernel scheduler. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 1 - Type: ProcessLinux 5.9.1Linux 5.10 Git 23 Oct1.07422.14843.22264.29685.371SE +/- 0.005, N = 3SE +/- 0.016, N = 34.4914.7741. (CC) gcc options: -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 1 - Type: ProcessLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 4.48 / Avg: 4.49 / Max: 4.5Min: 4.75 / Avg: 4.77 / Max: 4.81. (CC) gcc options: -lpthread

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 1 - Type: ThreadLinux 5.9.1Linux 5.10 Git 23 Oct1.12052.2413.36154.4825.6025SE +/- 0.030, N = 3SE +/- 0.016, N = 34.7144.9801. (CC) gcc options: -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 1 - Type: ThreadLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 4.65 / Avg: 4.71 / Max: 4.75Min: 4.95 / Avg: 4.98 / Max: 5.011. (CC) gcc options: -lpthread

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Socket ActivityLinux 5.9.1Linux 5.10 Git 23 Oct15003000450060007500SE +/- 18.37, N = 3SE +/- 65.15, N = 36765.816454.541. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Socket ActivityLinux 5.9.1Linux 5.10 Git 23 Oct12002400360048006000Min: 6730.07 / Avg: 6765.81 / Max: 6791.06Min: 6339.6 / Avg: 6454.54 / Max: 6565.161. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

PostMark

This is a test of NetApp's PostMark benchmark designed to simulate small-file testing similar to the tasks endured by web and mail servers. This test profile will set PostMark to perform 25,000 transactions with 500 files simultaneously with the file sizes ranging between 5 and 512 kilobytes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTPS, More Is BetterPostMark 1.51Disk Transaction PerformanceLinux 5.9.1Linux 5.10 Git 23 Oct2K4K6K8K10KSE +/- 84.00, N = 3789675751. (CC) gcc options: -O3
OpenBenchmarking.orgTPS, More Is BetterPostMark 1.51Disk Transaction PerformanceLinux 5.9.1Linux 5.10 Git 23 Oct14002800420056007000Min: 7812 / Avg: 7896 / Max: 80641. (CC) gcc options: -O3

OSBench

OSBench is a collection of micro-benchmarks for measuring operating system primitives like time to create threads/processes, launching programs, creating files, and memory allocation. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgus Per Event, Fewer Is BetterOSBenchTest: Create ThreadsLinux 5.9.1Linux 5.10 Git 23 Oct3691215SE +/- 0.091455, N = 10SE +/- 0.069936, N = 39.7699179.3801821. (CC) gcc options: -lm
OpenBenchmarking.orgus Per Event, Fewer Is BetterOSBenchTest: Create ThreadsLinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 9.37 / Avg: 9.77 / Max: 10.17Min: 9.31 / Avg: 9.38 / Max: 9.521. (CC) gcc options: -lm

Apache Siege

This is a test of the Apache web server performance being facilitated by the Siege web serverb enchmark program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 50Linux 5.9.1Linux 5.10 Git 23 Oct20K40K60K80K100KSE +/- 812.43, N = 3SE +/- 629.92, N = 388406.2384898.971. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto
OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 50Linux 5.9.1Linux 5.10 Git 23 Oct15K30K45K60K75KMin: 86956.52 / Avg: 88406.23 / Max: 89766.6Min: 83892.62 / Avg: 84898.97 / Max: 86058.521. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: mobilenetLinux 5.9.1Linux 5.10 Git 23 Oct3691215SE +/- 0.04, N = 3SE +/- 0.02, N = 311.3011.74MIN: 10.53 / MAX: 24.46MIN: 10.99 / MAX: 19.761. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: mobilenetLinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 11.22 / Avg: 11.3 / Max: 11.34Min: 11.7 / Avg: 11.74 / Max: 11.771. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

LeelaChessZero

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

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.26Backend: EigenLinux 5.9.1Linux 5.10 Git 23 Oct130260390520650SE +/- 6.46, N = 9SE +/- 7.80, N = 35936161. (CXX) g++ options: -flto -pthread
OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.26Backend: EigenLinux 5.9.1Linux 5.10 Git 23 Oct110220330440550Min: 570 / Avg: 593.11 / Max: 627Min: 603 / Avg: 616.33 / Max: 6301. (CXX) g++ options: -flto -pthread

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: ForkingLinux 5.9.1Linux 5.10 Git 23 Oct11K22K33K44K55KSE +/- 214.03, N = 3SE +/- 293.91, N = 351191.5649488.991. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: ForkingLinux 5.9.1Linux 5.10 Git 23 Oct9K18K27K36K45KMin: 50956.59 / Avg: 51191.56 / Max: 51618.91Min: 49060.11 / Avg: 49488.99 / Max: 50051.551. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

Apache Siege

This is a test of the Apache web server performance being facilitated by the Siege web serverb enchmark program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 10Linux 5.9.1Linux 5.10 Git 23 Oct7K14K21K28K35KSE +/- 361.59, N = 3SE +/- 377.38, N = 634254.1533189.871. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto
OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 10Linux 5.9.1Linux 5.10 Git 23 Oct6K12K18K24K30KMin: 33783.79 / Avg: 34254.15 / Max: 34965.04Min: 31347.96 / Avg: 33189.87 / Max: 33783.791. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: MMAPLinux 5.9.1Linux 5.10 Git 23 Oct4080120160200SE +/- 0.28, N = 3SE +/- 0.11, N = 3184.29189.961. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: MMAPLinux 5.9.1Linux 5.10 Git 23 Oct306090120150Min: 183.74 / Avg: 184.29 / Max: 184.65Min: 189.75 / Avg: 189.96 / Max: 190.081. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

OSBench

OSBench is a collection of micro-benchmarks for measuring operating system primitives like time to create threads/processes, launching programs, creating files, and memory allocation. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgus Per Event, Fewer Is BetterOSBenchTest: Create FilesLinux 5.9.1Linux 5.10 Git 23 Oct3691215SE +/- 0.03, N = 3SE +/- 0.01, N = 312.0612.401. (CC) gcc options: -lm
OpenBenchmarking.orgus Per Event, Fewer Is BetterOSBenchTest: Create FilesLinux 5.9.1Linux 5.10 Git 23 Oct48121620Min: 12.04 / Avg: 12.06 / Max: 12.11Min: 12.39 / Avg: 12.4 / Max: 12.421. (CC) gcc options: -lm

GPAW

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

OpenBenchmarking.orgSeconds, Fewer Is BetterGPAW 20.1Input: Carbon NanotubeLinux 5.9.1Linux 5.10 Git 23 Oct80160240320400SE +/- 0.31, N = 3SE +/- 0.69, N = 3379.35389.711. (CC) gcc options: -pthread -shared -fwrapv -O2 -lxc -lblas -lmpi
OpenBenchmarking.orgSeconds, Fewer Is BetterGPAW 20.1Input: Carbon NanotubeLinux 5.9.1Linux 5.10 Git 23 Oct70140210280350Min: 379.02 / Avg: 379.35 / Max: 379.96Min: 388.42 / Avg: 389.71 / Max: 390.781. (CC) gcc options: -pthread -shared -fwrapv -O2 -lxc -lblas -lmpi

Selenium

This test profile uses the Selenium WebDriver for running various browser benchmarks in different available web browsers. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterSeleniumBenchmark: CanvasMark - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct3K6K9K12K15KSE +/- 80.60, N = 3SE +/- 139.74, N = 314235138581. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0
OpenBenchmarking.orgScore, More Is BetterSeleniumBenchmark: CanvasMark - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct2K4K6K8K10KMin: 14140 / Avg: 14234.67 / Max: 14395Min: 13664 / Avg: 13857.67 / Max: 141291. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Glibc C String FunctionsLinux 5.9.1Linux 5.10 Git 23 Oct130K260K390K520K650KSE +/- 6366.90, N = 3SE +/- 9660.65, N = 3583277.59598962.951. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Glibc C String FunctionsLinux 5.9.1Linux 5.10 Git 23 Oct100K200K300K400K500KMin: 570952.95 / Avg: 583277.59 / Max: 592212.94Min: 579642.3 / Avg: 598962.95 / Max: 608760.551. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: CPU CacheLinux 5.9.1Linux 5.10 Git 23 Oct510152025SE +/- 0.26, N = 15SE +/- 0.24, N = 1521.3720.831. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: CPU CacheLinux 5.9.1Linux 5.10 Git 23 Oct510152025Min: 20.03 / Avg: 21.37 / Max: 23.83Min: 19.4 / Avg: 20.83 / Max: 22.51. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

Apache Siege

This is a test of the Apache web server performance being facilitated by the Siege web serverb enchmark program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 250Linux 5.9.1Linux 5.10 Git 23 Oct16K32K48K64K80KSE +/- 56.33, N = 3SE +/- 693.79, N = 376561.9474632.411. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto
OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 250Linux 5.9.1Linux 5.10 Git 23 Oct13K26K39K52K65KMin: 76452.59 / Avg: 76561.94 / Max: 76640.1Min: 73313.79 / Avg: 74632.41 / Max: 75665.861. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto

Selenium

This test profile uses the Selenium WebDriver for running various browser benchmarks in different available web browsers. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterSeleniumBenchmark: ARES-6 - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct918273645SE +/- 0.08, N = 3SE +/- 0.15, N = 340.0539.061. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0
OpenBenchmarking.orgms, Fewer Is BetterSeleniumBenchmark: ARES-6 - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct816243240Min: 39.96 / Avg: 40.05 / Max: 40.2Min: 38.76 / Avg: 39.06 / Max: 39.231. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0

WebP Image Encode

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

OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, LosslessLinux 5.9.1Linux 5.10 Git 23 Oct48121620SE +/- 0.03, N = 3SE +/- 0.06, N = 315.4315.051. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff
OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, LosslessLinux 5.9.1Linux 5.10 Git 23 Oct48121620Min: 15.38 / Avg: 15.43 / Max: 15.49Min: 14.98 / Avg: 15.05 / Max: 15.171. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff

Selenium

This test profile uses the Selenium WebDriver for running various browser benchmarks in different available web browsers. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOverall Score, More Is BetterSeleniumBenchmark: Basemark - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct2004006008001000SE +/- 13.26, N = 3SE +/- 12.88, N = 3906.01884.841. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0
OpenBenchmarking.orgOverall Score, More Is BetterSeleniumBenchmark: Basemark - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct160320480640800Min: 890.94 / Avg: 906.01 / Max: 932.45Min: 865.68 / Avg: 884.84 / Max: 909.331. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: MEMFDLinux 5.9.1Linux 5.10 Git 23 Oct130260390520650SE +/- 1.72, N = 3SE +/- 2.11, N = 3589.68602.891. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: MEMFDLinux 5.9.1Linux 5.10 Git 23 Oct110220330440550Min: 586.83 / Avg: 589.68 / Max: 592.76Min: 599.92 / Avg: 602.89 / Max: 606.971. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

LibRaw

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

OpenBenchmarking.orgMpix/sec, More Is BetterLibRaw 0.20Post-Processing BenchmarkLinux 5.9.1Linux 5.10 Git 23 Oct918273645SE +/- 0.40, N = 3SE +/- 0.28, N = 340.1639.301. (CXX) g++ options: -O2 -fopenmp -ljpeg -lz -lm
OpenBenchmarking.orgMpix/sec, More Is BetterLibRaw 0.20Post-Processing BenchmarkLinux 5.9.1Linux 5.10 Git 23 Oct816243240Min: 39.38 / Avg: 40.16 / Max: 40.66Min: 38.81 / Avg: 39.3 / Max: 39.781. (CXX) g++ options: -O2 -fopenmp -ljpeg -lz -lm

Waifu2x-NCNN Vulkan

Waifu2x-NCNN is an NCNN neural network implementation of the Waifu2x converter project and accelerated using the Vulkan API. NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. This test profile times how long it takes to increase the resolution of a sample image with Vulkan. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterWaifu2x-NCNN Vulkan 20200818Scale: 2x - Denoise: 3 - TAA: NoLinux 5.9.1Linux 5.10 Git 23 Oct0.69731.39462.09192.78923.4865SE +/- 0.006, N = 3SE +/- 0.050, N = 33.0353.099
OpenBenchmarking.orgSeconds, Fewer Is BetterWaifu2x-NCNN Vulkan 20200818Scale: 2x - Denoise: 3 - TAA: NoLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 3.03 / Avg: 3.03 / Max: 3.05Min: 3.04 / Avg: 3.1 / Max: 3.2

WebP Image Encode

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

OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Highest CompressionLinux 5.9.1Linux 5.10 Git 23 Oct246810SE +/- 0.081, N = 3SE +/- 0.010, N = 36.7236.8511. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff
OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Highest CompressionLinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 6.6 / Avg: 6.72 / Max: 6.88Min: 6.83 / Avg: 6.85 / Max: 6.861. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff

KeyDB

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

OpenBenchmarking.orgOps/sec, More Is BetterKeyDB 6.0.16Linux 5.9.1Linux 5.10 Git 23 Oct130K260K390K520K650KSE +/- 1265.44, N = 3SE +/- 679.54, N = 3623244.73612175.881. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
OpenBenchmarking.orgOps/sec, More Is BetterKeyDB 6.0.16Linux 5.9.1Linux 5.10 Git 23 Oct110K220K330K440K550KMin: 620755.17 / Avg: 623244.73 / Max: 624883.92Min: 611286.12 / Avg: 612175.88 / Max: 613510.471. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

OpenVINO

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

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Age Gender Recognition Retail 0013 FP32 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct0.13050.2610.39150.5220.6525SE +/- 0.00, N = 6SE +/- 0.01, N = 30.570.58
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Age Gender Recognition Retail 0013 FP32 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 0.57 / Avg: 0.57 / Max: 0.59Min: 0.57 / Avg: 0.58 / Max: 0.59

Sockperf

This is a network socket API performance benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgusec, Fewer Is BetterSockperf 3.4Test: Latency Ping PongLinux 5.9.1Linux 5.10 Git 23 Oct0.81591.63182.44773.26364.0795SE +/- 0.023, N = 21SE +/- 0.038, N = 83.5663.6261. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread
OpenBenchmarking.orgusec, Fewer Is BetterSockperf 3.4Test: Latency Ping PongLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 3.31 / Avg: 3.57 / Max: 3.85Min: 3.43 / Avg: 3.63 / Max: 3.741. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: yolov4-tinyLinux 5.9.1Linux 5.10 Git 23 Oct510152025SE +/- 0.06, N = 3SE +/- 0.12, N = 318.7619.06MIN: 16.96 / MAX: 30.81MIN: 17.4 / MAX: 28.481. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: yolov4-tinyLinux 5.9.1Linux 5.10 Git 23 Oct510152025Min: 18.66 / Avg: 18.76 / Max: 18.87Min: 18.82 / Avg: 19.06 / Max: 19.21. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: SENDFILELinux 5.9.1Linux 5.10 Git 23 Oct30K60K90K120K150KSE +/- 527.85, N = 3SE +/- 137.45, N = 3151380.14148999.831. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: SENDFILELinux 5.9.1Linux 5.10 Git 23 Oct30K60K90K120K150KMin: 150639.94 / Avg: 151380.14 / Max: 152402.12Min: 148848.19 / Avg: 148999.83 / Max: 149274.221. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

LevelDB

LevelDB is a key-value storage library developed by Google that supports making use of Snappy for data compression and has other modern features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Hot ReadLinux 5.9.1Linux 5.10 Git 23 Oct246810SE +/- 0.050, N = 3SE +/- 0.001, N = 37.3507.4671. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Hot ReadLinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 7.3 / Avg: 7.35 / Max: 7.45Min: 7.47 / Avg: 7.47 / Max: 7.471. (CXX) g++ options: -O3 -lsnappy -lpthread

Apache Siege

This is a test of the Apache web server performance being facilitated by the Siege web serverb enchmark program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 100Linux 5.9.1Linux 5.10 Git 23 Oct20K40K60K80K100KSE +/- 475.78, N = 3SE +/- 1178.45, N = 392711.9491299.421. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto
OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 100Linux 5.9.1Linux 5.10 Git 23 Oct16K32K48K64K80KMin: 91827.36 / Avg: 92711.94 / Max: 93457.95Min: 88967.98 / Avg: 91299.42 / Max: 92764.381. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto

Dolfyn

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

OpenBenchmarking.orgSeconds, Fewer Is BetterDolfyn 0.527Computational Fluid DynamicsLinux 5.9.1Linux 5.10 Git 23 Oct48121620SE +/- 0.12, N = 3SE +/- 0.19, N = 315.4215.65
OpenBenchmarking.orgSeconds, Fewer Is BetterDolfyn 0.527Computational Fluid DynamicsLinux 5.9.1Linux 5.10 Git 23 Oct48121620Min: 15.23 / Avg: 15.42 / Max: 15.64Min: 15.26 / Avg: 15.65 / Max: 15.86

Apache CouchDB

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

OpenBenchmarking.orgSeconds, Fewer Is BetterApache CouchDB 3.1.1Bulk Size: 100 - Inserts: 1000 - Rounds: 24Linux 5.9.1Linux 5.10 Git 23 Oct20406080100SE +/- 1.46, N = 3SE +/- 0.65, N = 394.6793.241. (CXX) g++ options: -std=c++14 -lmozjs-68 -lm -lerl_interface -lei -fPIC -MMD
OpenBenchmarking.orgSeconds, Fewer Is BetterApache CouchDB 3.1.1Bulk Size: 100 - Inserts: 1000 - Rounds: 24Linux 5.9.1Linux 5.10 Git 23 Oct20406080100Min: 91.89 / Avg: 94.67 / Max: 96.83Min: 91.98 / Avg: 93.24 / Max: 94.151. (CXX) g++ options: -std=c++14 -lmozjs-68 -lm -lerl_interface -lei -fPIC -MMD

LevelDB

LevelDB is a key-value storage library developed by Google that supports making use of Snappy for data compression and has other modern features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Random ReadLinux 5.9.1Linux 5.10 Git 23 Oct246810SE +/- 0.013, N = 3SE +/- 0.022, N = 37.3167.4281. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Random ReadLinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 7.29 / Avg: 7.32 / Max: 7.34Min: 7.39 / Avg: 7.43 / Max: 7.471. (CXX) g++ options: -O3 -lsnappy -lpthread

Selenium

This test profile uses the Selenium WebDriver for running various browser benchmarks in different available web browsers. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterSeleniumBenchmark: Kraken - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct160320480640800SE +/- 1.63, N = 3SE +/- 6.38, N = 3743.2732.01. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0
OpenBenchmarking.orgms, Fewer Is BetterSeleniumBenchmark: Kraken - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct130260390520650Min: 741.5 / Avg: 743.23 / Max: 746.5Min: 723.2 / Avg: 732 / Max: 744.41. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0

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 0.26Backend: BLASLinux 5.9.1Linux 5.10 Git 23 Oct130260390520650SE +/- 8.57, N = 3SE +/- 6.57, N = 36006091. (CXX) g++ options: -flto -pthread
OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.26Backend: BLASLinux 5.9.1Linux 5.10 Git 23 Oct110220330440550Min: 584 / Avg: 600.33 / Max: 613Min: 600 / Avg: 609.33 / Max: 6221. (CXX) g++ options: -flto -pthread

eSpeak-NG Speech Engine

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

OpenBenchmarking.orgSeconds, Fewer Is BettereSpeak-NG Speech Engine 20200907Text-To-Speech SynthesisLinux 5.9.1Linux 5.10 Git 23 Oct714212835SE +/- 0.02, N = 4SE +/- 0.20, N = 428.2527.841. (CC) gcc options: -O2 -std=c99
OpenBenchmarking.orgSeconds, Fewer Is BettereSpeak-NG Speech Engine 20200907Text-To-Speech SynthesisLinux 5.9.1Linux 5.10 Git 23 Oct612182430Min: 28.22 / Avg: 28.25 / Max: 28.29Min: 27.41 / Avg: 27.84 / Max: 28.341. (CC) gcc options: -O2 -std=c99

Apache Siege

This is a test of the Apache web server performance being facilitated by the Siege web serverb enchmark program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 200Linux 5.9.1Linux 5.10 Git 23 Oct20K40K60K80K100KSE +/- 569.47, N = 3SE +/- 833.80, N = 382538.8681351.091. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto
OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 200Linux 5.9.1Linux 5.10 Git 23 Oct14K28K42K56K70KMin: 81400.09 / Avg: 82538.86 / Max: 83125.52Min: 79744.82 / Avg: 81351.09 / Max: 82542.31. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto

Waifu2x-NCNN Vulkan

Waifu2x-NCNN is an NCNN neural network implementation of the Waifu2x converter project and accelerated using the Vulkan API. NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. This test profile times how long it takes to increase the resolution of a sample image with Vulkan. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterWaifu2x-NCNN Vulkan 20200818Scale: 2x - Denoise: 3 - TAA: YesLinux 5.9.1Linux 5.10 Git 23 Oct510152025SE +/- 0.07, N = 3SE +/- 0.04, N = 319.9420.24
OpenBenchmarking.orgSeconds, Fewer Is BetterWaifu2x-NCNN Vulkan 20200818Scale: 2x - Denoise: 3 - TAA: YesLinux 5.9.1Linux 5.10 Git 23 Oct510152025Min: 19.87 / Avg: 19.94 / Max: 20.09Min: 20.2 / Avg: 20.23 / Max: 20.31

WebP Image Encode

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

OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100Linux 5.9.1Linux 5.10 Git 23 Oct0.48760.97521.46281.95042.438SE +/- 0.029, N = 3SE +/- 0.028, N = 32.1362.1671. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff
OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100Linux 5.9.1Linux 5.10 Git 23 Oct246810Min: 2.11 / Avg: 2.14 / Max: 2.19Min: 2.11 / Avg: 2.17 / Max: 2.21. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Context SwitchingLinux 5.9.1Linux 5.10 Git 23 Oct900K1800K2700K3600K4500KSE +/- 51803.89, N = 3SE +/- 18567.71, N = 34257616.044198734.921. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Context SwitchingLinux 5.9.1Linux 5.10 Git 23 Oct700K1400K2100K2800K3500KMin: 4156986.87 / Avg: 4257616.04 / Max: 4329290.8Min: 4178846.21 / Avg: 4198734.92 / Max: 4235838.241. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

TNN

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

OpenBenchmarking.orgms, Fewer Is BetterTNN 0.2.3Target: CPU - Model: SqueezeNet v1.1Linux 5.9.1Linux 5.10 Git 23 Oct50100150200250SE +/- 0.40, N = 3SE +/- 1.91, N = 3237.72234.50MIN: 234.13 / MAX: 246.23MIN: 227.72 / MAX: 238.751. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -O3 -rdynamic -ldl
OpenBenchmarking.orgms, Fewer Is BetterTNN 0.2.3Target: CPU - Model: SqueezeNet v1.1Linux 5.9.1Linux 5.10 Git 23 Oct4080120160200Min: 237.1 / Avg: 237.72 / Max: 238.48Min: 230.69 / Avg: 234.49 / Max: 236.641. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -O3 -rdynamic -ldl

VkFFT

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

OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 2020-09-29Linux 5.9.1Linux 5.10 Git 23 Oct2K4K6K8K10KSE +/- 5.21, N = 3SE +/- 7.67, N = 31085110706
OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 2020-09-29Linux 5.9.1Linux 5.10 Git 23 Oct2K4K6K8K10KMin: 10842 / Avg: 10850.67 / Max: 10860Min: 10691 / Avg: 10706.33 / Max: 10714

OSBench

OSBench is a collection of micro-benchmarks for measuring operating system primitives like time to create threads/processes, launching programs, creating files, and memory allocation. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgus Per Event, Fewer Is BetterOSBenchTest: Launch ProgramsLinux 5.9.1Linux 5.10 Git 23 Oct918273645SE +/- 0.30, N = 3SE +/- 0.50, N = 338.1838.691. (CC) gcc options: -lm
OpenBenchmarking.orgus Per Event, Fewer Is BetterOSBenchTest: Launch ProgramsLinux 5.9.1Linux 5.10 Git 23 Oct816243240Min: 37.84 / Avg: 38.18 / Max: 38.79Min: 37.71 / Avg: 38.69 / Max: 39.331. (CC) gcc options: -lm

Zstd Compression

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

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.5Compression Level: 19Linux 5.9.1Linux 5.10 Git 23 Oct714212835SE +/- 0.00, N = 3SE +/- 0.10, N = 330.830.41. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.5Compression Level: 19Linux 5.9.1Linux 5.10 Git 23 Oct714212835Min: 30.8 / Avg: 30.8 / Max: 30.8Min: 30.2 / Avg: 30.4 / Max: 30.51. (CC) gcc options: -O3 -pthread -lz -llzma

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: MallocLinux 5.9.1Linux 5.10 Git 23 Oct20M40M60M80M100MSE +/- 81223.46, N = 3SE +/- 53241.35, N = 383874667.8382826646.861. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: MallocLinux 5.9.1Linux 5.10 Git 23 Oct15M30M45M60M75MMin: 83713115.29 / Avg: 83874667.83 / Max: 83970186.38Min: 82734082.36 / Avg: 82826646.86 / Max: 82918511.821. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: vgg16Linux 5.9.1Linux 5.10 Git 23 Oct1224364860SE +/- 0.11, N = 3SE +/- 0.01, N = 353.6454.30MIN: 53.18 / MAX: 62.42MIN: 53.75 / MAX: 58.981. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: vgg16Linux 5.9.1Linux 5.10 Git 23 Oct1122334455Min: 53.5 / Avg: 53.64 / Max: 53.85Min: 54.29 / Avg: 54.3 / Max: 54.311. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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 4KLinux 5.9.1Linux 5.10 Git 23 Oct3691215SE +/- 0.12, N = 8SE +/- 0.16, N = 311.5411.401. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 4KLinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 10.88 / Avg: 11.54 / Max: 11.92Min: 11.14 / Avg: 11.4 / Max: 11.71. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

RealSR-NCNN

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

OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: NoLinux 5.9.1Linux 5.10 Git 23 Oct1122334455SE +/- 0.06, N = 3SE +/- 0.07, N = 347.5348.11
OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: NoLinux 5.9.1Linux 5.10 Git 23 Oct1020304050Min: 47.44 / Avg: 47.53 / Max: 47.64Min: 47.98 / Avg: 48.11 / Max: 48.21

Selenium

This test profile uses the Selenium WebDriver for running various browser benchmarks in different available web browsers. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterSeleniumBenchmark: Jetstream - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct50100150200250SE +/- 0.59, N = 3SE +/- 0.48, N = 3228.28225.691. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0
OpenBenchmarking.orgScore, More Is BetterSeleniumBenchmark: Jetstream - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct4080120160200Min: 227.58 / Avg: 228.28 / Max: 229.46Min: 224.95 / Avg: 225.69 / Max: 226.61. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0

RealSR-NCNN

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

OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: YesLinux 5.9.1Linux 5.10 Git 23 Oct80160240320400SE +/- 0.20, N = 3SE +/- 0.04, N = 3363.35367.42
OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: YesLinux 5.9.1Linux 5.10 Git 23 Oct70140210280350Min: 363.14 / Avg: 363.35 / Max: 363.75Min: 367.37 / Avg: 367.42 / Max: 367.5

dav1d

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

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Chimera 1080p 10-bitLinux 5.9.1Linux 5.10 Git 23 Oct20406080100SE +/- 0.09, N = 3SE +/- 0.21, N = 3102.41101.36MIN: 67.85 / MAX: 233.27MIN: 67.25 / MAX: 225.761. (CC) gcc options: -pthread
OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Chimera 1080p 10-bitLinux 5.9.1Linux 5.10 Git 23 Oct20406080100Min: 102.27 / Avg: 102.41 / Max: 102.57Min: 100.97 / Avg: 101.36 / Max: 101.71. (CC) gcc options: -pthread

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 MobileLinux 5.9.1Linux 5.10 Git 23 Oct30K60K90K120K150KSE +/- 597.97, N = 3SE +/- 1913.90, N = 3160650162185
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: NASNet MobileLinux 5.9.1Linux 5.10 Git 23 Oct30K60K90K120K150KMin: 159908 / Avg: 160649.67 / Max: 161833Min: 160113 / Avg: 162184.67 / Max: 166008

Kvazaar

This is a test of Kvazaar as a CPU-based H.265 video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 1080p - Video Preset: Very FastLinux 5.9.1Linux 5.10 Git 23 Oct1326395265SE +/- 0.18, N = 3SE +/- 0.12, N = 359.4158.891. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 1080p - Video Preset: Very FastLinux 5.9.1Linux 5.10 Git 23 Oct1224364860Min: 59.1 / Avg: 59.41 / Max: 59.72Min: 58.71 / Avg: 58.89 / Max: 59.131. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

SVT-AV1

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

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8Encoder Mode: Enc Mode 0 - Input: 1080pLinux 5.9.1Linux 5.10 Git 23 Oct0.02590.05180.07770.10360.1295SE +/- 0.000, N = 3SE +/- 0.000, N = 30.1150.1141. (CXX) g++ options: -O3 -fcommon -fPIE -fPIC -pie
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8Encoder Mode: Enc Mode 0 - Input: 1080pLinux 5.9.1Linux 5.10 Git 23 Oct12345Min: 0.12 / Avg: 0.12 / Max: 0.12Min: 0.11 / Avg: 0.11 / Max: 0.111. (CXX) g++ options: -O3 -fcommon -fPIE -fPIC -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 0.11.07Test: SemaphoresLinux 5.9.1Linux 5.10 Git 23 Oct200K400K600K800K1000KSE +/- 4093.36, N = 3SE +/- 248.14, N = 31161533.531151702.001. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: SemaphoresLinux 5.9.1Linux 5.10 Git 23 Oct200K400K600K800K1000KMin: 1153657.91 / Avg: 1161533.53 / Max: 1167407.28Min: 1151207.72 / Avg: 1151702 / Max: 1151987.681. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: NUMALinux 5.9.1Linux 5.10 Git 23 Oct50100150200250SE +/- 1.31, N = 3SE +/- 0.91, N = 3219.61217.831. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: NUMALinux 5.9.1Linux 5.10 Git 23 Oct4080120160200Min: 217.02 / Avg: 219.61 / Max: 221.25Min: 216.82 / Avg: 217.83 / Max: 219.651. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

Kvazaar

This is a test of Kvazaar as a CPU-based H.265 video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 1080p - Video Preset: Ultra FastLinux 5.9.1Linux 5.10 Git 23 Oct20406080100SE +/- 0.08, N = 3SE +/- 0.28, N = 3110.29109.421. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 1080p - Video Preset: Ultra FastLinux 5.9.1Linux 5.10 Git 23 Oct20406080100Min: 110.16 / Avg: 110.29 / Max: 110.44Min: 109.07 / Avg: 109.42 / Max: 109.981. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -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 20200916Target: CPU - Model: resnet18Linux 5.9.1Linux 5.10 Git 23 Oct510152025SE +/- 0.06, N = 3SE +/- 0.03, N = 319.1519.00MIN: 18.92 / MAX: 34.93MIN: 18.84 / MAX: 19.511. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: resnet18Linux 5.9.1Linux 5.10 Git 23 Oct510152025Min: 19.03 / Avg: 19.15 / Max: 19.24Min: 18.96 / Avg: 19 / Max: 19.051. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

TNN

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

OpenBenchmarking.orgms, Fewer Is BetterTNN 0.2.3Target: CPU - Model: MobileNet v2Linux 5.9.1Linux 5.10 Git 23 Oct50100150200250SE +/- 2.08, N = 3SE +/- 2.46, N = 3247.26245.44MIN: 241.93 / MAX: 271.83MIN: 237.89 / MAX: 264.71. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -O3 -rdynamic -ldl
OpenBenchmarking.orgms, Fewer Is BetterTNN 0.2.3Target: CPU - Model: MobileNet v2Linux 5.9.1Linux 5.10 Git 23 Oct4080120160200Min: 244.27 / Avg: 247.26 / Max: 251.27Min: 240.52 / Avg: 245.44 / Max: 248.021. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -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 20200916Target: CPU-v3-v3 - Model: mobilenet-v3Linux 5.9.1Linux 5.10 Git 23 Oct1.262.523.785.046.3SE +/- 0.02, N = 3SE +/- 0.02, N = 35.605.56MIN: 5.44 / MAX: 17.88MIN: 5.44 / MAX: 7.781. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU-v3-v3 - Model: mobilenet-v3Linux 5.9.1Linux 5.10 Git 23 Oct246810Min: 5.57 / Avg: 5.6 / Max: 5.63Min: 5.53 / Avg: 5.56 / Max: 5.581. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Zstd Compression

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

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.5Compression Level: 3Linux 5.9.1Linux 5.10 Git 23 Oct8001600240032004000SE +/- 8.41, N = 3SE +/- 6.89, N = 33749.43723.21. (CC) gcc options: -O3 -pthread -lz -llzma
OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.4.5Compression Level: 3Linux 5.9.1Linux 5.10 Git 23 Oct7001400210028003500Min: 3732.8 / Avg: 3749.4 / Max: 3760Min: 3712.2 / Avg: 3723.23 / Max: 3735.91. (CC) gcc options: -O3 -pthread -lz -llzma

OpenVINO

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

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Person Detection 0106 FP32 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct0.33530.67061.00591.34121.6765SE +/- 0.00, N = 3SE +/- 0.00, N = 31.491.48
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Person Detection 0106 FP32 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 1.48 / Avg: 1.49 / Max: 1.49Min: 1.48 / Avg: 1.48 / Max: 1.49

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: vgg16Linux 5.9.1Linux 5.10 Git 23 Oct20406080100SE +/- 0.09, N = 3SE +/- 0.15, N = 376.3075.80MIN: 75.89 / MAX: 81.79MIN: 75.24 / MAX: 104.631. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: vgg16Linux 5.9.1Linux 5.10 Git 23 Oct1530456075Min: 76.12 / Avg: 76.3 / Max: 76.39Min: 75.49 / Avg: 75.8 / Max: 75.981. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: efficientnet-b0Linux 5.9.1Linux 5.10 Git 23 Oct246810SE +/- 0.02, N = 3SE +/- 0.02, N = 37.967.91MIN: 7.88 / MAX: 8.79MIN: 7.81 / MAX: 8.691. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: efficientnet-b0Linux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 7.94 / Avg: 7.96 / Max: 7.99Min: 7.88 / Avg: 7.91 / Max: 7.951. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

LevelDB

LevelDB is a key-value storage library developed by Google that supports making use of Snappy for data compression and has other modern features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterLevelDB 1.22Benchmark: OverwriteLinux 5.9.1Linux 5.10 Git 23 Oct816243240SE +/- 0.12, N = 3SE +/- 0.06, N = 332.432.21. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMB/s, More Is BetterLevelDB 1.22Benchmark: OverwriteLinux 5.9.1Linux 5.10 Git 23 Oct714212835Min: 32.2 / Avg: 32.37 / Max: 32.6Min: 32.1 / Avg: 32.2 / Max: 32.31. (CXX) g++ options: -O3 -lsnappy -lpthread

SVT-AV1

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

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8Encoder Mode: Enc Mode 4 - Input: 1080pLinux 5.9.1Linux 5.10 Git 23 Oct0.73041.46082.19122.92163.652SE +/- 0.007, N = 3SE +/- 0.002, N = 33.2463.2261. (CXX) g++ options: -O3 -fcommon -fPIE -fPIC -pie
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8Encoder Mode: Enc Mode 4 - Input: 1080pLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 3.24 / Avg: 3.25 / Max: 3.26Min: 3.22 / Avg: 3.23 / Max: 3.231. (CXX) g++ options: -O3 -fcommon -fPIE -fPIC -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 20200916Target: CPU - Model: mnasnetLinux 5.9.1Linux 5.10 Git 23 Oct1.14532.29063.43594.58125.7265SE +/- 0.03, N = 3SE +/- 0.01, N = 35.095.06MIN: 4.99 / MAX: 6.24MIN: 4.97 / MAX: 5.971. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: mnasnetLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 5.05 / Avg: 5.09 / Max: 5.15Min: 5.03 / Avg: 5.06 / Max: 5.071. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OSBench

OSBench is a collection of micro-benchmarks for measuring operating system primitives like time to create threads/processes, launching programs, creating files, and memory allocation. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNs Per Event, Fewer Is BetterOSBenchTest: Memory AllocationsLinux 5.9.1Linux 5.10 Git 23 Oct20406080100SE +/- 0.19, N = 3SE +/- 0.05, N = 375.0874.651. (CC) gcc options: -lm
OpenBenchmarking.orgNs Per Event, Fewer Is BetterOSBenchTest: Memory AllocationsLinux 5.9.1Linux 5.10 Git 23 Oct1428425670Min: 74.8 / Avg: 75.08 / Max: 75.46Min: 74.61 / Avg: 74.65 / Max: 74.751. (CC) gcc options: -lm

SVT-AV1

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

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8Encoder Mode: Enc Mode 8 - Input: 1080pLinux 5.9.1Linux 5.10 Git 23 Oct612182430SE +/- 0.08, N = 3SE +/- 0.02, N = 327.2827.131. (CXX) g++ options: -O3 -fcommon -fPIE -fPIC -pie
OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8Encoder Mode: Enc Mode 8 - Input: 1080pLinux 5.9.1Linux 5.10 Git 23 Oct612182430Min: 27.16 / Avg: 27.28 / Max: 27.43Min: 27.1 / Avg: 27.13 / Max: 27.151. (CXX) g++ options: -O3 -fcommon -fPIE -fPIC -pie

OpenVINO

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

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Face Detection 0106 FP16 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct0.4230.8461.2691.6922.115SE +/- 0.01, N = 3SE +/- 0.00, N = 31.871.88
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Face Detection 0106 FP16 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 1.86 / Avg: 1.87 / Max: 1.88Min: 1.87 / Avg: 1.88 / Max: 1.88

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: blazefaceLinux 5.9.1Linux 5.10 Git 23 Oct0.42530.85061.27591.70122.1265SE +/- 0.00, N = 3SE +/- 0.00, N = 31.891.88MIN: 1.84 / MAX: 2.24MIN: 1.83 / MAX: 2.051. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: blazefaceLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 1.88 / Avg: 1.89 / Max: 1.89Min: 1.88 / Avg: 1.88 / Max: 1.891. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Kvazaar

This is a test of Kvazaar as a CPU-based H.265 video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 4K - Video Preset: SlowLinux 5.9.1Linux 5.10 Git 23 Oct1.30282.60563.90845.21126.514SE +/- 0.01, N = 3SE +/- 0.01, N = 35.795.761. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 4K - Video Preset: SlowLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 5.77 / Avg: 5.79 / Max: 5.8Min: 5.75 / Avg: 5.76 / Max: 5.771. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

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 1080pLinux 5.9.1Linux 5.10 Git 23 Oct1122334455SE +/- 0.23, N = 3SE +/- 0.22, N = 350.6150.351. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 1080pLinux 5.9.1Linux 5.10 Git 23 Oct1020304050Min: 50.26 / Avg: 50.61 / Max: 51.04Min: 50.01 / Avg: 50.35 / Max: 50.751. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

Selenium

This test profile uses the Selenium WebDriver for running various browser benchmarks in different available web browsers. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGeometric Mean, More Is BetterSeleniumBenchmark: Octane - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct8K16K24K32K40KSE +/- 255.99, N = 3SE +/- 221.23, N = 337823376421. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0
OpenBenchmarking.orgGeometric Mean, More Is BetterSeleniumBenchmark: Octane - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct7K14K21K28K35KMin: 37391 / Avg: 37823.33 / Max: 38277Min: 37326 / Avg: 37641.67 / Max: 380681. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0

LevelDB

LevelDB is a key-value storage library developed by Google that supports making use of Snappy for data compression and has other modern features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: OverwriteLinux 5.9.1Linux 5.10 Git 23 Oct1224364860SE +/- 0.17, N = 3SE +/- 0.11, N = 354.6554.901. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: OverwriteLinux 5.9.1Linux 5.10 Git 23 Oct1122334455Min: 54.32 / Avg: 54.64 / Max: 54.91Min: 54.71 / Avg: 54.9 / Max: 55.091. (CXX) g++ options: -O3 -lsnappy -lpthread

WebP Image Encode

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

OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: DefaultLinux 5.9.1Linux 5.10 Git 23 Oct0.31590.63180.94771.26361.5795SE +/- 0.015, N = 3SE +/- 0.017, N = 31.3981.4041. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff
OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: DefaultLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 1.38 / Avg: 1.4 / Max: 1.43Min: 1.38 / Avg: 1.4 / Max: 1.441. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff

Timed MAFFT Alignment

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

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed MAFFT Alignment 7.471Multiple Sequence Alignment - LSU RNALinux 5.9.1Linux 5.10 Git 23 Oct3691215SE +/- 0.07, N = 3SE +/- 0.06, N = 311.4911.441. (CC) gcc options: -std=c99 -O3 -lm -lpthread
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed MAFFT Alignment 7.471Multiple Sequence Alignment - LSU RNALinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 11.36 / Avg: 11.49 / Max: 11.6Min: 11.32 / Avg: 11.44 / Max: 11.531. (CC) gcc options: -std=c99 -O3 -lm -lpthread

OpenVINO

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

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Face Detection 0106 FP16 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct5001000150020002500SE +/- 7.43, N = 3SE +/- 4.97, N = 32137.282128.90
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Face Detection 0106 FP16 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct400800120016002000Min: 2122.56 / Avg: 2137.28 / Max: 2146.39Min: 2120.23 / Avg: 2128.9 / Max: 2137.45

Kvazaar

This is a test of Kvazaar as a CPU-based H.265 video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 1080p - Video Preset: SlowLinux 5.9.1Linux 5.10 Git 23 Oct612182430SE +/- 0.05, N = 3SE +/- 0.08, N = 324.9624.871. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 1080p - Video Preset: SlowLinux 5.9.1Linux 5.10 Git 23 Oct612182430Min: 24.88 / Avg: 24.96 / Max: 25.04Min: 24.71 / Avg: 24.87 / Max: 24.991. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Glibc Qsort Data SortingLinux 5.9.1Linux 5.10 Git 23 Oct306090120150SE +/- 0.73, N = 3SE +/- 0.45, N = 3141.17141.681. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Glibc Qsort Data SortingLinux 5.9.1Linux 5.10 Git 23 Oct306090120150Min: 140.17 / Avg: 141.17 / Max: 142.6Min: 141.13 / Avg: 141.68 / Max: 142.571. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

Kvazaar

This is a test of Kvazaar as a CPU-based H.265 video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 1080p - Video Preset: MediumLinux 5.9.1Linux 5.10 Git 23 Oct612182430SE +/- 0.04, N = 3SE +/- 0.03, N = 325.5925.501. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 1080p - Video Preset: MediumLinux 5.9.1Linux 5.10 Git 23 Oct612182430Min: 25.55 / Avg: 25.59 / Max: 25.66Min: 25.43 / Avg: 25.5 / Max: 25.541. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 4K - Video Preset: MediumLinux 5.9.1Linux 5.10 Git 23 Oct1.32082.64163.96245.28326.604SE +/- 0.00, N = 3SE +/- 0.01, N = 35.875.851. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 4K - Video Preset: MediumLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 5.86 / Avg: 5.87 / Max: 5.87Min: 5.84 / Avg: 5.85 / Max: 5.861. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

LevelDB

LevelDB is a key-value storage library developed by Google that supports making use of Snappy for data compression and has other modern features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Random FillLinux 5.9.1Linux 5.10 Git 23 Oct1224364860SE +/- 0.03, N = 3SE +/- 0.12, N = 354.7854.951. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Random FillLinux 5.9.1Linux 5.10 Git 23 Oct1122334455Min: 54.71 / Avg: 54.78 / Max: 54.82Min: 54.72 / Avg: 54.95 / Max: 55.111. (CXX) g++ options: -O3 -lsnappy -lpthread

Blender

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

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Pabellon Barcelona - Compute: CPU-OnlyLinux 5.9.1Linux 5.10 Git 23 Oct110220330440550SE +/- 0.79, N = 3SE +/- 0.69, N = 3498.42500.00
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Pabellon Barcelona - Compute: CPU-OnlyLinux 5.9.1Linux 5.10 Git 23 Oct90180270360450Min: 497.01 / Avg: 498.42 / Max: 499.73Min: 498.72 / Avg: 500 / Max: 501.08

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: resnet50Linux 5.9.1Linux 5.10 Git 23 Oct714212835SE +/- 0.04, N = 3SE +/- 0.07, N = 332.1032.00MIN: 31.79 / MAX: 56.36MIN: 31.76 / MAX: 33.971. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: resnet50Linux 5.9.1Linux 5.10 Git 23 Oct714212835Min: 32.06 / Avg: 32.1 / Max: 32.18Min: 31.9 / Avg: 32 / Max: 32.141. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

LevelDB

LevelDB is a key-value storage library developed by Google that supports making use of Snappy for data compression and has other modern features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterLevelDB 1.22Benchmark: Random FillLinux 5.9.1Linux 5.10 Git 23 Oct816243240SE +/- 0.00, N = 3SE +/- 0.07, N = 332.332.21. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMB/s, More Is BetterLevelDB 1.22Benchmark: Random FillLinux 5.9.1Linux 5.10 Git 23 Oct714212835Min: 32.3 / Avg: 32.3 / Max: 32.3Min: 32.1 / Avg: 32.17 / Max: 32.31. (CXX) g++ options: -O3 -lsnappy -lpthread

OpenBenchmarking.orgMB/s, More Is BetterLevelDB 1.22Benchmark: Sequential FillLinux 5.9.1Linux 5.10 Git 23 Oct816243240SE +/- 0.07, N = 3SE +/- 0.12, N = 333.433.31. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMB/s, More Is BetterLevelDB 1.22Benchmark: Sequential FillLinux 5.9.1Linux 5.10 Git 23 Oct714212835Min: 33.3 / Avg: 33.37 / Max: 33.5Min: 33.1 / Avg: 33.27 / Max: 33.51. (CXX) g++ options: -O3 -lsnappy -lpthread

Sockperf

This is a network socket API performance benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMessages Per Second, More Is BetterSockperf 3.4Test: ThroughputLinux 5.9.1Linux 5.10 Git 23 Oct150K300K450K600K750KSE +/- 6634.59, N = 5SE +/- 6321.01, N = 56964866985131. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread
OpenBenchmarking.orgMessages Per Second, More Is BetterSockperf 3.4Test: ThroughputLinux 5.9.1Linux 5.10 Git 23 Oct120K240K360K480K600KMin: 681608 / Avg: 696486 / Max: 719383Min: 690061 / Avg: 698513.4 / Max: 7231461. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread

LevelDB

LevelDB is a key-value storage library developed by Google that supports making use of Snappy for data compression and has other modern features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Sequential FillLinux 5.9.1Linux 5.10 Git 23 Oct1224364860SE +/- 0.10, N = 3SE +/- 0.18, N = 353.0353.171. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Sequential FillLinux 5.9.1Linux 5.10 Git 23 Oct1122334455Min: 52.83 / Avg: 53.03 / Max: 53.14Min: 52.82 / Avg: 53.17 / Max: 53.431. (CXX) g++ options: -O3 -lsnappy -lpthread

Blender

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

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Classroom - Compute: CPU-OnlyLinux 5.9.1Linux 5.10 Git 23 Oct100200300400500SE +/- 0.23, N = 3SE +/- 0.57, N = 3447.04448.22
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Classroom - Compute: CPU-OnlyLinux 5.9.1Linux 5.10 Git 23 Oct80160240320400Min: 446.74 / Avg: 447.04 / Max: 447.48Min: 447.42 / Avg: 448.22 / Max: 449.33

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Matrix MathLinux 5.9.1Linux 5.10 Git 23 Oct9K18K27K36K45KSE +/- 216.00, N = 3SE +/- 427.33, N = 340473.6940575.621. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Matrix MathLinux 5.9.1Linux 5.10 Git 23 Oct7K14K21K28K35KMin: 40167.98 / Avg: 40473.69 / Max: 40890.88Min: 39724.4 / Avg: 40575.62 / Max: 41067.551. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

Blender

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

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Barbershop - Compute: CPU-OnlyLinux 5.9.1Linux 5.10 Git 23 Oct140280420560700SE +/- 0.58, N = 3SE +/- 0.55, N = 3648.28646.68
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Barbershop - Compute: CPU-OnlyLinux 5.9.1Linux 5.10 Git 23 Oct110220330440550Min: 647.12 / Avg: 648.28 / Max: 648.89Min: 645.95 / Avg: 646.68 / Max: 647.75

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Fishy Cat - Compute: CPU-OnlyLinux 5.9.1Linux 5.10 Git 23 Oct50100150200250SE +/- 0.47, N = 3SE +/- 0.12, N = 3208.00208.50
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Fishy Cat - Compute: CPU-OnlyLinux 5.9.1Linux 5.10 Git 23 Oct4080120160200Min: 207.51 / Avg: 208 / Max: 208.94Min: 208.26 / Avg: 208.5 / Max: 208.63

OpenVINO

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

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Age Gender Recognition Retail 0013 FP16 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct15003000450060007500SE +/- 15.22, N = 3SE +/- 20.78, N = 36806.096822.19
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Age Gender Recognition Retail 0013 FP16 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct12002400360048006000Min: 6777.96 / Avg: 6806.09 / Max: 6830.23Min: 6781.06 / Avg: 6822.19 / Max: 6847.91

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: shufflenet-v2Linux 5.9.1Linux 5.10 Git 23 Oct1.0622.1243.1864.2485.31SE +/- 0.01, N = 3SE +/- 0.02, N = 34.714.72MIN: 4.6 / MAX: 5.6MIN: 4.65 / MAX: 6.241. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: shufflenet-v2Linux 5.9.1Linux 5.10 Git 23 Oct246810Min: 4.68 / Avg: 4.71 / Max: 4.72Min: 4.69 / Avg: 4.72 / Max: 4.741. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Kvazaar

This is a test of Kvazaar as a CPU-based H.265 video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 4K - Video Preset: Ultra FastLinux 5.9.1Linux 5.10 Git 23 Oct714212835SE +/- 0.11, N = 3SE +/- 0.02, N = 329.1729.111. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 4K - Video Preset: Ultra FastLinux 5.9.1Linux 5.10 Git 23 Oct612182430Min: 28.95 / Avg: 29.17 / Max: 29.29Min: 29.07 / Avg: 29.11 / Max: 29.151. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

Caffe

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

OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: GoogleNet - Acceleration: CPU - Iterations: 100Linux 5.9.1Linux 5.10 Git 23 Oct30K60K90K120K150KSE +/- 351.51, N = 3SE +/- 187.00, N = 31286071288581. (CXX) g++ options: -fPIC -O3 -rdynamic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas
OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: GoogleNet - Acceleration: CPU - Iterations: 100Linux 5.9.1Linux 5.10 Git 23 Oct20K40K60K80K100KMin: 128252 / Avg: 128607 / Max: 129310Min: 128489 / Avg: 128858.33 / Max: 1290941. (CXX) g++ options: -fPIC -O3 -rdynamic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: alexnetLinux 5.9.1Linux 5.10 Git 23 Oct48121620SE +/- 0.01, N = 3SE +/- 0.01, N = 317.2817.25MIN: 17.2 / MAX: 18.6MIN: 17.17 / MAX: 17.411. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: alexnetLinux 5.9.1Linux 5.10 Git 23 Oct48121620Min: 17.26 / Avg: 17.28 / Max: 17.3Min: 17.23 / Avg: 17.25 / Max: 17.281. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

LevelDB

LevelDB is a key-value storage library developed by Google that supports making use of Snappy for data compression and has other modern features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Seek RandomLinux 5.9.1Linux 5.10 Git 23 Oct3691215SE +/- 0.04, N = 3SE +/- 0.03, N = 311.1411.161. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Seek RandomLinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 11.08 / Avg: 11.14 / Max: 11.22Min: 11.13 / Avg: 11.16 / Max: 11.211. (CXX) g++ options: -O3 -lsnappy -lpthread

Blender

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

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: BMW27 - Compute: CPU-OnlyLinux 5.9.1Linux 5.10 Git 23 Oct306090120150SE +/- 0.22, N = 3SE +/- 0.06, N = 3155.68155.43
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: BMW27 - Compute: CPU-OnlyLinux 5.9.1Linux 5.10 Git 23 Oct306090120150Min: 155.32 / Avg: 155.68 / Max: 156.08Min: 155.32 / Avg: 155.43 / Max: 155.52

Caffe

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

OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: AlexNet - Acceleration: CPU - Iterations: 100Linux 5.9.1Linux 5.10 Git 23 Oct11K22K33K44K55KSE +/- 83.67, N = 3SE +/- 94.17, N = 352917528331. (CXX) g++ options: -fPIC -O3 -rdynamic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas
OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: AlexNet - Acceleration: CPU - Iterations: 100Linux 5.9.1Linux 5.10 Git 23 Oct9K18K27K36K45KMin: 52750 / Avg: 52917.33 / Max: 53002Min: 52685 / Avg: 52833.33 / Max: 530081. (CXX) g++ options: -fPIC -O3 -rdynamic -lglog -lgflags -lprotobuf -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: CryptoLinux 5.9.1Linux 5.10 Git 23 Oct5001000150020002500SE +/- 1.73, N = 3SE +/- 3.59, N = 32406.522402.771. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: CryptoLinux 5.9.1Linux 5.10 Git 23 Oct400800120016002000Min: 2404.34 / Avg: 2406.52 / Max: 2409.95Min: 2395.61 / Avg: 2402.77 / Max: 2406.741. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

TensorFlow Lite

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

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet QuantLinux 5.9.1Linux 5.10 Git 23 Oct30K60K90K120K150KSE +/- 94.84, N = 3SE +/- 64.22, N = 3126774126948
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet QuantLinux 5.9.1Linux 5.10 Git 23 Oct20K40K60K80K100KMin: 126586 / Avg: 126774.33 / Max: 126888Min: 126820 / Avg: 126948.33 / Max: 127017

NAMD

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

OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.14ATPase Simulation - 327,506 AtomsLinux 5.9.1Linux 5.10 Git 23 Oct0.49420.98841.48261.97682.471SE +/- 0.00186, N = 3SE +/- 0.00110, N = 32.193582.19643
OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.14ATPase Simulation - 327,506 AtomsLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 2.19 / Avg: 2.19 / Max: 2.2Min: 2.2 / Avg: 2.2 / Max: 2.2

OpenVINO

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

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Face Detection 0106 FP32 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct5001000150020002500SE +/- 3.98, N = 3SE +/- 10.65, N = 32141.332138.56
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Face Detection 0106 FP32 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct400800120016002000Min: 2137.26 / Avg: 2141.33 / Max: 2149.29Min: 2118.65 / Avg: 2138.56 / Max: 2155.08

GROMACS

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

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2020.3Water BenchmarkLinux 5.9.1Linux 5.10 Git 23 Oct0.17440.34880.52320.69760.872SE +/- 0.001, N = 3SE +/- 0.002, N = 30.7750.7741. (CXX) g++ options: -O3 -pthread -lrt -lpthread -lm
OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2020.3Water BenchmarkLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 0.77 / Avg: 0.77 / Max: 0.78Min: 0.77 / Avg: 0.77 / Max: 0.781. (CXX) g++ options: -O3 -pthread -lrt -lpthread -lm

Selenium

This test profile uses the Selenium WebDriver for running various browser benchmarks in different available web browsers. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRuns / Minute, More Is BetterSeleniumBenchmark: StyleBench - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct20406080100SE +/- 0.27, N = 3SE +/- 0.38, N = 385.685.71. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0
OpenBenchmarking.orgRuns / Minute, More Is BetterSeleniumBenchmark: StyleBench - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct1632486480Min: 85.1 / Avg: 85.63 / Max: 86Min: 85.1 / Avg: 85.73 / Max: 86.41. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0

OpenVINO

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

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Age Gender Recognition Retail 0013 FP32 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct14002800420056007000SE +/- 78.63, N = 6SE +/- 51.64, N = 36759.076751.86
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Age Gender Recognition Retail 0013 FP32 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct12002400360048006000Min: 6368.92 / Avg: 6759.07 / Max: 6872.96Min: 6676.68 / Avg: 6751.86 / Max: 6850.77

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: mobilenetLinux 5.9.1Linux 5.10 Git 23 Oct510152025SE +/- 0.09, N = 3SE +/- 0.11, N = 319.5019.48MIN: 18.92 / MAX: 47.26MIN: 19.04 / MAX: 21.71. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: mobilenetLinux 5.9.1Linux 5.10 Git 23 Oct510152025Min: 19.38 / Avg: 19.5 / Max: 19.68Min: 19.27 / Avg: 19.48 / Max: 19.641. (CXX) g++ options: -O3 -rdynamic -lgomp -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: Inception ResNet V2Linux 5.9.1Linux 5.10 Git 23 Oct500K1000K1500K2000K2500KSE +/- 1443.44, N = 3SE +/- 1302.40, N = 324154332417670
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Inception ResNet V2Linux 5.9.1Linux 5.10 Git 23 Oct400K800K1200K1600K2000KMin: 2412580 / Avg: 2415433.33 / Max: 2417240Min: 2415130 / Avg: 2417670 / Max: 2419440

OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: SqueezeNetLinux 5.9.1Linux 5.10 Git 23 Oct40K80K120K160K200KSE +/- 63.53, N = 3SE +/- 94.84, N = 3185313185481
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: SqueezeNetLinux 5.9.1Linux 5.10 Git 23 Oct30K60K90K120K150KMin: 185189 / Avg: 185313 / Max: 185399Min: 185291 / Avg: 185480.67 / Max: 185577

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: AtomicLinux 5.9.1Linux 5.10 Git 23 Oct160K320K480K640K800KSE +/- 511.18, N = 3SE +/- 104.09, N = 3764049.46764678.981. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: AtomicLinux 5.9.1Linux 5.10 Git 23 Oct130K260K390K520K650KMin: 763063.95 / Avg: 764049.46 / Max: 764777.76Min: 764570.21 / Avg: 764678.98 / Max: 764887.091. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

OpenVINO

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

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Person Detection 0106 FP32 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct6001200180024003000SE +/- 4.38, N = 3SE +/- 6.09, N = 32669.232667.29
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Person Detection 0106 FP32 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct5001000150020002500Min: 2660.48 / Avg: 2669.23 / Max: 2673.78Min: 2655.15 / Avg: 2667.29 / Max: 2674.14

Kvazaar

This is a test of Kvazaar as a CPU-based H.265 video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 4K - Video Preset: Very FastLinux 5.9.1Linux 5.10 Git 23 Oct48121620SE +/- 0.02, N = 3SE +/- 0.02, N = 316.3916.381. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 4K - Video Preset: Very FastLinux 5.9.1Linux 5.10 Git 23 Oct48121620Min: 16.35 / Avg: 16.39 / Max: 16.41Min: 16.34 / Avg: 16.38 / Max: 16.421. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: CPU StressLinux 5.9.1Linux 5.10 Git 23 Oct8001600240032004000SE +/- 23.34, N = 3SE +/- 15.17, N = 33567.533565.421. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: CPU StressLinux 5.9.1Linux 5.10 Git 23 Oct6001200180024003000Min: 3524 / Avg: 3567.53 / Max: 3603.91Min: 3535.56 / Avg: 3565.42 / Max: 3584.981. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: squeezenetLinux 5.9.1Linux 5.10 Git 23 Oct510152025SE +/- 0.09, N = 3SE +/- 0.10, N = 318.6718.66MIN: 18.34 / MAX: 19.43MIN: 18.24 / MAX: 32.621. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: squeezenetLinux 5.9.1Linux 5.10 Git 23 Oct510152025Min: 18.54 / Avg: 18.67 / Max: 18.84Min: 18.49 / Avg: 18.66 / Max: 18.841. (CXX) g++ options: -O3 -rdynamic -lgomp -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: Inception V4Linux 5.9.1Linux 5.10 Git 23 Oct600K1200K1800K2400K3000KSE +/- 1325.07, N = 3SE +/- 1229.12, N = 326728432674193
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Inception V4Linux 5.9.1Linux 5.10 Git 23 Oct500K1000K1500K2000K2500KMin: 2670300 / Avg: 2672843.33 / Max: 2674760Min: 2671860 / Avg: 2674193.33 / Max: 2676030

OpenVINO

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

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Person Detection 0106 FP16 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct6001200180024003000SE +/- 1.24, N = 3SE +/- 4.17, N = 32665.582666.60
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Person Detection 0106 FP16 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct5001000150020002500Min: 2663.22 / Avg: 2665.58 / Max: 2667.44Min: 2661.87 / Avg: 2666.6 / Max: 2674.91

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: yolov4-tinyLinux 5.9.1Linux 5.10 Git 23 Oct714212835SE +/- 0.08, N = 3SE +/- 0.06, N = 331.3531.36MIN: 31.1 / MAX: 32.16MIN: 30.94 / MAX: 50.631. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: yolov4-tinyLinux 5.9.1Linux 5.10 Git 23 Oct714212835Min: 31.23 / Avg: 31.35 / Max: 31.5Min: 31.24 / Avg: 31.36 / Max: 31.461. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

LevelDB

LevelDB is a key-value storage library developed by Google that supports making use of Snappy for data compression and has other modern features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Random DeleteLinux 5.9.1Linux 5.10 Git 23 Oct1122334455SE +/- 0.13, N = 3SE +/- 0.06, N = 350.6050.621. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Random DeleteLinux 5.9.1Linux 5.10 Git 23 Oct1020304050Min: 50.35 / Avg: 50.6 / Max: 50.74Min: 50.55 / Avg: 50.62 / Max: 50.731. (CXX) g++ options: -O3 -lsnappy -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 FloatLinux 5.9.1Linux 5.10 Git 23 Oct30K60K90K120K150KSE +/- 45.17, N = 3SE +/- 67.04, N = 3124015124051
OpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet FloatLinux 5.9.1Linux 5.10 Git 23 Oct20K40K60K80K100KMin: 123925 / Avg: 124015.33 / Max: 124061Min: 123917 / Avg: 124051 / Max: 124122

WebP Image Encode

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

OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Lossless, Highest CompressionLinux 5.9.1Linux 5.10 Git 23 Oct714212835SE +/- 0.09, N = 3SE +/- 0.09, N = 332.1632.171. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff
OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Lossless, Highest CompressionLinux 5.9.1Linux 5.10 Git 23 Oct714212835Min: 32.04 / Avg: 32.16 / Max: 32.35Min: 32.06 / Avg: 32.17 / Max: 32.341. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Vector MathLinux 5.9.1Linux 5.10 Git 23 Oct16K32K48K64K80KSE +/- 2.74, N = 3SE +/- 6.32, N = 376006.7776005.991. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.11.07Test: Vector MathLinux 5.9.1Linux 5.10 Git 23 Oct13K26K39K52K65KMin: 76001.54 / Avg: 76006.77 / Max: 76010.78Min: 75994.75 / Avg: 76005.99 / Max: 76016.611. (CC) gcc options: -O2 -std=gnu99 -lm -laio -lbsd -lcrypt -lrt -lz -ldl -lpthread -lc

Basemark GPU

This is a benchmark of Basemark GPU. For this test profile to work, you must have a valid license/copy of BasemarkGPU in your Phoronix Test Suite download cache. This test profile simply automates the execution of BasemarkGPU and you must already have the Windows .zip or Linux .tar.gz in the download cache. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterBasemark GPU 1.2Renderer: Vulkan - Resolution: 1920 x 1080 - Graphics Preset: MediumLinux 5.9.1130260390520650SE +/- 2.74, N = 3606.02MIN: 332.72 / MAX: 1088.99

OpenBenchmarking.orgFPS, More Is BetterBasemark GPU 1.2Renderer: Vulkan - Resolution: 1920 x 1080 - Graphics Preset: HighLinux 5.9.1306090120150SE +/- 0.70, N = 3119.63MIN: 96.97 / MAX: 192.07

Xonotic

This is a benchmark of Xonotic, which is a fork of the DarkPlaces-based Nexuiz game. Development began in March of 2010 on the Xonotic game. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterXonotic 0.8.2Resolution: 1920 x 1080 - Effects Quality: UltimateLinux 5.9.170140210280350SE +/- 3.01, N = 3315.02MIN: 41 / MAX: 835

Unigine Superposition

This test calculates the average frame-rate within the Superposition demo for the Unigine engine, released in 2017. This engine is extremely demanding on the system's graphics card. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterUnigine Superposition 1.0Resolution: 1920 x 1080 - Mode: Fullscreen - Quality: Ultra - Renderer: OpenGLLinux 5.9.148121620SE +/- 0.03, N = 318.0MAX: 21.6

OpenBenchmarking.orgFrames Per Second, More Is BetterUnigine Superposition 1.0Resolution: 1920 x 1080 - Mode: Fullscreen - Quality: High - Renderer: OpenGLLinux 5.9.1918273645SE +/- 0.06, N = 341.2MAX: 52.6

Tesseract

Tesseract is a fork of Cube 2 Sauerbraten with numerous graphics and game-play improvements. Tesseract has been in development since 2012 while its first release happened in May of 2014. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterTesseract 2014-05-12Resolution: 1920 x 1080Linux 5.9.1110220330440550SE +/- 0.92, N = 3529.03

InfluxDB

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

OpenBenchmarking.orgval/sec, More Is BetterInfluxDB 1.8.2Concurrent Streams: 1024 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000Linux 5.9.1300K600K900K1200K1500KSE +/- 714.22, N = 31451332.5

OpenBenchmarking.orgval/sec, More Is BetterInfluxDB 1.8.2Concurrent Streams: 64 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000Linux 5.9.1300K600K900K1200K1500KSE +/- 2616.83, N = 31420949.5

OpenBenchmarking.orgval/sec, More Is BetterInfluxDB 1.8.2Concurrent Streams: 4 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000Linux 5.9.1300K600K900K1200K1500KSE +/- 4975.23, N = 31225807.0

Selenium

This test profile uses the Selenium WebDriver for running various browser benchmarks in different available web browsers. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterSeleniumBenchmark: WASM collisionDetection - Browser: FirefoxLinux 5.9.180160240320400SE +/- 1.70, N = 3364.71. firefox 81.0.2

OpenBenchmarking.orgms, Fewer Is BetterSeleniumBenchmark: WASM imageConvolute - Browser: FirefoxLinux 5.9.1612182430SE +/- 0.25, N = 327.11. firefox 81.0.2

OpenBenchmarking.orgRuns Per Minute, More Is BetterSeleniumBenchmark: Speedometer - Browser: FirefoxLinux 5.9.120406080100SE +/- 1.37, N = 398.31. firefox 81.0.2

OpenBenchmarking.orgSeconds, Fewer Is BetterSeleniumBenchmark: Maze Solver - Browser: FirefoxLinux 5.9.10.991.982.973.964.95SE +/- 0.00, N = 34.41. firefox 81.0.2

OpenBenchmarking.orgScore, More Is BetterSeleniumBenchmark: Jetstream 2 - Browser: FirefoxLinux 5.9.120406080100SE +/- 0.44, N = 3107.971. firefox 81.0.2

OpenBenchmarking.orgScore, More Is BetterSeleniumBenchmark: WebXPRT - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct60120180240300SE +/- 1.00, N = 32732731. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0
OpenBenchmarking.orgScore, More Is BetterSeleniumBenchmark: WebXPRT - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct50100150200250Min: 271 / Avg: 273 / Max: 2741. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0

OpenVINO

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

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Age Gender Recognition Retail 0013 FP16 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct0.12830.25660.38490.51320.6415SE +/- 0.00, N = 3SE +/- 0.00, N = 30.570.57
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2021.1Model: Age Gender Recognition Retail 0013 FP16 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 0.57 / Avg: 0.57 / Max: 0.58Min: 0.57 / Avg: 0.57 / Max: 0.58

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Person Detection 0106 FP16 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct0.33530.67061.00591.34121.6765SE +/- 0.00, N = 3SE +/- 0.00, N = 31.491.49
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Person Detection 0106 FP16 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 1.49 / Avg: 1.49 / Max: 1.49Min: 1.49 / Avg: 1.49 / Max: 1.5

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Face Detection 0106 FP32 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct0.42080.84161.26241.68322.104SE +/- 0.00, N = 3SE +/- 0.01, N = 31.871.87
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2021.1Model: Face Detection 0106 FP32 - Device: CPULinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 1.86 / Avg: 1.87 / Max: 1.87Min: 1.85 / Avg: 1.87 / Max: 1.89

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: alexnetLinux 5.9.1Linux 5.10 Git 23 Oct3691215SE +/- 0.09, N = 3SE +/- 0.05, N = 39.989.98MIN: 9.01 / MAX: 20.09MIN: 9.31 / MAX: 15.621. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: alexnetLinux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 9.81 / Avg: 9.98 / Max: 10.09Min: 9.91 / Avg: 9.98 / Max: 10.081. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: blazefaceLinux 5.9.1Linux 5.10 Git 23 Oct0.34430.68861.03291.37721.7215SE +/- 0.00, N = 3SE +/- 0.00, N = 31.531.53MIN: 1.5 / MAX: 1.69MIN: 1.49 / MAX: 1.651. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: blazefaceLinux 5.9.1Linux 5.10 Git 23 Oct246810Min: 1.53 / Avg: 1.53 / Max: 1.54Min: 1.52 / Avg: 1.53 / Max: 1.531. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: googlenetLinux 5.9.1Linux 5.10 Git 23 Oct48121620SE +/- 0.06, N = 3SE +/- 0.10, N = 318.2418.24MIN: 18.03 / MAX: 18.97MIN: 17.96 / MAX: 72.151. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU - Model: googlenetLinux 5.9.1Linux 5.10 Git 23 Oct510152025Min: 18.16 / Avg: 18.24 / Max: 18.36Min: 18.11 / Avg: 18.24 / Max: 18.431. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU-v2-v2 - Model: mobilenet-v2Linux 5.9.1Linux 5.10 Git 23 Oct246810SE +/- 0.02, N = 3SE +/- 0.02, N = 36.386.38MIN: 6.26 / MAX: 7.79MIN: 6.25 / MAX: 7.851. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: CPU-v2-v2 - Model: mobilenet-v2Linux 5.9.1Linux 5.10 Git 23 Oct3691215Min: 6.34 / Avg: 6.38 / Max: 6.42Min: 6.34 / Avg: 6.38 / Max: 6.421. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Selenium

This test profile uses the Selenium WebDriver for running various browser benchmarks in different available web browsers. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterSeleniumBenchmark: MotionMark - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct306090120150SE +/- 9.00, N = 6SE +/- 6.41, N = 9116.88111.811. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0
OpenBenchmarking.orgScore, More Is BetterSeleniumBenchmark: MotionMark - Browser: FirefoxLinux 5.9.1Linux 5.10 Git 23 Oct20406080100Min: 81.01 / Avg: 116.88 / Max: 150.09Min: 84 / Avg: 111.81 / Max: 144.191. Linux 5.9.1: firefox 81.0.22. Linux 5.10 Git 23 Oct: firefox 82.0

LevelDB

LevelDB is a key-value storage library developed by Google that supports making use of Snappy for data compression and has other modern features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Fill SyncLinux 5.9.1Linux 5.10 Git 23 Oct14002800420056007000SE +/- 154.24, N = 15SE +/- 381.08, N = 36684.946147.491. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Fill SyncLinux 5.9.1Linux 5.10 Git 23 Oct12002400360048006000Min: 5387.56 / Avg: 6684.94 / Max: 7308.39Min: 5520.2 / Avg: 6147.49 / Max: 6836.041. (CXX) g++ options: -O3 -lsnappy -lpthread

OpenBenchmarking.orgMB/s, More Is BetterLevelDB 1.22Benchmark: Fill SyncLinux 5.9.1Linux 5.10 Git 23 Oct0.06750.1350.20250.270.3375SE +/- 0.01, N = 15SE +/- 0.00, N = 30.30.31. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMB/s, More Is BetterLevelDB 1.22Benchmark: Fill SyncLinux 5.9.1Linux 5.10 Git 23 Oct12345Min: 0.2 / Avg: 0.26 / Max: 0.3Min: 0.3 / Avg: 0.3 / Max: 0.31. (CXX) g++ options: -O3 -lsnappy -lpthread

OSBench

OSBench is a collection of micro-benchmarks for measuring operating system primitives like time to create threads/processes, launching programs, creating files, and memory allocation. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgus Per Event, Fewer Is BetterOSBenchTest: Create ProcessesLinux 5.9.1Linux 5.10 Git 23 Oct510152025SE +/- 0.72, N = 13SE +/- 0.15, N = 320.6120.931. (CC) gcc options: -lm
OpenBenchmarking.orgus Per Event, Fewer Is BetterOSBenchTest: Create ProcessesLinux 5.9.1Linux 5.10 Git 23 Oct510152025Min: 17.86 / Avg: 20.61 / Max: 28.58Min: 20.62 / Avg: 20.93 / Max: 21.081. (CC) gcc options: -lm

Sockperf

This is a network socket API performance benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgusec, Fewer Is BetterSockperf 3.4Test: Latency Under LoadLinux 5.9.1Linux 5.10 Git 23 Oct20406080100SE +/- 13.12, N = 25SE +/- 12.91, N = 2467.3798.691. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread
OpenBenchmarking.orgusec, Fewer Is BetterSockperf 3.4Test: Latency Under LoadLinux 5.9.1Linux 5.10 Git 23 Oct20406080100Min: 6.74 / Avg: 67.37 / Max: 174.8Min: 7.19 / Avg: 98.69 / Max: 191.51. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread

IOR

IOR is a parallel I/O storage benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterIOR 3.2.1Read TestLinux 5.9.1Linux 5.10 Git 23 Oct90180270360450SE +/- 27.46, N = 12SE +/- 31.04, N = 12347.81394.46MIN: 159.63 / MAX: 584.11MIN: 202.79 / MAX: 560.751. (CC) gcc options: -O2 -lm -pthread -lmpi
OpenBenchmarking.orgMB/s, More Is BetterIOR 3.2.1Read TestLinux 5.9.1Linux 5.10 Git 23 Oct70140210280350Min: 274.58 / Avg: 347.81 / Max: 510.36Min: 273.09 / Avg: 394.46 / Max: 511.751. (CC) gcc options: -O2 -lm -pthread -lmpi

OpenBenchmarking.orgMB/s, More Is BetterIOR 3.2.1Write TestLinux 5.9.1Linux 5.10 Git 23 Oct4080120160200SE +/- 18.20, N = 12SE +/- 21.71, N = 12159.91195.45MIN: 8.52 / MAX: 303.02MIN: 8.49 / MAX: 357.271. (CC) gcc options: -O2 -lm -pthread -lmpi
OpenBenchmarking.orgMB/s, More Is BetterIOR 3.2.1Write TestLinux 5.9.1Linux 5.10 Git 23 Oct4080120160200Min: 110.51 / Avg: 159.91 / Max: 270.52Min: 112.58 / Avg: 195.45 / Max: 286.561. (CC) gcc options: -O2 -lm -pthread -lmpi

178 Results Shown

NCNN:
  Vulkan GPU-v2-v2 - mobilenet-v2
  Vulkan GPU - resnet18
  Vulkan GPU - efficientnet-b0
  Vulkan GPU-v3-v3 - mobilenet-v3
  Vulkan GPU - mnasnet
  Vulkan GPU - googlenet
  Vulkan GPU - squeezenet
Stress-NG
NCNN
Hackbench:
  8 - Process
  4 - Process
  8 - Thread
  4 - Thread
  16 - Process
  16 - Thread
  2 - Process
  2 - Thread
dav1d
NCNN
Hackbench
LAMMPS Molecular Dynamics Simulator
dav1d:
  Summer Nature 4K
  Summer Nature 1080p
LAMMPS Molecular Dynamics Simulator
Stress-NG
Hackbench:
  1 - Process
  1 - Thread
Stress-NG
PostMark
OSBench
Apache Siege
NCNN
LeelaChessZero
Stress-NG
Apache Siege
Stress-NG
OSBench
GPAW
Selenium
Stress-NG:
  Glibc C String Functions
  CPU Cache
Apache Siege
Selenium
WebP Image Encode
Selenium
Stress-NG
LibRaw
Waifu2x-NCNN Vulkan
WebP Image Encode
KeyDB
OpenVINO
Sockperf
NCNN
Stress-NG
LevelDB
Apache Siege
Dolfyn
Apache CouchDB
LevelDB
Selenium
LeelaChessZero
eSpeak-NG Speech Engine
Apache Siege
Waifu2x-NCNN Vulkan
WebP Image Encode
Stress-NG
TNN
VkFFT
OSBench
Zstd Compression
Stress-NG
NCNN
x265
RealSR-NCNN
Selenium
RealSR-NCNN
dav1d
TensorFlow Lite
Kvazaar
SVT-AV1
Stress-NG:
  Semaphores
  NUMA
Kvazaar
NCNN
TNN
NCNN
Zstd Compression
OpenVINO
NCNN:
  CPU - vgg16
  CPU - efficientnet-b0
LevelDB
SVT-AV1
NCNN
OSBench
SVT-AV1
OpenVINO
NCNN
Kvazaar
x265
Selenium
LevelDB
WebP Image Encode
Timed MAFFT Alignment
OpenVINO
Kvazaar
Stress-NG
Kvazaar:
  Bosphorus 1080p - Medium
  Bosphorus 4K - Medium
LevelDB
Blender
NCNN
LevelDB:
  Rand Fill
  Seq Fill
Sockperf
LevelDB
Blender
Stress-NG
Blender:
  Barbershop - CPU-Only
  Fishy Cat - CPU-Only
OpenVINO
NCNN
Kvazaar
Caffe
NCNN
LevelDB
Blender
Caffe
Stress-NG
TensorFlow Lite
NAMD
OpenVINO
GROMACS
Selenium
OpenVINO
NCNN
TensorFlow Lite:
  Inception ResNet V2
  SqueezeNet
Stress-NG
OpenVINO
Kvazaar
Stress-NG
NCNN
TensorFlow Lite
OpenVINO
NCNN
LevelDB
TensorFlow Lite
WebP Image Encode
Stress-NG
Basemark GPU:
  Vulkan - 1920 x 1080 - Medium
  Vulkan - 1920 x 1080 - High
Xonotic
Unigine Superposition:
  1920 x 1080 - Fullscreen - Ultra - OpenGL
  1920 x 1080 - Fullscreen - High - OpenGL
Tesseract
InfluxDB:
  1024 - 10000 - 2,5000,1 - 10000
  64 - 10000 - 2,5000,1 - 10000
  4 - 10000 - 2,5000,1 - 10000
Selenium:
  WASM collisionDetection - Firefox
  WASM imageConvolute - Firefox
  Speedometer - Firefox
  Maze Solver - Firefox
  Jetstream 2 - Firefox
  WebXPRT - Firefox
OpenVINO:
  Age Gender Recognition Retail 0013 FP16 - CPU
  Person Detection 0106 FP16 - CPU
  Face Detection 0106 FP32 - CPU
NCNN:
  Vulkan GPU - alexnet
  Vulkan GPU - blazeface
  CPU - googlenet
  CPU-v2-v2 - mobilenet-v2
Selenium
LevelDB:
  Fill Sync:
    Microseconds Per Op
    MB/s
OSBench
Sockperf
IOR:
  Read Test
  Write Test

Linux 5.9.1

Processor: AMD Ryzen 7 3800XT 8-Core @ 3.90GHz (8 Cores / 16 Threads), Motherboard: ASRock X570 Pro4 (P3.00 BIOS), Chipset: AMD Starship/Matisse, Memory: 16GB, Disk: 500GB CT500P2SSD8, Graphics: Sapphire AMD Radeon RX 470/480/570/570X/580/580X/590 8GB (1560/2100MHz), Audio: AMD Ellesmere HDMI Audio, Monitor: G237HL, Network: Intel I211

OS: Ubuntu 20.04, Kernel: 5.9.1-050901-generic (x86_64), Desktop: GNOME Shell 3.36.4, Display Server: X Server 1.20.8, Display Driver: modesetting 1.20.8, OpenGL: 4.6 Mesa 20.0.4 (LLVM 9.0.1), Vulkan: 1.2.128, Compiler: GCC 9.3.0, File-System: ext4, Screen Resolution: 1920x1080

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,gm2 --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-9-HskZEa/gcc-9-9.3.0/debian/tmp-nvptx/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Disk Notes: NONE / errors=remount-ro,relatime,rw
Processor Notes: Scaling Governor: acpi-cpufreq ondemand - CPU Microcode: 0x8701021
Graphics Notes: GLAMOR
Python Notes: Python 3.8.5
Security Notes: 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 STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 23 October 2020 14:19 by user phoronix.

Linux 5.10 Git 23 Oct

Processor: AMD Ryzen 7 3800XT 8-Core @ 3.90GHz (8 Cores / 16 Threads), Motherboard: ASRock X570 Pro4 (P3.00 BIOS), Chipset: AMD Starship/Matisse, Memory: 16GB, Disk: 500GB CT500P2SSD8, Graphics: Sapphire AMD Radeon RX 470/480/570/570X/580/580X/590 8GB (1560/2100MHz), Audio: AMD Ellesmere HDMI Audio, Monitor: G237HL, Network: Intel I211

OS: Ubuntu 20.04, Kernel: 5.9.0-050900daily20201023-generic (x86_64), Desktop: GNOME Shell 3.36.4, Display Server: X Server 1.20.8, Display Driver: modesetting 1.20.8, OpenGL: 4.6 Mesa 20.0.4 (LLVM 9.0.1), Vulkan: 1.2.128, Compiler: GCC 9.3.0, File-System: ext4, Screen Resolution: 1920x1080

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,gm2 --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-9-HskZEa/gcc-9-9.3.0/debian/tmp-nvptx/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Disk Notes: NONE / errors=remount-ro,no_fc,relatime,rw
Processor Notes: Scaling Governor: acpi-cpufreq ondemand - CPU Microcode: 0x8701021
Graphics Notes: GLAMOR
Python Notes: Python 3.8.5
Security Notes: 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 STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 24 October 2020 06:07 by user phoronix.