Tests for a future article. 2 x AMD EPYC 7573X 32-Core testing with a AMD DAYTONA_X (RYM1009B BIOS) and ASPEED on Ubuntu 22.04 via the Phoronix Test Suite.
a Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa001229Python Notes: Python 3.10.6Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected
b Processor: AMD EPYC 7773X 64-Core @ 2.20GHz (64 Cores / 128 Threads), Motherboard: AMD DAYTONA_X (RYM1009B BIOS), Chipset: AMD Starship/Matisse, Memory: 256GB, Disk: 3841GB Micron_9300_MTFDHAL3T8TDP, Graphics: ASPEED, Monitor: VE228, Network: 2 x Mellanox MT27710
OS: Ubuntu 22.04, Kernel: 5.15.0-47-generic (x86_64), Desktop: GNOME Shell 42.4, Display Server: X Server 1.21.1.3, Vulkan: 1.2.204, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 1920x1080
5 a 5 b Changed Processor to AMD EPYC 7573X 32-Core @ 2.80GHz (32 Cores / 64 Threads) .
5 2p a 5 2p b Changed Processor to 2 x AMD EPYC 7573X 32-Core @ 2.80GHz (64 Cores / 128 Threads) .
Changed Memory to 512GB .
PETSc PETSc, the Portable, Extensible Toolkit for Scientific Computation, is for the scalable (parallel) solution of scientific applications modeled by partial differential equations. This test profile runs the PETSc "make streams" benchmark and records the throughput rate when all available cores are utilized for the MPI Streams build. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MB/s, More Is Better PETSc 3.19 Test: Streams 5 2p b b 5 a 16K 32K 48K 64K 80K 74130.18 56185.75 31992.61 1. (CC) gcc options: -fPIC -O3 -O2 -lpthread -ludev -lpciaccess -lm
OpenVKL OpenVKL is the Intel Open Volume Kernel Library that offers high-performance volume computation kernels and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Items / Sec, More Is Better OpenVKL 1.3.1 Benchmark: vklBenchmark ISPC a b 5 2p b 5 2p a 5 a 5 b 100 200 300 400 500 SE +/- 0.33, N = 3 470 469 452 452 340 339 MIN: 84 / MAX: 2616 MIN: 84 / MAX: 2565 MIN: 99 / MAX: 2013 MIN: 98 / MAX: 1875 MIN: 55 / MAX: 2309 MIN: 54 / MAX: 2307
OpenFOAM OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Medium Mesh Size - Execution Time a 5 2p b 5 2p a b 5 b 5 a 90 180 270 360 450 40.48 203.65 205.46 363.74 427.71 428.66 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm
OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Medium Mesh Size - Mesh Time a 5 2p a 5 2p b 5 a 5 b b 30 60 90 120 150 25.35 99.38 99.93 117.44 117.60 120.39 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm
Blender Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.5 Blend File: Barbershop - Compute: CPU-Only 5 2p b 5 2p a b a 5 a 90 180 270 360 450 SE +/- 0.09, N = 3 213.46 213.60 259.87 260.83 408.79
OpenCV This is a benchmark of the OpenCV (Computer Vision) library's built-in performance tests. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better OpenCV 4.7 Test: Graph API 5 a b 5 2p b 90K 180K 270K 360K 450K 206970 235810 419702 1. (CXX) g++ options: -fPIC -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -shared
FFmpeg This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org FPS, More Is Better FFmpeg 6.0 Encoder: libx265 - Scenario: Platform 5 a 5 2p a b a 5 2p b 11 22 33 44 55 SE +/- 0.02, N = 3 46.70 44.31 43.90 43.73 43.64 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.org Seconds, Fewer Is Better FFmpeg 6.0 Encoder: libx265 - Scenario: Platform 5 a 5 2p a b a 5 2p b 40 80 120 160 200 SE +/- 0.08, N = 3 162.19 170.94 172.57 173.23 173.56 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.org FPS, More Is Better FFmpeg 6.0 Encoder: libx265 - Scenario: Video On Demand 5 a 5 2p a b 5 2p b a 11 22 33 44 55 SE +/- 0.04, N = 3 46.89 44.97 43.90 43.77 43.67 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.org Seconds, Fewer Is Better FFmpeg 6.0 Encoder: libx265 - Scenario: Video On Demand 5 a 5 2p a b 5 2p b a 40 80 120 160 200 SE +/- 0.15, N = 3 161.54 168.46 172.53 173.05 173.47 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
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.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: FastestDet 5 a b 5 2p a 5 2p b 12 24 36 48 60 9.25 19.08 40.45 53.25 MIN: 9.12 / MAX: 9.81 MIN: 13.65 / MAX: 21.7 MIN: 27.4 / MAX: 462.07 MIN: 30.07 / MAX: 66.77 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: vision_transformer 5 a b 5 2p a 5 2p b 30 60 90 120 150 126.26 133.39 144.31 145.49 MIN: 125.42 / MAX: 132.03 MIN: 129.7 / MAX: 252.77 MIN: 140.5 / MAX: 157.63 MIN: 141.27 / MAX: 245.34 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: regnety_400m 5 a b 5 2p a 5 2p b 30 60 90 120 150 26.34 53.52 100.65 128.35 MIN: 25.99 / MAX: 28.26 MIN: 50.93 / MAX: 71.06 MIN: 97.81 / MAX: 136.12 MIN: 111.43 / MAX: 240.27 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: squeezenet_ssd 5 a b 5 2p b 5 2p a 9 18 27 36 45 14.85 19.32 35.81 37.17 MIN: 14.49 / MAX: 25.32 MIN: 18.86 / MAX: 22.39 MIN: 29.14 / MAX: 58.45 MIN: 30.62 / MAX: 51.54 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: yolov4-tiny 5 a b 5 2p a 5 2p b 10 20 30 40 50 19.98 23.92 38.10 45.89 MIN: 19.53 / MAX: 23.19 MIN: 23.17 / MAX: 30.48 MIN: 29.08 / MAX: 101.85 MIN: 34.08 / MAX: 64.66 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: resnet50 5 a b 5 2p a 5 2p b 30 60 90 120 150 14.80 19.58 80.81 120.27 MIN: 14.6 / MAX: 16.84 MIN: 19.15 / MAX: 44.79 MIN: 62.51 / MAX: 112.38 MIN: 42.99 / MAX: 192.41 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: alexnet 5 a b 5 2p a 5 2p b 7 14 21 28 35 5.48 7.13 18.60 28.55 MIN: 5.36 / MAX: 6.34 MIN: 6.97 / MAX: 7.76 MIN: 17.59 / MAX: 34.64 MIN: 12.6 / MAX: 62.18 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: resnet18 5 a b 5 2p a 5 2p b 12 24 36 48 60 8.12 10.89 38.94 51.80 MIN: 8.01 / MAX: 10.04 MIN: 10.68 / MAX: 11.74 MIN: 16.05 / MAX: 124.98 MIN: 16.09 / MAX: 93.56 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: vgg16 5 a b 5 2p b 5 2p a 8 16 24 32 40 21.08 24.04 32.81 33.55 MIN: 20.78 / MAX: 24.35 MIN: 23.49 / MAX: 30.66 MIN: 30.06 / MAX: 44.76 MIN: 28.65 / MAX: 42.53 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: googlenet 5 a b 5 2p a 5 2p b 30 60 90 120 150 14.11 19.09 93.29 112.43 MIN: 13.96 / MAX: 17.22 MIN: 18.77 / MAX: 25.76 MIN: 52.46 / MAX: 137.19 MIN: 29.99 / MAX: 148.59 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: blazeface 5 a b 5 2p a 5 2p b 6 12 18 24 30 3.82 8.12 14.65 24.56 MIN: 3.45 / MAX: 79.74 MIN: 6.94 / MAX: 11.1 MIN: 11.34 / MAX: 73.36 MIN: 20.64 / MAX: 91.2 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: efficientnet-b0 5 a b 5 2p a 5 2p b 14 28 42 56 70 9.03 15.44 35.12 60.79 MIN: 8.93 / MAX: 11.07 MIN: 13.63 / MAX: 18.67 MIN: 33.83 / MAX: 41.96 MIN: 47.67 / MAX: 141.85 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: mnasnet 5 a b 5 2p a 5 2p b 11 22 33 44 55 6.08 11.71 38.49 49.46 MIN: 6.01 / MAX: 6.54 MIN: 9.43 / MAX: 20.81 MIN: 25.51 / MAX: 75.24 MIN: 36.42 / MAX: 176.46 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: shufflenet-v2 5 a b 5 2p a 5 2p b 9 18 27 36 45 7.73 15.60 33.27 37.89 MIN: 7.58 / MAX: 9.71 MIN: 12.95 / MAX: 19.73 MIN: 29.37 / MAX: 96.69 MIN: 34.63 / MAX: 113.32 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU-v3-v3 - Model: mobilenet-v3 5 a b 5 2p a 5 2p b 7 14 21 28 35 6.19 10.29 21.26 28.88 MIN: 6.05 / MAX: 6.93 MIN: 9.51 / MAX: 11.93 MIN: 20.79 / MAX: 28.25 MIN: 23.66 / MAX: 168.43 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU-v2-v2 - Model: mobilenet-v2 5 a b 5 2p a 5 2p b 9 18 27 36 45 6.46 11.30 27.70 38.67 MIN: 6.35 / MAX: 8.73 MIN: 9.74 / MAX: 14.96 MIN: 23.18 / MAX: 43.54 MIN: 28.02 / MAX: 119.64 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: mobilenet 5 a b 5 2p a 5 2p b 30 60 90 120 150 13.70 17.84 77.77 115.68 MIN: 13.55 / MAX: 14.38 MIN: 17.55 / MAX: 25.78 MIN: 67.11 / MAX: 156 MIN: 64.45 / MAX: 159.93 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
FFmpeg This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org FPS, More Is Better FFmpeg 6.0 Encoder: libx265 - Scenario: Upload 5 a 5 b 5 2p a 5 2p b b a 5 10 15 20 25 SE +/- 0.01, N = 3 22.95 22.93 22.38 22.10 21.62 21.59 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.org Seconds, Fewer Is Better FFmpeg 6.0 Encoder: libx265 - Scenario: Upload 5 a 5 b 5 2p a 5 2p b b a 30 60 90 120 150 SE +/- 0.05, N = 3 110.01 110.14 112.82 114.26 116.81 116.93 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenCV This is a benchmark of the OpenCV (Computer Vision) library's built-in performance tests. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better OpenCV 4.7 Test: Stitching 5 a b 5 2p b 60K 120K 180K 240K 300K 184032 201160 287492 1. (CXX) g++ options: -fPIC -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -shared
ASKAP ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Million Grid Points Per Second, More Is Better ASKAP 1.0 Test: tConvolve MT - Degridding 5 2p a 5 2p b 5 a b 3K 6K 9K 12K 15K 12506.70 11068.80 9611.05 8308.33 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
OpenBenchmarking.org Million Grid Points Per Second, More Is Better ASKAP 1.0 Test: tConvolve MT - Gridding 5 2p a 5 2p b 5 a b 1700 3400 5100 6800 8500 7738.59 7176.41 6599.68 5493.35 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
ClickHouse ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, Third Run 5 2p a 5 a 5 2p b b 100 200 300 400 500 461.53 456.30 448.03 437.35 MIN: 41.49 / MAX: 3000 MIN: 24.65 / MAX: 5454.55 MIN: 41.38 / MAX: 2608.7 MIN: 35.59 / MAX: 5454.55
OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, Second Run 5 2p b 5 2p a 5 a b 100 200 300 400 500 467.46 463.68 457.81 439.07 MIN: 40.6 / MAX: 4615.38 MIN: 41.64 / MAX: 4615.38 MIN: 24.13 / MAX: 5454.55 MIN: 35.82 / MAX: 6000
OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, First Run / Cold Cache 5 2p a 5 2p b 5 a b 100 200 300 400 500 445.18 440.56 437.20 428.51 MIN: 41.18 / MAX: 3157.89 MIN: 40.98 / MAX: 4000 MIN: 24.3 / MAX: 6000 MIN: 34.8 / MAX: 5454.55
VVenC VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better VVenC 1.8 Video Input: Bosphorus 4K - Video Preset: Fast 5 a 5 b a b 5 2p b 5 2p a 2 4 6 8 10 SE +/- 0.010, N = 3 6.406 6.390 6.286 6.128 5.768 5.548 1. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects
Blender Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.5 Blend File: Pabellon Barcelona - Compute: CPU-Only 5 2p a 5 2p b b a 5 a 30 60 90 120 150 SE +/- 0.10, N = 3 71.64 71.73 88.73 88.76 136.58
OpenCV This is a benchmark of the OpenCV (Computer Vision) library's built-in performance tests. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better OpenCV 4.7 Test: Core 5 a b 5 2p b 50K 100K 150K 200K 250K 68343 77061 236445 1. (CXX) g++ options: -fPIC -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -shared
Blender Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.5 Blend File: Classroom - Compute: CPU-Only 5 2p a 5 2p b b a 5 a 30 60 90 120 150 SE +/- 0.17, N = 3 58.55 58.57 72.02 72.45 113.31
FFmpeg This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org FPS, More Is Better FFmpeg 6.0 Encoder: libx265 - Scenario: Live 5 a 5 b 5 2p b b a 5 2p a 20 40 60 80 100 SE +/- 0.29, N = 3 110.49 110.29 107.29 106.98 105.78 92.52 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.org Seconds, Fewer Is Better FFmpeg 6.0 Encoder: libx265 - Scenario: Live 5 a 5 b 5 2p b b a 5 2p a 12 24 36 48 60 SE +/- 0.13, N = 3 45.71 45.79 47.07 47.21 47.74 54.58 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
oneDNN This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.1 Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU 5 a b 5 2p b 5 2p a 300 600 900 1200 1500 1182.61 1183.63 1385.33 1390.36 MIN: 1161.83 MIN: 1162.73 MIN: 1314.76 MIN: 1287.43 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.1 Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU 5 a b 5 2p a 5 2p b 300 600 900 1200 1500 679.65 749.62 1212.84 1305.96 MIN: 664.64 MIN: 731.75 MIN: 1070.52 MIN: 1060.61 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: HMAC-SHA512 a b 5 a 5 b 5 2p a 5 2p b 30M 60M 90M 120M 150M SE +/- 266743.32, N = 3 136796000 135165000 96059000 94924000 79002000 73260000 1. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2
Zstd Compression This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 19, Long Mode - Decompression Speed 5 a 5 2p b 5 2p a b 300 600 900 1200 1500 1229.1 1227.1 1222.9 1190.8 1. (CC) gcc options: -O3 -pthread -lz -llzma -llz4
VVenC VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better VVenC 1.8 Video Input: Bosphorus 4K - Video Preset: Faster 5 b 5 a a b 5 2p b 5 2p a 3 6 9 12 15 SE +/- 0.022, N = 3 12.020 11.961 11.367 11.364 9.906 9.574 1. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects
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.org ms, Fewer Is Better OpenVINO 2022.3 Model: Person Detection FP16 - Device: CPU 5 a 5 2p b 5 2p a b 800 1600 2400 3200 4000 2881.04 2948.39 2975.15 3911.97 MIN: 1536.21 / MAX: 3142.35 MIN: 1547.54 / MAX: 3537.3 MIN: 2241.37 / MAX: 3616.82 MIN: 3337.4 / MAX: 4451.63 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Person Detection FP16 - Device: CPU 5 2p b 5 2p a b 5 a 3 6 9 12 15 10.63 10.54 8.05 5.48 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenFOAM OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Small Mesh Size - Execution Time 5 2p b 5 2p a b a 5 b 5 a 11 22 33 44 55 32.97 33.00 40.24 40.45 50.70 50.75 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm
OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Small Mesh Size - Mesh Time 5 a 5 2p a 5 b 5 2p b a b 6 12 18 24 30 22.34 22.56 22.71 23.34 24.73 25.60 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm
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.org ms, Fewer Is Better OpenVINO 2022.3 Model: Person Detection FP32 - Device: CPU 5 a 5 2p a 5 2p b b 800 1600 2400 3200 4000 2886.12 2946.97 2952.76 3927.00 MIN: 1694.68 / MAX: 3104.67 MIN: 2004.15 / MAX: 3534.38 MIN: 2193.59 / MAX: 3652.32 MIN: 3402.58 / MAX: 4474.47 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Person Detection FP32 - Device: CPU 5 2p b 5 2p a b 5 a 3 6 9 12 15 10.64 10.64 8.04 5.47 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
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.org ms, Fewer Is Better OpenVINO 2022.3 Model: Face Detection FP16 - Device: CPU 5 a 5 2p a 5 2p b b 600 1200 1800 2400 3000 1935.60 2013.52 2020.21 2608.34 MIN: 1852.05 / MAX: 1974.19 MIN: 1890.96 / MAX: 2802.82 MIN: 1823.96 / MAX: 3111.51 MIN: 2421.38 / MAX: 2754.89 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Face Detection FP16 - Device: CPU 5 2p a 5 2p b b 5 a 4 8 12 16 20 15.86 15.76 12.16 8.23 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Face Detection FP16-INT8 - Device: CPU 5 a 5 2p b 5 2p a b 300 600 900 1200 1500 886.41 912.42 914.24 1175.01 MIN: 851.09 / MAX: 900.44 MIN: 878.63 / MAX: 988.74 MIN: 797.83 / MAX: 966.84 MIN: 982.69 / MAX: 1202.21 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Face Detection FP16-INT8 - Device: CPU 5 2p b 5 2p a b 5 a 8 16 24 32 40 34.91 34.83 27.03 17.92 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Machine Translation EN To DE FP16 - Device: CPU 5 a 5 2p b 5 2p a b 50 100 150 200 250 172.53 180.91 182.10 230.27 MIN: 81.35 / MAX: 207.95 MIN: 124.49 / MAX: 288.01 MIN: 117.14 / MAX: 548.18 MIN: 166.99 / MAX: 311.89 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Machine Translation EN To DE FP16 - Device: CPU 5 2p b 5 2p a b 5 a 40 80 120 160 200 176.76 175.48 138.84 92.61 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
Zstd Compression This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 3, Long Mode - Decompression Speed 5 a 5 2p b 5 2p a b 300 600 900 1200 1500 1361.5 1336.5 1332.3 1300.8 1. (CC) gcc options: -O3 -pthread -lz -llzma -llz4
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 3, Long Mode - Compression Speed 5 a 5 2p a b 5 2p b 200 400 600 800 1000 854.2 808.5 752.6 741.5 1. (CC) gcc options: -O3 -pthread -lz -llzma -llz4
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 8, Long Mode - Decompression Speed 5 a 5 2p a 5 2p b b 300 600 900 1200 1500 1460.1 1454.7 1445.1 1434.2 1. (CC) gcc options: -O3 -pthread -lz -llzma -llz4
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 8, Long Mode - Compression Speed 5 a b 5 2p a 5 2p b 200 400 600 800 1000 792.1 744.5 702.8 702.2 1. (CC) gcc options: -O3 -pthread -lz -llzma -llz4
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 12 - Decompression Speed 5 2p b 5 2p a 5 a b 300 600 900 1200 1500 1498.9 1482.7 1478.1 1445.9 1. (CC) gcc options: -O3 -pthread -lz -llzma -llz4
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 8 - Decompression Speed 5 2p b 5 a 5 2p a b 300 600 900 1200 1500 1441.1 1438.8 1437.5 1390.0 1. (CC) gcc options: -O3 -pthread -lz -llzma -llz4
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 3 - Decompression Speed 5 a 5 2p b 5 2p a b 300 600 900 1200 1500 1326.5 1309.5 1306.7 1277.3 1. (CC) gcc options: -O3 -pthread -lz -llzma -llz4
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 3 - Compression Speed 5 a b 5 2p a 5 2p b 600 1200 1800 2400 3000 2856.6 2819.0 2783.7 2775.9 1. (CC) gcc options: -O3 -pthread -lz -llzma -llz4
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.org ms, Fewer Is Better OpenVINO 2022.3 Model: Person Vehicle Bike Detection FP16 - Device: CPU 5 2p a 5 2p b 5 a b 4 8 12 16 20 12.66 12.66 12.77 16.93 MIN: 7.58 / MAX: 53.45 MIN: 8.38 / MAX: 48.88 MIN: 7.39 / MAX: 23.62 MIN: 13.88 / MAX: 33.11 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Person Vehicle Bike Detection FP16 - Device: CPU 5 2p b 5 2p a b 5 a 500 1000 1500 2000 2500 2524.05 2523.57 1888.20 1252.05 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Vehicle Detection FP16-INT8 - Device: CPU 5 2p a 5 2p b 5 a b 4 8 12 16 20 12.76 12.79 12.84 16.93 MIN: 7.65 / MAX: 43.58 MIN: 7.68 / MAX: 43.18 MIN: 6.96 / MAX: 23.28 MIN: 10.73 / MAX: 31.24 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Vehicle Detection FP16-INT8 - Device: CPU 5 2p a 5 2p b b 5 a 500 1000 1500 2000 2500 2504.36 2498.74 1889.05 1244.86 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Vehicle Detection FP16 - Device: CPU 5 a 5 2p a 5 2p b b 6 12 18 24 30 19.00 19.58 19.69 26.54 MIN: 13.62 / MAX: 32.75 MIN: 11.25 / MAX: 73.87 MIN: 11.03 / MAX: 75.83 MIN: 14.19 / MAX: 63.19 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Vehicle Detection FP16 - Device: CPU 5 2p a 5 2p b b 5 a 400 800 1200 1600 2000 1632.81 1623.83 1205.11 841.49 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Weld Porosity Detection FP16 - Device: CPU 5 a 5 2p a 5 2p b b 6 12 18 24 30 19.45 20.06 20.07 26.68 MIN: 11.98 / MAX: 28.4 MIN: 11.41 / MAX: 47.48 MIN: 13.26 / MAX: 75.83 MIN: 15.32 / MAX: 46.54 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Weld Porosity Detection FP16 - Device: CPU 5 2p a 5 2p b b 5 a 300 600 900 1200 1500 1594.14 1592.71 1198.45 822.19 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Weld Porosity Detection FP16-INT8 - Device: CPU 5 a 5 2p b 5 2p a b 6 12 18 24 30 18.10 18.25 18.26 23.92 MIN: 9.23 / MAX: 28.11 MIN: 8.84 / MAX: 40.73 MIN: 10.53 / MAX: 60.19 MIN: 15.02 / MAX: 35.74 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Weld Porosity Detection FP16-INT8 - Device: CPU 5 2p b 5 2p a b 5 a 800 1600 2400 3200 4000 3504.15 3501.69 2674.29 1767.13 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU 5 a 5 2p a 5 2p b b 0.3915 0.783 1.1745 1.566 1.9575 1.26 1.37 1.40 1.74 MIN: 0.69 / MAX: 12.96 MIN: 0.67 / MAX: 28.86 MIN: 0.68 / MAX: 42.1 MIN: 0.84 / MAX: 14.57 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU 5 2p a 5 2p b b 5 a 8K 16K 24K 32K 40K 39455.86 38800.34 35737.82 25209.62 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU 5 a 5 2p b b 0.3645 0.729 1.0935 1.458 1.8225 1.16 1.33 1.62 MIN: 0.66 / MAX: 12.23 MIN: 0.64 / MAX: 26.34 MIN: 0.69 / MAX: 13.34 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU 5 2p b b 5 a 9K 18K 27K 36K 45K 40352.48 38091.46 27340.96 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenCV This is a benchmark of the OpenCV (Computer Vision) library's built-in performance tests. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better OpenCV 4.7 Test: DNN - Deep Neural Network b 5 a 5 2p b 20K 40K 60K 80K 100K 39756 39997 87133 1. (CXX) g++ options: -fPIC -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -shared
VVenC VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better VVenC 1.8 Video Input: Bosphorus 1080p - Video Preset: Fast 5 b 5 a a b 5 2p b 5 2p a 4 8 12 16 20 SE +/- 0.10, N = 3 17.25 16.85 16.55 16.37 15.04 14.81 1. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects
Blender Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.5 Blend File: Fishy Cat - Compute: CPU-Only 5 2p b 5 2p a b a 5 a 12 24 36 48 60 SE +/- 0.04, N = 3 27.90 27.96 34.61 34.77 52.68
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.5 Encoder Mode: Preset 4 - Input: Bosphorus 4K a 5 2p b b 5 2p a 5 b 5 a 1.0942 2.1884 3.2826 4.3768 5.471 SE +/- 0.021, N = 3 4.863 4.842 4.826 4.784 4.760 4.743 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenCV This is a benchmark of the OpenCV (Computer Vision) library's built-in performance tests. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better OpenCV 4.7 Test: Object Detection 5 a b 5 2p b 20K 40K 60K 80K 100K 27739 31910 88553 1. (CXX) g++ options: -fPIC -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -shared
SPECFEM3D simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.0 Model: Layered Halfspace 5 2p a 5 2p b b a 5 b 5 a 11 22 33 44 55 SE +/- 0.29, N = 3 25.22 25.47 30.12 30.52 48.74 49.14 1. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
GROMACS The GROMACS (GROningen MAchine for Chemical Simulations) molecular dynamics package testing with the water_GMX50 data. This test profile allows selecting between CPU and GPU-based GROMACS builds. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Ns Per Day, More Is Better GROMACS 2023 Implementation: MPI CPU - Input: water_GMX50_bare 5 2p a 5 2p b b a 5 a 2 4 6 8 10 SE +/- 0.026, N = 3 8.285 8.222 7.299 7.290 5.022 1. (CXX) g++ options: -O3
SPECFEM3D simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.0 Model: Water-layered Halfspace 5 2p b 5 2p a b a 5 b 5 a 10 20 30 40 50 SE +/- 0.13, N = 3 23.27 23.94 27.47 28.08 44.33 45.71 1. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
Blender Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.5 Blend File: BMW27 - Compute: CPU-Only 5 2p b 5 2p a b a 5 a 10 20 30 40 50 SE +/- 0.03, N = 3 22.51 22.59 27.60 27.61 42.52
OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: bcrypt 5 2p a 5 2p b b a 5 a 5 b 30K 60K 90K 120K 150K SE +/- 99.80, N = 3 118502 117542 91276 86230 60192 60152 1. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2
OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: Blowfish 5 2p a 5 2p b b a 5 b 5 a 30K 60K 90K 120K 150K SE +/- 185.66, N = 3 118579 117771 87360 86460 60345 60325 1. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2
QuantLib QuantLib is an open-source library/framework around quantitative finance for modeling, trading and risk management scenarios. QuantLib is written in C++ with Boost and its built-in benchmark used reports the QuantLib Benchmark Index benchmark score. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MFLOPS, More Is Better QuantLib 1.30 5 2p b 5 b 5 2p a 5 a a b 600 1200 1800 2400 3000 SE +/- 25.52, N = 3 2850.8 2820.4 2760.9 2756.9 2746.5 2692.4 1. (CXX) g++ options: -O3 -march=native -fPIE -pie
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.5 Encoder Mode: Preset 8 - Input: Bosphorus 4K a 5 b 5 2p b b 5 a 5 2p a 16 32 48 64 80 SE +/- 0.56, N = 12 73.44 70.86 69.54 67.55 65.34 62.84 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
VVenC VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better VVenC 1.8 Video Input: Bosphorus 1080p - Video Preset: Faster 5 a 5 b b a 5 2p a 5 2p b 7 14 21 28 35 SE +/- 0.03, N = 3 31.84 31.80 30.11 30.01 24.49 24.27 1. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects
Embree OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.0.1 Binary: Pathtracer - Model: Asian Dragon Obj a 5 2p b 5 2p a 5 b 5 a 16 32 48 64 80 SE +/- 0.08, N = 3 70.52 68.70 68.69 44.19 43.96 MIN: 69.67 / MAX: 71.94 MIN: 67.96 / MAX: 69.64 MIN: 68.03 / MAX: 70.15 MIN: 43.94 / MAX: 44.51 MIN: 43.73 / MAX: 44.3
OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.0.1 Binary: Pathtracer ISPC - Model: Asian Dragon Obj b a 5 2p a 5 2p b 5 b 5 a 14 28 42 56 70 SE +/- 0.03, N = 3 64.64 64.51 64.41 64.04 39.22 39.19 MIN: 64.13 / MAX: 65.46 MIN: 63.96 / MAX: 66.18 MIN: 63.78 / MAX: 65.49 MIN: 63.29 / MAX: 64.92 MIN: 39 / MAX: 39.63 MIN: 38.97 / MAX: 39.59
SPECFEM3D simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.0 Model: Homogeneous Halfspace 5 2p b 5 2p a b a 5 a 5 b 6 12 18 24 30 SE +/- 0.20, N = 3 14.42 14.49 16.82 17.21 24.99 25.09 1. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.5 Encoder Mode: Preset 4 - Input: Bosphorus 1080p 5 b 5 a 5 2p b 5 2p a a b 3 6 9 12 15 SE +/- 0.02, N = 3 10.83 10.76 10.50 10.41 10.30 10.23 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
SPECFEM3D simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.0 Model: Tomographic Model 5 2p a 5 2p b b a 5 a 5 b 5 10 15 20 25 SE +/- 0.15, N = 3 11.22 11.80 13.53 13.86 19.09 19.63 1. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
oneDNN This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.1 Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU 5 a b 5 2p b 5 2p a 3 6 9 12 15 4.49565 6.92145 10.36850 10.52210 MIN: 3.9 MIN: 6.32 MIN: 8.54 MIN: 8.88 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
ASKAP ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Mpix/sec, More Is Better ASKAP 1.0 Test: tConvolve MPI - Gridding 5 2p b 5 2p a b 5 a 11K 22K 33K 44K 55K 51199.1 49980.1 36827.5 24990.1 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
OpenBenchmarking.org Mpix/sec, More Is Better ASKAP 1.0 Test: tConvolve MPI - Degridding 5 2p b 5 2p a b 5 a 9K 18K 27K 36K 45K 41983.3 41983.3 32799.5 20991.7 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
SPECFEM3D simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.0 Model: Mount St. Helens 5 2p b 5 2p a a b 5 a 5 b 5 10 15 20 25 SE +/- 0.038048290, N = 3 9.594769350 9.609453868 11.678414616 11.753321460 18.862199070 19.355553092 1. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
Embree OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.0.1 Binary: Pathtracer ISPC - Model: Crown 5 2p a 5 2p b a b 5 a 5 b 16 32 48 64 80 SE +/- 0.11, N = 3 73.33 73.26 63.37 63.36 40.24 40.12 MIN: 71.99 / MAX: 75.06 MIN: 72.13 / MAX: 74.82 MIN: 62.18 / MAX: 66.58 MIN: 62.44 / MAX: 66.52 MIN: 39.64 / MAX: 40.95 MIN: 39.64 / MAX: 40.56
Xcompact3d Incompact3d Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 Input: input.i3d 193 Cells Per Direction 5 2p a 5 2p b b 5 a 5 10 15 20 25 10.62 10.76 17.21 19.63 1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
Embree OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.0.1 Binary: Pathtracer - Model: Crown 5 2p a 5 2p b a 5 b 5 a 20 40 60 80 100 SE +/- 0.05, N = 3 81.45 81.09 69.11 43.98 43.95 MIN: 80.49 / MAX: 83.06 MIN: 80.03 / MAX: 82.56 MIN: 68.18 / MAX: 71.75 MIN: 43.29 / MAX: 44.67 MIN: 43.49 / MAX: 44.42
oneDNN This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.1 Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU b 5 a 5 2p a 5 2p b 0.649 1.298 1.947 2.596 3.245 1.26107 1.37996 2.67801 2.88451 MIN: 1.06 MIN: 1.25 MIN: 1.89 MIN: 1.79 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
Embree OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.0.1 Binary: Pathtracer - Model: Asian Dragon a 5 2p a 5 2p b 5 a 5 b 20 40 60 80 100 SE +/- 0.14, N = 3 78.48 76.81 76.63 48.52 48.51 MIN: 77.64 / MAX: 80.46 MIN: 76.07 / MAX: 77.73 MIN: 75.97 / MAX: 78.17 MIN: 48.32 / MAX: 48.91 MIN: 48.29 / MAX: 48.79
OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.0.1 Binary: Pathtracer ISPC - Model: Asian Dragon 5 2p a a b 5 2p b 5 b 5 a 16 32 48 64 80 SE +/- 0.01, N = 3 73.96 73.84 73.83 73.78 45.11 44.91 MIN: 73.12 / MAX: 75.27 MIN: 73.31 / MAX: 75.9 MIN: 73.34 / MAX: 74.66 MIN: 72.91 / MAX: 74.82 MIN: 44.88 / MAX: 45.48 MIN: 44.68 / MAX: 45.26
ASKAP ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Iterations Per Second, More Is Better ASKAP 1.0 Test: Hogbom Clean OpenMP 5 a b 5 2p a 5 2p b 200 400 600 800 1000 869.57 806.45 436.68 434.78 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
CloverLeaf CloverLeaf is a Lagrangian-Eulerian hydrodynamics benchmark. This test profile currently makes use of CloverLeaf's OpenMP version and benchmarked with the clover_bm.in input file (Problem 5). Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better CloverLeaf Lagrangian-Eulerian Hydrodynamics b 5 a 5 2p b 5 2p a 4 8 12 16 20 11.34 12.09 15.51 15.67 1. (F9X) gfortran options: -O3 -march=native -funroll-loops -fopenmp
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.5 Encoder Mode: Preset 8 - Input: Bosphorus 1080p a b 5 b 5 a 5 2p a 5 2p b 20 40 60 80 100 SE +/- 1.23, N = 5 109.98 109.71 108.36 107.38 107.10 104.17 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
Google Draco Draco is a library developed by Google for compressing/decompressing 3D geometric meshes and point clouds. This test profile uses some Artec3D PLY models as the sample 3D model input formats for Draco compression/decompression. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better Google Draco 1.5.6 Model: Church Facade 5 2p a 5 a 5 2p b a b 1300 2600 3900 5200 6500 SE +/- 3.21, N = 3 5682 5718 5729 5793 5833 1. (CXX) g++ options: -O3
oneDNN This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.1 Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU 5 a 5 2p a 5 2p b b 0.2804 0.5608 0.8412 1.1216 1.402 0.625930 0.911806 0.917226 1.246180 MIN: 0.57 MIN: 0.78 MIN: 0.77 MIN: 1.13 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
ASKAP ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Million Grid Points Per Second, More Is Better ASKAP 1.0 Test: tConvolve OpenMP - Degridding 5 a b 5 2p a 5 2p b 4K 8K 12K 16K 20K 19018.3 19018.3 16641.0 15662.1 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
OpenBenchmarking.org Million Grid Points Per Second, More Is Better ASKAP 1.0 Test: tConvolve OpenMP - Gridding b 5 a 5 2p b 5 2p a 4K 8K 12K 16K 20K 20481.2 17750.4 12678.9 12102.5 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
Google Draco Draco is a library developed by Google for compressing/decompressing 3D geometric meshes and point clouds. This test profile uses some Artec3D PLY models as the sample 3D model input formats for Draco compression/decompression. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better Google Draco 1.5.6 Model: Lion 5 2p a 5 a 5 2p b a b 1000 2000 3000 4000 5000 SE +/- 7.02, N = 3 4734 4760 4818 4867 4885 1. (CXX) g++ options: -O3
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.5 Encoder Mode: Preset 12 - Input: Bosphorus 4K 5 b a 5 a b 5 2p b 5 2p a 50 100 150 200 250 SE +/- 0.73, N = 3 226.73 222.84 220.81 214.93 176.11 169.11 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.5 Encoder Mode: Preset 13 - Input: Bosphorus 4K 5 a b 5 b a 5 2p b 5 2p a 40 80 120 160 200 SE +/- 0.70, N = 3 199.69 198.14 196.06 195.92 172.57 169.66 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
oneDNN This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.1 Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU 5 2p b 5 2p a b 5 a 0.258 0.516 0.774 1.032 1.29 0.669349 0.674336 0.827622 1.146560 MIN: 0.6 MIN: 0.62 MIN: 0.78 MIN: 1.09 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
Xcompact3d Incompact3d Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 Input: input.i3d 129 Cells Per Direction 5 2p b 5 2p a b 5 a 1.1402 2.2804 3.4206 4.5608 5.701 2.48386908 2.48654294 4.43804121 5.06756401 1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
oneDNN This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.1 Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU 5 2p a 5 2p b 5 a b 0.7031 1.4062 2.1093 2.8124 3.5155 1.85167 1.89859 2.76234 3.12487 MIN: 1.61 MIN: 1.58 MIN: 2.59 MIN: 2.06 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.5 Encoder Mode: Preset 12 - Input: Bosphorus 1080p 5 a 5 b a b 5 2p b 5 2p a 140 280 420 560 700 SE +/- 2.97, N = 3 629.30 622.73 604.84 601.98 565.06 540.73 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.5 Encoder Mode: Preset 13 - Input: Bosphorus 1080p 5 a 5 b a b 5 2p b 5 2p a 120 240 360 480 600 SE +/- 5.77, N = 3 571.08 560.57 552.24 549.83 546.03 525.15 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
a Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa001229Python Notes: Python 3.10.6Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 2 May 2023 05:24 by user phoronix.
b Processor: AMD EPYC 7773X 64-Core @ 2.20GHz (64 Cores / 128 Threads), Motherboard: AMD DAYTONA_X (RYM1009B BIOS), Chipset: AMD Starship/Matisse, Memory: 256GB, Disk: 3841GB Micron_9300_MTFDHAL3T8TDP, Graphics: ASPEED, Monitor: VE228, Network: 2 x Mellanox MT27710
OS: Ubuntu 22.04, Kernel: 5.15.0-47-generic (x86_64), Desktop: GNOME Shell 42.4, Display Server: X Server 1.21.1.3, Vulkan: 1.2.204, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 1920x1080
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa001229Python Notes: Python 3.10.6Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 2 May 2023 08:49 by user phoronix.
5 a Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa001229Python Notes: Python 3.10.6Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 4 May 2023 05:23 by user phoronix.
5 b Processor: AMD EPYC 7573X 32-Core @ 2.80GHz (32 Cores / 64 Threads), Motherboard: AMD DAYTONA_X (RYM1009B BIOS), Chipset: AMD Starship/Matisse, Memory: 256GB, Disk: 3841GB Micron_9300_MTFDHAL3T8TDP, Graphics: ASPEED, Monitor: VE228, Network: 2 x Mellanox MT27710
OS: Ubuntu 22.04, Kernel: 5.15.0-47-generic (x86_64), Desktop: GNOME Shell 42.4, Display Server: X Server 1.21.1.3, Vulkan: 1.2.204, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 1920x1080
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa001229Python Notes: Python 3.10.6Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 4 May 2023 09:24 by user phoronix.
5 2p a Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa001229Python Notes: Python 3.10.6Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 4 May 2023 10:34 by user phoronix.
5 2p b Processor: 2 x AMD EPYC 7573X 32-Core @ 2.80GHz (64 Cores / 128 Threads), Motherboard: AMD DAYTONA_X (RYM1009B BIOS), Chipset: AMD Starship/Matisse, Memory: 512GB, Disk: 3841GB Micron_9300_MTFDHAL3T8TDP, Graphics: ASPEED, Monitor: VE228, Network: 2 x Mellanox MT27710
OS: Ubuntu 22.04, Kernel: 5.15.0-47-generic (x86_64), Desktop: GNOME Shell 42.4, Display Server: X Server 1.21.1.3, Vulkan: 1.2.204, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 1920x1080
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa001229Python Notes: Python 3.10.6Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 4 May 2023 12:32 by user phoronix.