TR 2970WX NN + More AMD Ryzen Threadripper 2970WX 24-Core testing with a Gigabyte X399 AORUS Gaming 7 (F12h BIOS) and Sapphire AMD Radeon RX 550 640SP / 560/560X 4GB on Ubuntu 20.04 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2009264-FI-TR2970WXN17&sro .
TR 2970WX NN + More Processor Motherboard Chipset Memory Disk Graphics Audio Monitor Network OS Kernel Desktop Display Server Display Driver OpenGL Compiler File-System Screen Resolution Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat AMD Ryzen Threadripper 2970WX 24-Core @ 3.00GHz (24 Cores / 48 Threads) Gigabyte X399 AORUS Gaming 7 (F12h BIOS) AMD 17h 16GB 120GB Corsair Force MP500 Sapphire AMD Radeon RX 550 640SP / 560/560X 4GB (1300/1750MHz) Realtek ALC1220 DELL S2409W Qualcomm Atheros Killer E2500 + 2 x QLogic cLOM8214 1/10GbE + Intel 8265 / 8275 Ubuntu 20.04 5.8.0-050800-generic (x86_64) GNOME Shell 3.36.4 X Server 1.20.8 modesetting 1.20.8 4.6 Mesa 20.0.8 (LLVM 10.0.0) GCC 9.3.0 ext4 1920x1080 5.9.0-050900rc6daily20200926-generic (x86_64) 20200925 OpenBenchmarking.org Compiler Details - --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,gm2 --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Processor Details - Scaling Governor: acpi-cpufreq ondemand - CPU Microcode: 0x800820d Graphics Details - GLAMOR Security Details - itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional STIBP: disabled RSB filling + srbds: Not affected + tsx_async_abort: Not affected
TR 2970WX NN + More realsr-ncnn: 4x - No realsr-ncnn: 4x - Yes osbench: Create Files osbench: Create Threads osbench: Launch Programs osbench: Create Processes osbench: Memory Allocations webp: Default webp: Quality 100 webp: Quality 100, Lossless webp: Quality 100, Highest Compression webp: Quality 100, Lossless, Highest Compression libraw: Post-Processing Benchmark espeak: Text-To-Speech Synthesis couchdb: 100 - 1000 - 24 ncnn: CPU - squeezenet ncnn: CPU - mobilenet ncnn: CPU-v2-v2 - mobilenet-v2 ncnn: CPU-v3-v3 - mobilenet-v3 ncnn: CPU - shufflenet-v2 ncnn: CPU - mnasnet ncnn: CPU - efficientnet-b0 ncnn: CPU - blazeface ncnn: CPU - googlenet ncnn: CPU - vgg16 ncnn: CPU - resnet18 ncnn: CPU - alexnet ncnn: CPU - resnet50 ncnn: CPU - yolov4-tiny ncnn: Vulkan GPU - squeezenet ncnn: Vulkan GPU - mobilenet ncnn: Vulkan GPU-v2-v2 - mobilenet-v2 ncnn: Vulkan GPU-v3-v3 - mobilenet-v3 ncnn: Vulkan GPU - shufflenet-v2 ncnn: Vulkan GPU - mnasnet ncnn: Vulkan GPU - efficientnet-b0 ncnn: Vulkan GPU - blazeface ncnn: Vulkan GPU - googlenet ncnn: Vulkan GPU - vgg16 ncnn: Vulkan GPU - resnet18 ncnn: Vulkan GPU - alexnet ncnn: Vulkan GPU - resnet50 ncnn: Vulkan GPU - yolov4-tiny tnn: CPU - MobileNet v2 tnn: CPU - SqueezeNet v1.1 influxdb: 4 - 10000 - 2,5000,1 - 10000 influxdb: 64 - 10000 - 2,5000,1 - 10000 influxdb: 1024 - 10000 - 2,5000,1 - 10000 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 124.758 974.000 15.492504 16.332467 64.817270 38.786729 79.048316 1.610 2.457 18.759 7.775 38.421 30.82 34.343 122.375 30.66 31.18 13.69 12.92 13.05 12.08 16.51 5.01 36.91 91.54 36.96 38.52 60.11 52.34 28.97 22.16 9.46 10.97 6.44 10.21 20.82 2.17 27.05 134.01 19.80 20.23 45.96 44.18 294.939 258.424 1020198.3 1245572.3 1266364.3 124.359 974.343 15.940509 17.846107 65.773130 40.818850 77.799638 1.626 2.445 18.599 7.732 39.254 30.77 33.421 124.279 34.10 32.25 14.06 13.33 13.48 12.18 18.58 5.11 42.47 95.74 52.28 37.00 71.67 51.08 28.60 22.09 9.45 10.96 6.46 10.16 20.83 2.27 27.20 133.85 19.65 20.28 45.66 44.16 295.263 259.009 996638.1 1243255.8 1256621.9 124.383 974.386 16.183030 16.439914 66.293081 40.656725 78.763008 1.607 2.483 18.526 7.754 38.578 30.90 33.273 124.452 32.95 32.87 16.22 13.43 13.95 11.99 17.50 4.95 39.01 99.70 58.59 35.66 65.65 51.32 28.35 22.05 9.46 10.97 6.52 10.28 20.51 2.15 26.90 133.74 19.75 20.59 45.84 44.06 295.805 258.216 995507.3 1238007.3 1251920.4 OpenBenchmarking.org
RealSR-NCNN Scale: 4x - TAA: No OpenBenchmarking.org Seconds, Fewer Is Better RealSR-NCNN 20200818 Scale: 4x - TAA: No Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 30 60 90 120 150 SE +/- 0.21, N = 3 SE +/- 0.06, N = 3 SE +/- 0.10, N = 3 124.76 124.36 124.38
RealSR-NCNN Scale: 4x - TAA: Yes OpenBenchmarking.org Seconds, Fewer Is Better RealSR-NCNN 20200818 Scale: 4x - TAA: Yes Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 200 400 600 800 1000 SE +/- 0.20, N = 3 SE +/- 0.03, N = 3 SE +/- 0.13, N = 3 974.00 974.34 974.39
OSBench Test: Create Files OpenBenchmarking.org us Per Event, Fewer Is Better OSBench Test: Create Files Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 4 8 12 16 20 SE +/- 0.05, N = 3 SE +/- 0.05, N = 3 SE +/- 0.15, N = 3 15.49 15.94 16.18 1. (CC) gcc options: -lm
OSBench Test: Create Threads OpenBenchmarking.org us Per Event, Fewer Is Better OSBench Test: Create Threads Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 4 8 12 16 20 SE +/- 0.17, N = 3 SE +/- 0.16, N = 15 SE +/- 0.20, N = 5 16.33 17.85 16.44 1. (CC) gcc options: -lm
OSBench Test: Launch Programs OpenBenchmarking.org us Per Event, Fewer Is Better OSBench Test: Launch Programs Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 15 30 45 60 75 SE +/- 0.33, N = 3 SE +/- 0.95, N = 4 SE +/- 0.15, N = 3 64.82 65.77 66.29 1. (CC) gcc options: -lm
OSBench Test: Create Processes OpenBenchmarking.org us Per Event, Fewer Is Better OSBench Test: Create Processes Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 9 18 27 36 45 SE +/- 0.08, N = 3 SE +/- 0.56, N = 3 SE +/- 0.61, N = 3 38.79 40.82 40.66 1. (CC) gcc options: -lm
OSBench Test: Memory Allocations OpenBenchmarking.org Ns Per Event, Fewer Is Better OSBench Test: Memory Allocations Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 20 40 60 80 100 SE +/- 0.21, N = 3 SE +/- 0.13, N = 3 SE +/- 0.03, N = 3 79.05 77.80 78.76 1. (CC) gcc options: -lm
WebP Image Encode Encode Settings: Default OpenBenchmarking.org Encode Time - Seconds, Fewer Is Better WebP Image Encode 1.1 Encode Settings: Default Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 0.3659 0.7318 1.0977 1.4636 1.8295 SE +/- 0.002, N = 3 SE +/- 0.001, N = 3 SE +/- 0.000, N = 3 1.610 1.626 1.607 1. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff
WebP Image Encode Encode Settings: Quality 100 OpenBenchmarking.org Encode Time - Seconds, Fewer Is Better WebP Image Encode 1.1 Encode Settings: Quality 100 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 0.5587 1.1174 1.6761 2.2348 2.7935 SE +/- 0.011, N = 3 SE +/- 0.007, N = 3 SE +/- 0.013, N = 3 2.457 2.445 2.483 1. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff
WebP Image Encode Encode Settings: Quality 100, Lossless OpenBenchmarking.org Encode Time - Seconds, Fewer Is Better WebP Image Encode 1.1 Encode Settings: Quality 100, Lossless Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 5 10 15 20 25 SE +/- 0.03, N = 3 SE +/- 0.26, N = 4 SE +/- 0.23, N = 3 18.76 18.60 18.53 1. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff
WebP Image Encode Encode Settings: Quality 100, Highest Compression OpenBenchmarking.org Encode Time - Seconds, Fewer Is Better WebP Image Encode 1.1 Encode Settings: Quality 100, Highest Compression Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 2 4 6 8 10 SE +/- 0.039, N = 3 SE +/- 0.021, N = 3 SE +/- 0.024, N = 3 7.775 7.732 7.754 1. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff
WebP Image Encode Encode Settings: Quality 100, Lossless, Highest Compression OpenBenchmarking.org Encode Time - Seconds, Fewer Is Better WebP Image Encode 1.1 Encode Settings: Quality 100, Lossless, Highest Compression Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 9 18 27 36 45 SE +/- 0.13, N = 3 SE +/- 0.11, N = 3 SE +/- 0.17, N = 3 38.42 39.25 38.58 1. (CC) gcc options: -fvisibility=hidden -O2 -pthread -lm -ljpeg -lpng16 -ltiff
LibRaw Post-Processing Benchmark OpenBenchmarking.org Mpix/sec, More Is Better LibRaw 0.20 Post-Processing Benchmark Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 7 14 21 28 35 SE +/- 0.03, N = 3 SE +/- 0.12, N = 3 SE +/- 0.04, N = 3 30.82 30.77 30.90 1. (CXX) g++ options: -O2 -fopenmp -ljpeg -lz -lm
eSpeak-NG Speech Engine Text-To-Speech Synthesis OpenBenchmarking.org Seconds, Fewer Is Better eSpeak-NG Speech Engine 20200907 Text-To-Speech Synthesis Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 8 16 24 32 40 SE +/- 0.53, N = 20 SE +/- 0.48, N = 20 SE +/- 0.55, N = 20 34.34 33.42 33.27 1. (CC) gcc options: -O2 -std=c99
Apache CouchDB Bulk Size: 100 - Inserts: 1000 - Rounds: 24 OpenBenchmarking.org Seconds, Fewer Is Better Apache CouchDB 3.1.1 Bulk Size: 100 - Inserts: 1000 - Rounds: 24 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 30 60 90 120 150 SE +/- 0.88, N = 3 SE +/- 1.07, N = 3 SE +/- 1.77, N = 3 122.38 124.28 124.45 1. (CXX) g++ options: -std=c++14 -lmozjs-68 -lm -lerl_interface -lei -fPIC -MMD
NCNN Target: CPU - Model: squeezenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: squeezenet Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 8 16 24 32 40 SE +/- 0.44, N = 3 SE +/- 1.65, N = 12 SE +/- 0.99, N = 12 30.66 34.10 32.95 MIN: 26.97 / MAX: 316.95 MIN: 26.75 / MAX: 425.49 MIN: 26.72 / MAX: 456.36 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU - Model: mobilenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: mobilenet Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 8 16 24 32 40 SE +/- 0.47, N = 3 SE +/- 0.80, N = 12 SE +/- 0.98, N = 12 31.18 32.25 32.87 MIN: 28.88 / MAX: 240.65 MIN: 28.41 / MAX: 398.09 MIN: 28.65 / MAX: 439.38 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU-v2-v2 - Model: mobilenet-v2 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU-v2-v2 - Model: mobilenet-v2 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 4 8 12 16 20 SE +/- 0.44, N = 3 SE +/- 0.22, N = 12 SE +/- 1.58, N = 12 13.69 14.06 16.22 MIN: 11.92 / MAX: 347.04 MIN: 11.96 / MAX: 335.89 MIN: 11.87 / MAX: 425.43 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU-v3-v3 - Model: mobilenet-v3 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU-v3-v3 - Model: mobilenet-v3 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 3 6 9 12 15 SE +/- 0.28, N = 3 SE +/- 0.30, N = 12 SE +/- 0.93, N = 12 12.92 13.33 13.43 MIN: 11.59 / MAX: 110.56 MIN: 11.55 / MAX: 394.82 MIN: 11.56 / MAX: 398.77 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU - Model: shufflenet-v2 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: shufflenet-v2 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 4 8 12 16 20 SE +/- 0.43, N = 3 SE +/- 0.22, N = 12 SE +/- 0.67, N = 12 13.05 13.48 13.95 MIN: 12.14 / MAX: 14.64 MIN: 12.18 / MAX: 390.55 MIN: 12.15 / MAX: 414.99 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU - Model: mnasnet OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: mnasnet Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 3 6 9 12 15 SE +/- 0.15, N = 3 SE +/- 0.12, N = 12 SE +/- 0.18, N = 12 12.08 12.18 11.99 MIN: 11.18 / MAX: 46.41 MIN: 11.24 / MAX: 110.31 MIN: 10.89 / MAX: 342.44 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU - Model: efficientnet-b0 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: efficientnet-b0 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 5 10 15 20 25 SE +/- 0.48, N = 3 SE +/- 1.28, N = 12 SE +/- 0.43, N = 12 16.51 18.58 17.50 MIN: 15.42 / MAX: 48.65 MIN: 15.35 / MAX: 510.96 MIN: 15.46 / MAX: 521.68 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU - Model: blazeface OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: blazeface Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 1.1498 2.2996 3.4494 4.5992 5.749 SE +/- 0.15, N = 3 SE +/- 0.07, N = 12 SE +/- 0.06, N = 12 5.01 5.11 4.95 MIN: 4.64 / MAX: 7.94 MIN: 4.64 / MAX: 32.58 MIN: 4.59 / MAX: 37.49 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU - Model: googlenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: googlenet Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 10 20 30 40 50 SE +/- 2.78, N = 3 SE +/- 3.05, N = 12 SE +/- 2.68, N = 12 36.91 42.47 39.01 MIN: 27.59 / MAX: 479.39 MIN: 27.92 / MAX: 504.41 MIN: 27.69 / MAX: 511.48 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU - Model: vgg16 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: vgg16 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 20 40 60 80 100 SE +/- 2.56, N = 3 SE +/- 2.26, N = 12 SE +/- 0.99, N = 12 91.54 95.74 99.70 MIN: 69.5 / MAX: 213.3 MIN: 63.79 / MAX: 220.6 MIN: 65.21 / MAX: 223.44 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU - Model: resnet18 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: resnet18 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 13 26 39 52 65 SE +/- 4.82, N = 3 SE +/- 6.08, N = 12 SE +/- 7.45, N = 12 36.96 52.28 58.59 MIN: 22.12 / MAX: 220.9 MIN: 21.12 / MAX: 215.72 MIN: 21.44 / MAX: 245.55 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU - Model: alexnet OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: alexnet Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 9 18 27 36 45 SE +/- 7.38, N = 3 SE +/- 2.20, N = 12 SE +/- 1.35, N = 12 38.52 37.00 35.66 MIN: 17.41 / MAX: 105.77 MIN: 16.59 / MAX: 112.15 MIN: 18.84 / MAX: 125.05 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU - Model: resnet50 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: resnet50 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 16 32 48 64 80 SE +/- 5.77, N = 3 SE +/- 5.77, N = 12 SE +/- 5.27, N = 12 60.11 71.67 65.65 MIN: 41.52 / MAX: 518.24 MIN: 38.92 / MAX: 530.97 MIN: 37.96 / MAX: 537.32 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: CPU - Model: yolov4-tiny OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: yolov4-tiny Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 12 24 36 48 60 SE +/- 0.83, N = 3 SE +/- 0.96, N = 12 SE +/- 0.79, N = 12 52.34 51.08 51.32 MIN: 44.27 / MAX: 207.57 MIN: 42.05 / MAX: 213.57 MIN: 42.04 / MAX: 214.86 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: squeezenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU - Model: squeezenet Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 7 14 21 28 35 SE +/- 0.24, N = 15 SE +/- 0.37, N = 4 SE +/- 0.16, N = 3 28.97 28.60 28.35 MIN: 26.62 / MAX: 36.99 MIN: 26.77 / MAX: 37.15 MIN: 27.06 / MAX: 36.04 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: mobilenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU - Model: mobilenet Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 5 10 15 20 25 SE +/- 0.02, N = 15 SE +/- 0.08, N = 4 SE +/- 0.05, N = 3 22.16 22.09 22.05 MIN: 20.83 / MAX: 28.17 MIN: 21.08 / MAX: 28.46 MIN: 21.01 / MAX: 23.75 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 3 6 9 12 15 SE +/- 0.01, N = 15 SE +/- 0.01, N = 4 SE +/- 0.01, N = 3 9.46 9.45 9.46 MIN: 8.62 / MAX: 14.48 MIN: 8.72 / MAX: 10.61 MIN: 8.58 / MAX: 10.96 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 3 6 9 12 15 SE +/- 0.02, N = 15 SE +/- 0.02, N = 4 SE +/- 0.02, N = 3 10.97 10.96 10.97 MIN: 9.94 / MAX: 15.31 MIN: 10.05 / MAX: 14.07 MIN: 9.98 / MAX: 14.14 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: shufflenet-v2 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU - Model: shufflenet-v2 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 2 4 6 8 10 SE +/- 0.01, N = 15 SE +/- 0.05, N = 4 SE +/- 0.08, N = 3 6.44 6.46 6.52 MIN: 5.62 / MAX: 9.35 MIN: 5.73 / MAX: 9.16 MIN: 5.75 / MAX: 8.72 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: mnasnet OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU - Model: mnasnet Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 3 6 9 12 15 SE +/- 0.03, N = 15 SE +/- 0.03, N = 4 SE +/- 0.16, N = 3 10.21 10.16 10.28 MIN: 9.55 / MAX: 16 MIN: 9.48 / MAX: 13.93 MIN: 9.54 / MAX: 14.65 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: efficientnet-b0 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU - Model: efficientnet-b0 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 5 10 15 20 25 SE +/- 0.11, N = 15 SE +/- 0.15, N = 4 SE +/- 0.01, N = 3 20.82 20.83 20.51 MIN: 18.95 / MAX: 26.73 MIN: 19.63 / MAX: 25.15 MIN: 19.52 / MAX: 23.32 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: blazeface OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU - Model: blazeface Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 0.5108 1.0216 1.5324 2.0432 2.554 SE +/- 0.03, N = 15 SE +/- 0.03, N = 4 SE +/- 0.03, N = 3 2.17 2.27 2.15 MIN: 1.83 / MAX: 5.86 MIN: 2.01 / MAX: 7.69 MIN: 1.82 / MAX: 3.74 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: googlenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU - Model: googlenet Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 6 12 18 24 30 SE +/- 0.04, N = 15 SE +/- 0.22, N = 4 SE +/- 0.07, N = 3 27.05 27.20 26.90 MIN: 25.23 / MAX: 33.38 MIN: 25.89 / MAX: 33.69 MIN: 26.12 / MAX: 39.16 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: vgg16 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU - Model: vgg16 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 30 60 90 120 150 SE +/- 0.09, N = 15 SE +/- 0.16, N = 4 SE +/- 0.19, N = 3 134.01 133.85 133.74 MIN: 132.1 / MAX: 142.92 MIN: 132.08 / MAX: 141.32 MIN: 132.16 / MAX: 140.12 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: resnet18 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU - Model: resnet18 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 5 10 15 20 25 SE +/- 0.09, N = 15 SE +/- 0.11, N = 4 SE +/- 0.13, N = 3 19.80 19.65 19.75 MIN: 18.98 / MAX: 28.54 MIN: 19.06 / MAX: 22.8 MIN: 19.07 / MAX: 22.67 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: alexnet OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU - Model: alexnet Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 5 10 15 20 25 SE +/- 0.05, N = 15 SE +/- 0.17, N = 4 SE +/- 0.45, N = 3 20.23 20.28 20.59 MIN: 19.47 / MAX: 28.79 MIN: 19.51 / MAX: 28.12 MIN: 19.38 / MAX: 28.43 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: resnet50 OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU - Model: resnet50 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 10 20 30 40 50 SE +/- 0.07, N = 15 SE +/- 0.04, N = 4 SE +/- 0.15, N = 3 45.96 45.66 45.84 MIN: 44.53 / MAX: 59.57 MIN: 44.45 / MAX: 51.45 MIN: 44.72 / MAX: 49.5 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: yolov4-tiny OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: Vulkan GPU - Model: yolov4-tiny Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 10 20 30 40 50 SE +/- 0.05, N = 15 SE +/- 0.06, N = 4 SE +/- 0.02, N = 3 44.18 44.16 44.06 MIN: 42.06 / MAX: 50.96 MIN: 41.67 / MAX: 48.98 MIN: 42.05 / MAX: 47.89 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
TNN Target: CPU - Model: MobileNet v2 OpenBenchmarking.org ms, Fewer Is Better TNN 0.2.3 Target: CPU - Model: MobileNet v2 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 60 120 180 240 300 SE +/- 1.00, N = 3 SE +/- 0.60, N = 3 SE +/- 0.44, N = 3 294.94 295.26 295.81 MIN: 278.83 / MAX: 315.53 MIN: 275.59 / MAX: 321.25 MIN: 275.48 / MAX: 334.45 1. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -O3 -rdynamic -ldl
TNN Target: CPU - Model: SqueezeNet v1.1 OpenBenchmarking.org ms, Fewer Is Better TNN 0.2.3 Target: CPU - Model: SqueezeNet v1.1 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 60 120 180 240 300 SE +/- 0.08, N = 3 SE +/- 0.77, N = 3 SE +/- 0.13, N = 3 258.42 259.01 258.22 MIN: 257.57 / MAX: 260.74 MIN: 257.05 / MAX: 264.71 MIN: 257.01 / MAX: 259.55 1. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -O3 -rdynamic -ldl
InfluxDB Concurrent Streams: 4 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 OpenBenchmarking.org val/sec, More Is Better InfluxDB 1.8.2 Concurrent Streams: 4 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 200K 400K 600K 800K 1000K SE +/- 2790.54, N = 3 SE +/- 12156.00, N = 3 SE +/- 7784.44, N = 3 1020198.3 996638.1 995507.3
InfluxDB Concurrent Streams: 64 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 OpenBenchmarking.org val/sec, More Is Better InfluxDB 1.8.2 Concurrent Streams: 64 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 300K 600K 900K 1200K 1500K SE +/- 4132.64, N = 3 SE +/- 2224.57, N = 3 SE +/- 2080.28, N = 3 1245572.3 1243255.8 1238007.3
InfluxDB Concurrent Streams: 1024 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 OpenBenchmarking.org val/sec, More Is Better InfluxDB 1.8.2 Concurrent Streams: 1024 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 Linux 5.8 Linux 5.9 Git Linux 5.9 Git Repeat 300K 600K 900K 1200K 1500K SE +/- 13092.34, N = 3 SE +/- 2116.83, N = 3 SE +/- 1710.08, N = 3 1266364.3 1256621.9 1251920.4
Phoronix Test Suite v10.8.4