NN TR 3960X + Radeon RX 5500 XT AMD Ryzen Threadripper 3960X 24-Core testing with a MSI Creator TRX40 (MS-7C59) v1.0 (1.12N1 BIOS) and Sapphire AMD Radeon RX 5500/5500M / Pro 5500M 4GB on Ubuntu 20.04 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2009245-PTS-NNTR396012&rdt .
NN TR 3960X + Radeon RX 5500 XT Processor Motherboard Chipset Memory Disk Graphics Audio Monitor Network OS Kernel Desktop Display Server Display Driver OpenGL Vulkan Compiler File-System Screen Resolution 1 2 3 AMD Ryzen Threadripper 3960X 24-Core @ 3.80GHz (24 Cores / 48 Threads) MSI Creator TRX40 (MS-7C59) v1.0 (1.12N1 BIOS) AMD Starship/Matisse 32GB 1000GB Sabrent Rocket 4.0 1TB Sapphire AMD Radeon RX 5500/5500M / Pro 5500M 4GB (1900/875MHz) AMD Navi 10 HDMI Audio ASUS MG28U Aquantia AQC107 NBase-T/IEEE + Intel I211 + Intel Wi-Fi 6 AX200 Ubuntu 20.04 5.9.0-rc5-14sep-patch (x86_64) 20200914 GNOME Shell 3.36.4 X Server 1.20.8 modesetting 1.20.8 4.6 Mesa 20.0.8 (LLVM 10.0.0) 1.2.128 GCC 9.3.0 ext4 3840x2160 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: 0x8301025 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: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected
NN TR 3960X + Radeon RX 5500 XT realsr-ncnn: 4x - No realsr-ncnn: 4x - Yes 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 1 2 3 25.178 183.307 34.16 16.99 7.22 6.74 7.13 6.52 8.57 2.90 39.53 69.58 20.25 11.43 56.97 27.38 36.30 9.59 4.37 8.43 3.20 4.44 12.06 0.89 41.21 86.05 20.82 28.16 58.79 21.28 257.861 237.816 25.135 183.344 33.51 16.93 7.19 6.71 7.15 6.49 8.57 2.88 39.27 69.73 19.80 11.71 56.35 28.00 37.23 9.30 4.35 8.22 3.20 4.45 11.98 0.90 41.37 70.43 20.92 28.60 58.78 21.30 257.621 237.236 25.184 183.380 34.17 16.93 7.25 6.90 7.15 6.50 8.64 2.89 41.22 71.61 21.01 11.88 58.16 28.21 36.78 9.54 4.36 8.21 3.22 4.44 12.11 0.90 41.08 69.45 20.60 28.45 58.06 21.35 257.427 237.826 OpenBenchmarking.org
RealSR-NCNN Scale: 4x - TAA: No OpenBenchmarking.org Seconds, Fewer Is Better RealSR-NCNN 20200818 Scale: 4x - TAA: No 1 2 3 6 12 18 24 30 SE +/- 0.07, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 25.18 25.14 25.18
RealSR-NCNN Scale: 4x - TAA: Yes OpenBenchmarking.org Seconds, Fewer Is Better RealSR-NCNN 20200818 Scale: 4x - TAA: Yes 1 2 3 40 80 120 160 200 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 183.31 183.34 183.38
NCNN Target: CPU - Model: squeezenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20200916 Target: CPU - Model: squeezenet 1 2 3 8 16 24 32 40 SE +/- 0.51, N = 3 SE +/- 0.24, N = 3 SE +/- 0.26, N = 3 34.16 33.51 34.17 MIN: 32.67 / MAX: 117.09 MIN: 32.21 / MAX: 92.06 MIN: 33.09 / MAX: 77.53 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 1 2 3 4 8 12 16 20 SE +/- 0.09, N = 3 SE +/- 0.03, N = 3 SE +/- 0.09, N = 3 16.99 16.93 16.93 MIN: 16.74 / MAX: 18.01 MIN: 16.66 / MAX: 17.45 MIN: 16.64 / MAX: 17.48 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 1 2 3 2 4 6 8 10 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.06, N = 3 7.22 7.19 7.25 MIN: 6.94 / MAX: 7.7 MIN: 6.97 / MAX: 7.73 MIN: 6.86 / MAX: 7.96 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 1 2 3 2 4 6 8 10 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.16, N = 3 6.74 6.71 6.90 MIN: 6.57 / MAX: 7.22 MIN: 6.56 / MAX: 7.43 MIN: 6.54 / MAX: 105.24 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 1 2 3 2 4 6 8 10 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 7.13 7.15 7.15 MIN: 6.84 / MAX: 7.57 MIN: 6.93 / MAX: 8.31 MIN: 6.91 / MAX: 8.35 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 1 2 3 2 4 6 8 10 SE +/- 0.03, N = 3 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 6.52 6.49 6.50 MIN: 6.35 / MAX: 8.85 MIN: 6.34 / MAX: 8.66 MIN: 6.23 / MAX: 6.91 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 1 2 3 2 4 6 8 10 SE +/- 0.05, N = 3 SE +/- 0.01, N = 3 SE +/- 0.06, N = 3 8.57 8.57 8.64 MIN: 8.35 / MAX: 9.02 MIN: 8.44 / MAX: 9.1 MIN: 8.43 / MAX: 9.13 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 1 2 3 0.6525 1.305 1.9575 2.61 3.2625 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 2.90 2.88 2.89 MIN: 2.76 / MAX: 3.07 MIN: 2.74 / MAX: 3.29 MIN: 2.77 / MAX: 3.02 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 1 2 3 9 18 27 36 45 SE +/- 0.06, N = 3 SE +/- 0.06, N = 3 SE +/- 0.14, N = 3 39.53 39.27 41.22 MIN: 38.19 / MAX: 97.58 MIN: 38.25 / MAX: 49.66 MIN: 40.07 / MAX: 42.85 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 1 2 3 16 32 48 64 80 SE +/- 0.37, N = 3 SE +/- 0.18, N = 3 SE +/- 0.33, N = 3 69.58 69.73 71.61 MIN: 67.57 / MAX: 82.43 MIN: 68.49 / MAX: 77.46 MIN: 69.64 / MAX: 129.25 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 1 2 3 5 10 15 20 25 SE +/- 0.52, N = 3 SE +/- 0.03, N = 3 SE +/- 0.17, N = 3 20.25 19.80 21.01 MIN: 19.51 / MAX: 24.44 MIN: 19.56 / MAX: 24.32 MIN: 20.51 / MAX: 33.69 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 1 2 3 3 6 9 12 15 SE +/- 0.08, N = 3 SE +/- 0.07, N = 3 SE +/- 0.04, N = 3 11.43 11.71 11.88 MIN: 11.22 / MAX: 12.3 MIN: 11.51 / MAX: 12.77 MIN: 11.76 / MAX: 12.7 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 1 2 3 13 26 39 52 65 SE +/- 0.80, N = 3 SE +/- 0.18, N = 3 SE +/- 0.67, N = 3 56.97 56.35 58.16 MIN: 55.76 / MAX: 116.12 MIN: 55.8 / MAX: 101.34 MIN: 56.46 / MAX: 72.94 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 1 2 3 7 14 21 28 35 SE +/- 0.21, N = 3 SE +/- 0.30, N = 3 SE +/- 0.13, N = 3 27.38 28.00 28.21 MIN: 26.82 / MAX: 28.17 MIN: 27.26 / MAX: 46.57 MIN: 27.77 / MAX: 32.51 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 1 2 3 9 18 27 36 45 SE +/- 0.40, N = 7 SE +/- 0.30, N = 3 SE +/- 0.18, N = 3 36.30 37.23 36.78 MIN: 34.06 / MAX: 93.7 MIN: 36.09 / MAX: 42.49 MIN: 34.68 / MAX: 102.95 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 1 2 3 3 6 9 12 15 SE +/- 0.24, N = 7 SE +/- 0.39, N = 3 SE +/- 0.47, N = 3 9.59 9.30 9.54 MIN: 6.09 / MAX: 36.77 MIN: 6.71 / MAX: 31.05 MIN: 6.58 / MAX: 28.71 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 1 2 3 0.9833 1.9666 2.9499 3.9332 4.9165 SE +/- 0.01, N = 7 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 4.37 4.35 4.36 MIN: 4.08 / MAX: 8.18 MIN: 4.18 / MAX: 5.04 MIN: 4.18 / MAX: 4.64 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 1 2 3 2 4 6 8 10 SE +/- 0.13, N = 7 SE +/- 0.15, N = 3 SE +/- 0.38, N = 3 8.43 8.22 8.21 MIN: 6.97 / MAX: 52 MIN: 6.98 / MAX: 50.24 MIN: 7 / MAX: 35.79 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 1 2 3 0.7245 1.449 2.1735 2.898 3.6225 SE +/- 0.00, N = 7 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 3.20 3.20 3.22 MIN: 3.14 / MAX: 3.47 MIN: 3.14 / MAX: 3.4 MIN: 3.14 / MAX: 5.74 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 1 2 3 1.0013 2.0026 3.0039 4.0052 5.0065 SE +/- 0.00, N = 7 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 4.44 4.45 4.44 MIN: 4.28 / MAX: 4.89 MIN: 4.3 / MAX: 4.78 MIN: 4.29 / MAX: 4.71 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 1 2 3 3 6 9 12 15 SE +/- 0.11, N = 7 SE +/- 0.03, N = 3 SE +/- 0.15, N = 3 12.06 11.98 12.11 MIN: 10.04 / MAX: 43.67 MIN: 10.04 / MAX: 38.23 MIN: 10.06 / MAX: 39.37 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 1 2 3 0.2025 0.405 0.6075 0.81 1.0125 SE +/- 0.00, N = 7 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 0.89 0.90 0.90 MIN: 0.87 / MAX: 1.19 MIN: 0.88 / MAX: 1.25 MIN: 0.88 / MAX: 1.15 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 1 2 3 9 18 27 36 45 SE +/- 0.13, N = 7 SE +/- 0.15, N = 3 SE +/- 0.20, N = 3 41.21 41.37 41.08 MIN: 39.9 / MAX: 107.26 MIN: 40.38 / MAX: 43.2 MIN: 39.95 / MAX: 42.38 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 1 2 3 20 40 60 80 100 SE +/- 16.43, N = 7 SE +/- 0.28, N = 3 SE +/- 0.18, N = 3 86.05 70.43 69.45 MIN: 67.47 / MAX: 1163.59 MIN: 68.71 / MAX: 81.81 MIN: 68.18 / MAX: 83.46 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 1 2 3 5 10 15 20 25 SE +/- 0.13, N = 7 SE +/- 0.21, N = 3 SE +/- 0.01, N = 3 20.82 20.92 20.60 MIN: 20.25 / MAX: 23.91 MIN: 20.37 / MAX: 21.98 MIN: 19.97 / MAX: 80.96 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 1 2 3 7 14 21 28 35 SE +/- 0.22, N = 7 SE +/- 0.18, N = 3 SE +/- 0.17, N = 3 28.16 28.60 28.45 MIN: 24.94 / MAX: 57.14 MIN: 24.84 / MAX: 71.78 MIN: 25.05 / MAX: 67.53 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 1 2 3 13 26 39 52 65 SE +/- 0.07, N = 7 SE +/- 0.18, N = 3 SE +/- 0.85, N = 3 58.79 58.78 58.06 MIN: 57.94 / MAX: 143.22 MIN: 58.17 / MAX: 141.3 MIN: 56.02 / MAX: 60.02 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 1 2 3 5 10 15 20 25 SE +/- 0.22, N = 7 SE +/- 0.06, N = 3 SE +/- 0.29, N = 3 21.28 21.30 21.35 MIN: 11.57 / MAX: 43.2 MIN: 10.97 / MAX: 44.03 MIN: 11.91 / MAX: 44.45 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 1 2 3 60 120 180 240 300 SE +/- 0.71, N = 3 SE +/- 0.04, N = 3 SE +/- 0.56, N = 3 257.86 257.62 257.43 MIN: 244.35 / MAX: 289.49 MIN: 248.05 / MAX: 298.14 MIN: 248.96 / MAX: 288.02 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 1 2 3 50 100 150 200 250 SE +/- 0.56, N = 3 SE +/- 0.79, N = 3 SE +/- 2.15, N = 3 237.82 237.24 237.83 MIN: 234.12 / MAX: 239.65 MIN: 234.05 / MAX: 239.25 MIN: 230.78 / MAX: 240.86 1. (CXX) g++ options: -fopenmp -pthread -fvisibility=hidden -O3 -rdynamic -ldl
Phoronix Test Suite v10.8.5