f1 AMD Ryzen 5 1600 Six-Core testing with a MSI B450-A PRO (MS-7B86) v2.0 (A.D0 BIOS) and XFX AMD Radeon RX 470/480/570/570X/580/580X/590 4GB on Ubuntu 20.04 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2108047-IB-F1767180826&grt .
f1 Processor Motherboard Chipset Memory Disk Graphics Audio Monitor Network OS Kernel Desktop Display Server OpenGL Vulkan Compiler File-System Screen Resolution OpenCL XFX AMD Radeon RX 470 120W-cap amdgpu 120W-cap-amdgpu 120W-cap-amdgpu-compute amdgpu2 amdgpu1 amdgpu3 120W-cap-amdgpu-2 cpu AMD Ryzen 5 1600 Six-Core @ 3.20GHz (6 Cores / 12 Threads) MSI B450-A PRO (MS-7B86) v2.0 (A.D0 BIOS) AMD 17h 16GB 1000GB Western Digital WD10EZEX-08M + 240GB Patriot Burst XFX AMD Radeon RX 470/480/570/570X/580/580X/590 4GB (1226/1750MHz) AMD Ellesmere HDMI Audio VW225 Realtek RTL8111/8168/8411 Ubuntu 20.04 5.11.0-25-generic (x86_64) GNOME Shell 3.36.9 X Server 1.20.9 4.6 Mesa 21.0.3 (LLVM 12.0.0) 1.2.145 Clang 12.0.1-++20210802051739+fed41342a82f-1~exp1~20210802152501.132 ext4 1680x1050 4.6 Mesa 21.1.0-devel (LLVM 12.0.0) OpenCL 2.0 AMD-APP (3137.0) OpenBenchmarking.org Kernel Details - libahci.ignore_sss=1 amdgpu.ppfeaturemask=0xfffd7fff - Transparent Huge Pages: madvise Processor Details - XFX AMD Radeon RX 470: Scaling Governor: acpi-cpufreq ondemand (Boost: Enabled) - CPU Microcode: 0x8001138 - 120W-cap: Scaling Governor: acpi-cpufreq ondemand (Boost: Enabled) - CPU Microcode: 0x8001138 - amdgpu: Scaling Governor: acpi-cpufreq ondemand (Boost: Enabled) - CPU Microcode: 0x8001138 - 120W-cap-amdgpu: Scaling Governor: acpi-cpufreq ondemand (Boost: Enabled) - CPU Microcode: 0x8001138 - 120W-cap-amdgpu-compute: Scaling Governor: acpi-cpufreq ondemand (Boost: Enabled) - CPU Microcode: 0x8001138 - amdgpu2: Scaling Governor: acpi-cpufreq ondemand (Boost: Enabled) - CPU Microcode: 0x8001138 - amdgpu1: Scaling Governor: acpi-cpufreq ondemand (Boost: Enabled) - CPU Microcode: 0x8001138 - amdgpu3: Scaling Governor: acpi-cpufreq ondemand (Boost: Enabled) - CPU Microcode: 0x8001138 - 120W-cap-amdgpu-2: Scaling Governor: acpi-cpufreq ondemand (Boost: Enabled) - CPU Microcode: 0x8001138 - cpu: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0x8001138 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 Graphics Details - amdgpu, 120W-cap-amdgpu, 120W-cap-amdgpu-compute, amdgpu2, amdgpu1: GLAMOR Python Details - amdgpu2: Python 3.8.10
f1 betsy: ETC1 - Highest betsy: ETC2 RGB - Highest glmark2: 1280 x 1024 gputest: Furmark - 1024 x 768 - Windowed 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 ncnn: Vulkan GPU - squeezenet_ssd ncnn: Vulkan GPU - regnety_400m 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: CPU - squeezenet_ssd ncnn: CPU - regnety_400m vkpeak: fp32-scalar vkpeak: fp32-vec4 vkpeak: fp64-scalar vkpeak: fp64-vec4 vkpeak: int32-scalar vkpeak: int32-vec4 XFX AMD Radeon RX 470 120W-cap amdgpu 120W-cap-amdgpu 120W-cap-amdgpu-compute amdgpu2 amdgpu1 amdgpu3 120W-cap-amdgpu-2 cpu 4205.53 4158.75 207.92 207.48 957.54 969.06 10.357 12.470 4258.81 4172.91 207.28 208.60 970.97 985.96 12.557 12.245 12.245 7035 7046 9186 8.65 3.42 4.65 2.94 3.60 11.61 1.40 7.20 17.43 3.48 6.47 9.08 12.71 7.10 6.00 8.71 3.5 4.67 2.95 3.62 11.62 1.41 7.23 17.43 3.48 6.43 9.08 12.73 7.13 5.97 27.11 8.42 7.47 7.27 7.45 11.40 2.70 22.59 85.04 21.22 17.38 44.24 41.85 29.61 18.71 OpenBenchmarking.org
Betsy GPU Compressor Codec: ETC1 - Quality: Highest OpenBenchmarking.org Seconds, Fewer Is Better Betsy GPU Compressor 1.1 Beta Codec: ETC1 - Quality: Highest 120W-cap 3 6 9 12 15 SE +/- 0.03, N = 3 10.36 1. (CXX) clang++ options: -O3 -O2 -lpthread -ldl
Betsy GPU Compressor Codec: ETC2 RGB - Quality: Highest OpenBenchmarking.org Seconds, Fewer Is Better Betsy GPU Compressor 1.1 Beta Codec: ETC2 RGB - Quality: Highest 120W-cap amdgpu 120W-cap-amdgpu 120W-cap-amdgpu-compute 3 6 9 12 15 SE +/- 0.27, N = 15 SE +/- 0.29, N = 14 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 12.47 12.56 12.25 12.25 1. (CXX) clang++ options: -O3 -O2 -lpthread -ldl
GLmark2 Resolution: 1280 x 1024 OpenBenchmarking.org Score, More Is Better GLmark2 2020.04 Resolution: 1280 x 1024 120W-cap-amdgpu-compute amdgpu2 1500 3000 4500 6000 7500 7035 7046
GpuTest Test: Furmark - Resolution: 1024 x 768 - Mode: Windowed OpenBenchmarking.org Points, More Is Better GpuTest 0.7.0 Test: Furmark - Resolution: 1024 x 768 - Mode: Windowed amdgpu1 2K 4K 6K 8K 10K SE +/- 129.83, N = 3 9186
NCNN Target: Vulkan GPU - Model: mobilenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: mobilenet amdgpu3 120W-cap-amdgpu-2 2 4 6 8 10 SE +/- 0.01, N = 3 SE +/- 0.09, N = 3 8.65 8.71 MIN: 8.42 / MAX: 28.57 MIN: 8.43 / MAX: 26.59 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2 amdgpu3 120W-cap-amdgpu-2 0.7875 1.575 2.3625 3.15 3.9375 SE +/- 0.02, N = 3 SE +/- 0.10, N = 3 3.42 3.50 MIN: 3.36 / MAX: 17.85 MIN: 3.36 / MAX: 6.23 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3 amdgpu3 120W-cap-amdgpu-2 1.0508 2.1016 3.1524 4.2032 5.254 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 4.65 4.67 MIN: 4.59 / MAX: 18.53 MIN: 4.54 / MAX: 18.03 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU - Model: shufflenet-v2 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: shufflenet-v2 amdgpu3 120W-cap-amdgpu-2 0.6638 1.3276 1.9914 2.6552 3.319 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 2.94 2.95 MIN: 2.82 / MAX: 4.7 MIN: 2.82 / MAX: 4.8 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU - Model: mnasnet OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: mnasnet amdgpu3 120W-cap-amdgpu-2 0.8145 1.629 2.4435 3.258 4.0725 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 3.60 3.62 MIN: 3.56 / MAX: 6.22 MIN: 3.56 / MAX: 16.02 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU - Model: efficientnet-b0 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: efficientnet-b0 amdgpu3 120W-cap-amdgpu-2 3 6 9 12 15 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 11.61 11.62 MIN: 11.1 / MAX: 17.11 MIN: 11.11 / MAX: 16.43 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU - Model: blazeface OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: blazeface amdgpu3 120W-cap-amdgpu-2 0.3173 0.6346 0.9519 1.2692 1.5865 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 1.40 1.41 MIN: 1.33 / MAX: 2.17 MIN: 1.34 / MAX: 3.81 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU - Model: googlenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: googlenet amdgpu3 120W-cap-amdgpu-2 2 4 6 8 10 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 7.20 7.23 MIN: 7.15 / MAX: 9.42 MIN: 7.14 / MAX: 21.57 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU - Model: vgg16 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: vgg16 amdgpu3 120W-cap-amdgpu-2 4 8 12 16 20 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 17.43 17.43 MIN: 17.01 / MAX: 20.67 MIN: 16.89 / MAX: 20.76 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU - Model: resnet18 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: resnet18 amdgpu3 120W-cap-amdgpu-2 0.783 1.566 2.349 3.132 3.915 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 3.48 3.48 MIN: 3.34 / MAX: 5.65 MIN: 3.36 / MAX: 6.72 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU - Model: alexnet OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: alexnet amdgpu3 120W-cap-amdgpu-2 2 4 6 8 10 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 6.47 6.43 MIN: 6.14 / MAX: 8.48 MIN: 6.14 / MAX: 20.06 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU - Model: resnet50 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: resnet50 amdgpu3 120W-cap-amdgpu-2 3 6 9 12 15 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 9.08 9.08 MIN: 8.83 / MAX: 20.52 MIN: 8.84 / MAX: 20.71 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU - Model: yolov4-tiny OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: yolov4-tiny amdgpu3 120W-cap-amdgpu-2 3 6 9 12 15 SE +/- 0.03, N = 3 SE +/- 0.01, N = 3 12.71 12.73 MIN: 12.46 / MAX: 18.08 MIN: 12.48 / MAX: 25.8 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU - Model: squeezenet_ssd OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: squeezenet_ssd amdgpu3 120W-cap-amdgpu-2 2 4 6 8 10 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 7.10 7.13 MIN: 6.93 / MAX: 10.04 MIN: 6.95 / MAX: 21.37 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: Vulkan GPU - Model: regnety_400m OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: Vulkan GPU - Model: regnety_400m amdgpu3 120W-cap-amdgpu-2 2 4 6 8 10 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 6.00 5.97 MIN: 5.72 / MAX: 18.9 MIN: 5.73 / MAX: 8.35 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: mobilenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: mobilenet cpu 6 12 18 24 30 SE +/- 0.17, N = 3 27.11 MIN: 26.49 / MAX: 46.94 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU-v2-v2 - Model: mobilenet-v2 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU-v2-v2 - Model: mobilenet-v2 cpu 2 4 6 8 10 SE +/- 0.24, N = 3 8.42 MIN: 7.95 / MAX: 36.57 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU-v3-v3 - Model: mobilenet-v3 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU-v3-v3 - Model: mobilenet-v3 cpu 2 4 6 8 10 SE +/- 0.05, N = 3 7.47 MIN: 7.13 / MAX: 21.22 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: shufflenet-v2 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: shufflenet-v2 cpu 2 4 6 8 10 SE +/- 0.09, N = 3 7.27 MIN: 7 / MAX: 10.53 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: mnasnet OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: mnasnet cpu 2 4 6 8 10 SE +/- 0.02, N = 3 7.45 MIN: 7.32 / MAX: 10.4 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: efficientnet-b0 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: efficientnet-b0 cpu 3 6 9 12 15 SE +/- 0.07, N = 3 11.40 MIN: 11.1 / MAX: 25.42 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: blazeface OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: blazeface cpu 0.6075 1.215 1.8225 2.43 3.0375 SE +/- 0.02, N = 3 2.70 MIN: 2.6 / MAX: 5.32 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: googlenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: googlenet cpu 5 10 15 20 25 SE +/- 0.13, N = 3 22.59 MIN: 22.08 / MAX: 38.69 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: vgg16 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: vgg16 cpu 20 40 60 80 100 SE +/- 1.35, N = 3 85.04 MIN: 82.51 / MAX: 122.64 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: resnet18 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: resnet18 cpu 5 10 15 20 25 SE +/- 0.06, N = 3 21.22 MIN: 20.8 / MAX: 38.82 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: alexnet OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: alexnet cpu 4 8 12 16 20 SE +/- 0.01, N = 3 17.38 MIN: 17.16 / MAX: 30.65 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: resnet50 OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: resnet50 cpu 10 20 30 40 50 SE +/- 0.39, N = 3 44.24 MIN: 43.13 / MAX: 60.99 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: yolov4-tiny OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: yolov4-tiny cpu 10 20 30 40 50 SE +/- 0.56, N = 3 41.85 MIN: 40.66 / MAX: 57.99 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: squeezenet_ssd OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: squeezenet_ssd cpu 7 14 21 28 35 SE +/- 0.20, N = 3 29.61 MIN: 28.91 / MAX: 46.64 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
NCNN Target: CPU - Model: regnety_400m OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: regnety_400m cpu 5 10 15 20 25 SE +/- 0.13, N = 3 18.71 MIN: 18.29 / MAX: 35.25 1. (CXX) clang++ options: -O3 -rdynamic -lomp -lpthread -pthread
vkpeak fp32-scalar OpenBenchmarking.org GFLOPS, More Is Better vkpeak 20210424 fp32-scalar XFX AMD Radeon RX 470 120W-cap 900 1800 2700 3600 4500 SE +/- 38.11, N = 3 SE +/- 11.57, N = 3 4205.53 4258.81
vkpeak fp32-vec4 OpenBenchmarking.org GFLOPS, More Is Better vkpeak 20210424 fp32-vec4 XFX AMD Radeon RX 470 120W-cap 900 1800 2700 3600 4500 SE +/- 15.75, N = 3 SE +/- 14.35, N = 3 4158.75 4172.91
vkpeak fp64-scalar OpenBenchmarking.org GFLOPS, More Is Better vkpeak 20210424 fp64-scalar XFX AMD Radeon RX 470 120W-cap 50 100 150 200 250 SE +/- 0.68, N = 3 SE +/- 1.30, N = 3 207.92 207.28
vkpeak fp64-vec4 OpenBenchmarking.org GFLOPS, More Is Better vkpeak 20210424 fp64-vec4 XFX AMD Radeon RX 470 120W-cap 50 100 150 200 250 SE +/- 1.03, N = 3 SE +/- 0.01, N = 3 207.48 208.60
vkpeak int32-scalar OpenBenchmarking.org GIOPS, More Is Better vkpeak 20210424 int32-scalar XFX AMD Radeon RX 470 120W-cap 200 400 600 800 1000 SE +/- 4.66, N = 3 SE +/- 0.40, N = 3 957.54 970.97
vkpeak int32-vec4 OpenBenchmarking.org GIOPS, More Is Better vkpeak 20210424 int32-vec4 XFX AMD Radeon RX 470 120W-cap 200 400 600 800 1000 SE +/- 5.15, N = 3 SE +/- 0.27, N = 3 969.06 985.96
Phoronix Test Suite v10.8.4