rasp-ncnn ARMv7 Cortex-A72 testing with a BCM2711 Raspberry Pi 4 Model B Rev 1.4 and vc4drmfb on Raspbian 11 via the Phoronix Test Suite.
Compare your own system(s) to this result file with the
Phoronix Test Suite by running the command:
phoronix-test-suite benchmark 2404227-NE-RASPNCNN619 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 Processor: ARMv7 Cortex-A72 @ 1.80GHz (4 Cores), Motherboard: BCM2711 Raspberry Pi 4 Model B Rev 1.4, Chipset: Broadcom BCM2711, Memory: 8GB, Disk: 32GB SD32G, Graphics: vc4drmfb
OS: Raspbian 11, Kernel: 5.15.84-v7l+ (armv7l), Display Server: X Server 1.20.11, Compiler: GCC 10.2.1 20210110, File-System: ext4, Screen Resolution: 1920x1080
Kernel Notes: snd_bcm2835.enable_compat_alsa=0 snd_bcm2835.enable_hdmi=1Compiler Notes: --build=arm-linux-gnueabihf --disable-libitm --disable-libquadmath --disable-libquadmath-support --disable-sjlj-exceptions --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-objc-gc=auto --enable-plugin --enable-shared --enable-threads=posix --host=arm-linux-gnueabihf --program-prefix=arm-linux-gnueabihf- --target=arm-linux-gnueabihf --with-arch=armv6 --with-default-libstdcxx-abi=new --with-float=hard --with-fpu=vfp --with-gcc-major-version-only --with-target-system-zlib=auto -vProcessor Notes: Scaling Governor: cpufreq-dt ondemandSecurity 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: Not affected + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Vulnerable + srbds: Not affected + tsx_async_abort: Not affected
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 20230517 Target: CPU - Model: mobilenet ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 160 320 480 640 800 SE +/- 25.04, N = 3 739.49 MIN: 577.98 / MAX: 924.06 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU-v2-v2 - Model: mobilenet-v2 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 40 80 120 160 200 SE +/- 16.90, N = 3 191.35 MIN: 120.41 / MAX: 276.08 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU-v3-v3 - Model: mobilenet-v3 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 40 80 120 160 200 SE +/- 12.91, N = 3 164.33 MIN: 106.31 / MAX: 246.87 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: shufflenet-v2 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 20 40 60 80 100 SE +/- 6.90, N = 3 99.61 MIN: 59.07 / MAX: 373.02 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: mnasnet ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 40 80 120 160 200 SE +/- 13.43, N = 3 189.49 MIN: 112.3 / MAX: 290.45 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: efficientnet-b0 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 70 140 210 280 350 SE +/- 18.86, N = 3 320.64 MIN: 195.73 / MAX: 449.63 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: blazeface ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 7 14 21 28 35 SE +/- 1.12, N = 3 31.62 MIN: 20.04 / MAX: 70.3 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: googlenet ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 130 260 390 520 650 SE +/- 22.33, N = 3 586.94 MIN: 428.64 / MAX: 751.67 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: vgg16 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 500 1000 1500 2000 2500 SE +/- 44.02, N = 3 2424.98 MIN: 2105.38 / MAX: 2786.92 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: resnet18 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 100 200 300 400 500 SE +/- 16.16, N = 3 469.61 MIN: 321.56 / MAX: 1089.59 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: alexnet ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 70 140 210 280 350 SE +/- 9.89, N = 3 320.31 MIN: 220.47 / MAX: 460.05 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: resnet50 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 300 600 900 1200 1500 SE +/- 22.30, N = 3 1461.65 MIN: 1193.51 / MAX: 1892.2 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: yolov4-tiny ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 200 400 600 800 1000 SE +/- 21.82, N = 3 870.83 MIN: 621.63 / MAX: 1084.37 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: squeezenet_ssd ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 80 160 240 320 400 SE +/- 7.34, N = 3 377.06 MIN: 273.26 / MAX: 739.84 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: regnety_400m ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 60 120 180 240 300 SE +/- 13.64, N = 3 279.45 MIN: 163.73 / MAX: 398.29 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: vision_transformer ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 1300 2600 3900 5200 6500 SE +/- 150.67, N = 3 6105.88 MIN: 5530.32 / MAX: 6883.26 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: CPU - Model: FastestDet ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 30 60 90 120 150 SE +/- 1.54, N = 3 112.11 MIN: 69.16 / MAX: 173.68 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: mobilenet ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 150 300 450 600 750 SE +/- 7.49, N = 3 713.25 MIN: 564.4 / MAX: 886.46 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 50 100 150 200 250 SE +/- 3.26, N = 3 207.03 MIN: 138.13 / MAX: 286.55 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 40 80 120 160 200 SE +/- 2.64, N = 3 175.74 MIN: 109.38 / MAX: 237.6 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: shufflenet-v2 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 20 40 60 80 100 SE +/- 2.28, N = 3 105.47 MIN: 59.37 / MAX: 162.28 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: mnasnet ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 40 80 120 160 200 SE +/- 3.88, N = 3 200.79 MIN: 115.29 / MAX: 282.95 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: efficientnet-b0 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 70 140 210 280 350 SE +/- 5.66, N = 3 331.99 MIN: 199.01 / MAX: 443.16 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: blazeface ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 8 16 24 32 40 SE +/- 0.46, N = 3 32.60 MIN: 20.09 / MAX: 70.51 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: googlenet ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 130 260 390 520 650 SE +/- 8.75, N = 3 598.85 MIN: 430.63 / MAX: 793.85 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: vgg16 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 500 1000 1500 2000 2500 SE +/- 40.56, N = 3 2423.40 MIN: 2109.8 / MAX: 2808.17 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: resnet18 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 100 200 300 400 500 SE +/- 8.10, N = 3 474.60 MIN: 323.99 / MAX: 585.89 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: alexnet ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 70 140 210 280 350 SE +/- 4.75, N = 3 322.83 MIN: 184.93 / MAX: 433.23 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: resnet50 ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 300 600 900 1200 1500 SE +/- 26.28, N = 3 1467.91 MIN: 1190.19 / MAX: 1737.16 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: yolov4-tiny ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 200 400 600 800 1000 SE +/- 4.61, N = 3 852.24 MIN: 666.33 / MAX: 1184.15 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: squeezenet_ssd ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 80 160 240 320 400 SE +/- 13.00, N = 3 369.39 MIN: 266.12 / MAX: 473.78 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: regnety_400m ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 60 120 180 240 300 SE +/- 4.79, N = 3 286.33 MIN: 164.8 / MAX: 417.84 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: vision_transformer ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 1200 2400 3600 4800 6000 SE +/- 46.74, N = 3 5779.54 MIN: 4960 / MAX: 6576.86 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: FastestDet ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 30 60 90 120 150 SE +/- 2.76, N = 3 111.76 MIN: 69.55 / MAX: 181.33 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
ARMv7 Cortex-A72 - vc4drmfb - BCM2711 Raspberry Pi 4 Processor: ARMv7 Cortex-A72 @ 1.80GHz (4 Cores), Motherboard: BCM2711 Raspberry Pi 4 Model B Rev 1.4, Chipset: Broadcom BCM2711, Memory: 8GB, Disk: 32GB SD32G, Graphics: vc4drmfb
OS: Raspbian 11, Kernel: 5.15.84-v7l+ (armv7l), Display Server: X Server 1.20.11, Compiler: GCC 10.2.1 20210110, File-System: ext4, Screen Resolution: 1920x1080
Kernel Notes: snd_bcm2835.enable_compat_alsa=0 snd_bcm2835.enable_hdmi=1Compiler Notes: --build=arm-linux-gnueabihf --disable-libitm --disable-libquadmath --disable-libquadmath-support --disable-sjlj-exceptions --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-objc-gc=auto --enable-plugin --enable-shared --enable-threads=posix --host=arm-linux-gnueabihf --program-prefix=arm-linux-gnueabihf- --target=arm-linux-gnueabihf --with-arch=armv6 --with-default-libstdcxx-abi=new --with-float=hard --with-fpu=vfp --with-gcc-major-version-only --with-target-system-zlib=auto -vProcessor Notes: Scaling Governor: cpufreq-dt ondemandSecurity 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: Not affected + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Vulnerable + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 22 April 2024 03:19 by user pi.