2024-10-17-ncnn-gpu AMD Ryzen Threadripper 3960X 24-Core testing with a ASUS ROG ZENITH II EXTREME ALPHA (0902 BIOS) and AMD Radeon RX 6900 XT 16GB on Fedora Linux 40 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2410170-ENB-2024101798&gru .
2024-10-17-ncnn-gpu Processor Motherboard Chipset Memory Disk Graphics Audio Monitor Network OS Kernel Desktop Display Server OpenGL OpenCL Compiler File-System Screen Resolution AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX AMD Ryzen Threadripper 3960X 24-Core @ 3.80GHz (24 Cores / 48 Threads) ASUS ROG ZENITH II EXTREME ALPHA (0902 BIOS) AMD Starship/Matisse 128GB 2 x 2000GB Sabrent Rocket Q AMD Radeon RX 6900 XT 16GB AMD Navi 21/23 MPCP28UHD + MP Monitor Aquantia AQtion AQC107 NBase-T/IEEE + Intel I211 + Intel Wi-Fi 6 AX200 Fedora Linux 40 6.11.3-200.fc40.x86_64 (x86_64) GNOME Shell 46.5 X Server 1.20.14 + Wayland 4.6 Mesa 24.1.7 (LLVM 18.1.6 DRM 3.59) OpenCL 2.1 AMD-APP (3614.0) + OpenCL 3.0 PoCL 5.0 Linux RELOC SPIR LLVM 17.0.6 SLEEF DISTRO POCL_DEBUG GCC 14.2.1 20240912 + Clang 18.1.8 + LLVM 18.1.8 btrfs 7680x2160 OpenBenchmarking.org - Transparent Huge Pages: madvise - --build=x86_64-redhat-linux --disable-libunwind-exceptions --enable-__cxa_atexit --enable-bootstrap --enable-cet --enable-checking=release --enable-gnu-indirect-function --enable-gnu-unique-object --enable-initfini-array --enable-languages=c,c++,fortran,objc,obj-c++,ada,go,d,m2,lto --enable-libstdcxx-backtrace --enable-link-serialization=1 --enable-multilib --enable-offload-defaulted --enable-offload-targets=nvptx-none,amdgcn-amdhsa --enable-plugin --enable-shared --enable-threads=posix --mandir=/usr/share/man --with-arch_32=i686 --with-build-config=bootstrap-lto --with-gcc-major-version-only --with-libstdcxx-zoneinfo=/usr/share/zoneinfo --with-linker-hash-style=gnu --with-tune=generic --without-cuda-driver - Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0x830107c - gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Mitigation of untrained return thunk; SMT enabled with STIBP protection + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected
2024-10-17-ncnn-gpu 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: Vulkan GPU - vision_transformer ncnn: Vulkan GPU - FastestDet AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 18.78 7.71 6.19 8.78 6.45 9.27 3.70 25.04 43.28 16.01 11.78 27.64 32.25 18.58 22.28 73.13 9.71 OpenBenchmarking.org
NCNN Target: Vulkan GPU - Model: mobilenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: mobilenet AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 5 10 15 20 25 SE +/- 0.36, N = 12 18.78 MIN: 15.72 / MAX: 589.09 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2 OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2 AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 2 4 6 8 10 SE +/- 0.71, N = 12 7.71 MIN: 6.19 / MAX: 556.12 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3 OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3 AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 2 4 6 8 10 SE +/- 0.24, N = 12 6.19 MIN: 5.46 / MAX: 262.23 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: shufflenet-v2 OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: shufflenet-v2 AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 2 4 6 8 10 SE +/- 0.50, N = 12 8.78 MIN: 7.24 / MAX: 504 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: mnasnet OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: mnasnet AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 2 4 6 8 10 SE +/- 0.36, N = 12 6.45 MIN: 5.4 / MAX: 916.71 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: efficientnet-b0 OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: efficientnet-b0 AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 3 6 9 12 15 SE +/- 0.50, N = 12 9.27 MIN: 7.86 / MAX: 315.35 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: blazeface OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: blazeface AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 0.8325 1.665 2.4975 3.33 4.1625 SE +/- 0.22, N = 12 3.70 MIN: 2.87 / MAX: 278.5 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: googlenet OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: googlenet AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 6 12 18 24 30 SE +/- 1.88, N = 12 25.04 MIN: 16.8 / MAX: 988.79 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: vgg16 OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: vgg16 AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 10 20 30 40 50 SE +/- 1.27, N = 12 43.28 MIN: 33.71 / MAX: 423.2 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: resnet18 OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: resnet18 AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 4 8 12 16 20 SE +/- 0.65, N = 12 16.01 MIN: 11.05 / MAX: 586.57 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: alexnet OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: alexnet AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 3 6 9 12 15 SE +/- 0.34, N = 12 11.78 MIN: 8.2 / MAX: 373.71 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: resnet50 OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: resnet50 AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 7 14 21 28 35 SE +/- 0.86, N = 12 27.64 MIN: 19.54 / MAX: 1095.65 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: yolov4-tiny OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: yolov4-tiny AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 7 14 21 28 35 SE +/- 0.74, N = 12 32.25 MIN: 25.08 / MAX: 650.3 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: squeezenet_ssd OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: squeezenet_ssd AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 5 10 15 20 25 SE +/- 0.61, N = 11 18.58 MIN: 15.49 / MAX: 1036.34 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: regnety_400m OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: regnety_400m AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 5 10 15 20 25 SE +/- 0.66, N = 12 22.28 MIN: 18.51 / MAX: 1014.4 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: vision_transformer OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: vision_transformer AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 16 32 48 64 80 SE +/- 1.76, N = 12 73.13 MIN: 62.69 / MAX: 1079.77 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
NCNN Target: Vulkan GPU - Model: FastestDet OpenBenchmarking.org ms, Fewer Is Better NCNN 20230517 Target: Vulkan GPU - Model: FastestDet AMD Ryzen Threadripper 3960X 24-Core - AMD Radeon RX 3 6 9 12 15 SE +/- 0.35, N = 12 9.71 MIN: 8.07 / MAX: 424.44 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
Phoronix Test Suite v10.8.5