AMD Ryzen 9 7950X compiler benchmarks by Michael Larabel for a future article.
AOCC 4.0 Processor: AMD Ryzen 9 7950X 16-Core @ 5.88GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR X670E HERO (0703 BIOS), Chipset: AMD Device 14d8, Memory: 32GB, Disk: 1000GB Sabrent Rocket 4.0 Plus, Graphics: AMD Radeon RX 6800 16GB (2475/1000MHz), Audio: AMD Navi 21/23, Monitor: ASUS MG28U, Network: Intel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411
OS: Ubuntu 22.10, Kernel: 6.1.0-060100rc3daily20221103-generic (x86_64), Desktop: GNOME Shell 43.0, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.2.1 (LLVM 15.0.2 DRM 3.49), Vulkan: 1.3.224, Compiler: Clang 14.0.6, File-System: ext4, Screen Resolution: 3840x2160
Kernel Notes: Transparent Huge Pages: madviseEnvironment Notes: CXXFLAGS="-O3 -march=native -flto" CFLAGS="-O3 -march=native -flto"Compiler Notes: Optimized build with assertions; Default target: x86_64-unknown-linux-gnu; Host CPU: znver4Processor Notes: Scaling Governor: amd-pstate schedutil (Boost: Enabled) - CPU Microcode: 0xa601203Graphics Notes: BAR1 / Visible vRAM Size: 16368 MB - vBIOS Version: 113-D4120900-101Python Notes: Python 3.10.7Security 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: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
GCC 12.2 OS: Ubuntu 22.10, Kernel: 6.1.0-060100rc3daily20221103-generic (x86_64), Desktop: GNOME Shell 43.0, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.2.1 (LLVM 15.0.2 DRM 3.49), Vulkan: 1.3.224, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 3840x2160
Kernel Notes: Transparent Huge Pages: madviseEnvironment Notes: CXXFLAGS="-O3 -march=native -flto" CFLAGS="-O3 -march=native -flto"Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --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-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-gcn/usr --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 -vProcessor Notes: Scaling Governor: amd-pstate schedutil (Boost: Enabled) - CPU Microcode: 0xa601203Graphics Notes: BAR1 / Visible vRAM Size: 16368 MB - vBIOS Version: 113-D4120900-101Python Notes: Python 3.10.7Security 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: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
GCC 13.0 14 Nov OS: Ubuntu 22.10, Kernel: 6.1.0-060100rc3daily20221103-generic (x86_64), Desktop: GNOME Shell 43.0, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.2.1 (LLVM 15.0.2 DRM 3.49), Vulkan: 1.3.224, Compiler: GCC 13.0.0 20221114 + clang (GCC) 13.0.0 20221114 (experimental), File-System: ext4, Screen Resolution: 3840x2160
Kernel Notes: Transparent Huge Pages: madviseEnvironment Notes: CXXFLAGS="-O3 -march=native -flto" CFLAGS="-O3 -march=native -flto"Compiler Notes: --disable-multilibProcessor Notes: Scaling Governor: amd-pstate schedutil (Boost: Enabled) - CPU Microcode: 0xa601203Graphics Notes: BAR1 / Visible vRAM Size: 16368 MB - vBIOS Version: 113-D4120900-101Python Notes: Python 3.10.7Security 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: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
LLVM Clang 14 OS: Ubuntu 22.10, Kernel: 6.1.0-060100rc3daily20221103-generic (x86_64), Desktop: GNOME Shell 43.0, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.2.1 (LLVM 15.0.2 DRM 3.49), Vulkan: 1.3.224, Compiler: Clang 14.0.6-2, File-System: ext4, Screen Resolution: 3840x2160
Kernel Notes: Transparent Huge Pages: madviseEnvironment Notes: CXXFLAGS="-O3 -march=native -flto" CFLAGS="-O3 -march=native -flto"Processor Notes: Scaling Governor: amd-pstate schedutil (Boost: Enabled) - CPU Microcode: 0xa601203Graphics Notes: BAR1 / Visible vRAM Size: 16368 MB - vBIOS Version: 113-D4120900-101Python Notes: Python 3.10.7Security 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: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
LLVM Clang 15.0.2 OS: Ubuntu 22.10, Kernel: 6.1.0-060100rc3daily20221103-generic (x86_64), Desktop: GNOME Shell 43.0, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.2.1 (LLVM 15.0.2 DRM 3.49), Vulkan: 1.3.224, Compiler: Clang 15.0.2-1, File-System: ext4, Screen Resolution: 3840x2160
OpenSSL OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.0 Algorithm: SHA256 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 9000M 18000M 27000M 36000M 45000M SE +/- 138604481.51, N = 3 SE +/- 115194474.21, N = 3 SE +/- 51284060.91, N = 3 SE +/- 97264370.12, N = 3 SE +/- 115974592.12, N = 3 40426163787 36864232800 37385343083 39948818857 37916267870 -Qunused-arguments -Qunused-arguments -Qunused-arguments 1. (CC) gcc options: -pthread -m64 -O3 -march=native -flto -lssl -lcrypto -ldl
QuadRay VectorChief's QuadRay is a real-time ray-tracing engine written to support SIMD across ARM, MIPS, PPC, and x86/x86_64 processors. QuadRay supports SSE/SSE2/SSE4 and AVX/AVX2/AVX-512 usage on Intel/AMD CPUs. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org FPS, More Is Better QuadRay 2022.05.25 Scene: 1 - Resolution: 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 6 12 18 24 30 SE +/- 0.12, N = 3 SE +/- 0.08, N = 3 SE +/- 0.09, N = 3 SE +/- 0.05, N = 3 SE +/- 0.05, N = 3 26.43 26.48 26.30 26.23 26.15 1. (CXX) g++ options: -O3 -pthread -lm -lstdc++ -lX11 -lXext -lpthread
OpenBenchmarking.org FPS, More Is Better QuadRay 2022.05.25 Scene: 5 - Resolution: 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.468 0.936 1.404 1.872 2.34 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 2.05 2.08 1.93 2.04 1.92 1. (CXX) g++ options: -O3 -pthread -lm -lstdc++ -lX11 -lXext -lpthread
OpenBenchmarking.org FPS, More Is Better QuadRay 2022.05.25 Scene: 1 - Resolution: 1080p AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 20 40 60 80 100 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.20, N = 3 SE +/- 0.05, N = 3 SE +/- 0.11, N = 3 102.79 102.31 102.06 101.89 102.34 1. (CXX) g++ options: -O3 -pthread -lm -lstdc++ -lX11 -lXext -lpthread
OpenBenchmarking.org FPS, More Is Better QuadRay 2022.05.25 Scene: 5 - Resolution: 1080p AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 2 4 6 8 10 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 8.12 8.22 7.65 8.07 7.60 1. (CXX) g++ options: -O3 -pthread -lm -lstdc++ -lX11 -lXext -lpthread
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.2.dev Model: Face Detection FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 4 8 12 16 20 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.06, N = 3 SE +/- 0.05, N = 3 14.29 14.29 14.37 14.35 -fno-strict-overflow -fwrapv 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.2.dev Model: Person Detection FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 2 4 6 8 10 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 7.53 7.50 7.46 7.47 -fno-strict-overflow -fwrapv 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.2.dev Model: Person Detection FP32 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 2 4 6 8 10 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 7.37 7.34 7.29 7.43 -fno-strict-overflow -fwrapv 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.2.dev Model: Vehicle Detection FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 160 320 480 640 800 SE +/- 6.50, N = 3 SE +/- 1.28, N = 3 SE +/- 4.19, N = 3 SE +/- 2.68, N = 3 752.27 757.51 752.41 742.44 -fno-strict-overflow -fwrapv 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.2.dev Model: Face Detection FP16-INT8 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 7 14 21 28 35 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.10, N = 3 SE +/- 0.11, N = 3 28.19 28.09 28.15 28.27 -fno-strict-overflow -fwrapv 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.2.dev Model: Vehicle Detection FP16-INT8 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 400 800 1200 1600 2000 SE +/- 1.21, N = 3 SE +/- 0.70, N = 3 SE +/- 1.75, N = 3 SE +/- 4.72, N = 3 1847.73 1837.19 1851.09 1845.31 -fno-strict-overflow -fwrapv 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.2.dev Model: Weld Porosity Detection FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 300 600 900 1200 1500 SE +/- 0.72, N = 3 SE +/- 1.86, N = 3 SE +/- 1.62, N = 3 SE +/- 0.50, N = 3 1439.92 1429.51 1434.10 1438.37 -fno-strict-overflow -fwrapv 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.2.dev Model: Machine Translation EN To DE FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 30 60 90 120 150 SE +/- 0.90, N = 3 SE +/- 1.44, N = 3 SE +/- 0.76, N = 3 SE +/- 0.62, N = 3 133.33 134.42 133.59 135.00 -fno-strict-overflow -fwrapv 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.2.dev Model: Weld Porosity Detection FP16-INT8 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 600 1200 1800 2400 3000 SE +/- 0.61, N = 3 SE +/- 2.62, N = 3 SE +/- 0.54, N = 3 SE +/- 1.94, N = 3 2896.90 2879.68 2887.98 2895.02 -fno-strict-overflow -fwrapv 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.2.dev Model: Person Vehicle Bike Detection FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 400 800 1200 1600 2000 SE +/- 6.39, N = 3 SE +/- 6.49, N = 3 SE +/- 4.79, N = 3 SE +/- 17.91, N = 4 1664.77 1658.30 1654.66 1645.57 -fno-strict-overflow -fwrapv 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.2.dev Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 10K 20K 30K 40K 50K SE +/- 34.34, N = 3 SE +/- 46.50, N = 3 SE +/- 43.19, N = 3 SE +/- 16.19, N = 3 44290.73 44470.59 44085.25 44075.75 -fno-strict-overflow -fwrapv 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.2.dev Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 14K 28K 42K 56K 70K SE +/- 68.00, N = 3 SE +/- 73.58, N = 3 SE +/- 21.27, N = 3 SE +/- 31.56, N = 3 63880.92 63460.60 63331.04 63386.57 -fno-strict-overflow -fwrapv 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
yquake2 This is a test of Yamagi Quake II. Yamagi Quake II is an enhanced client for id Software's Quake II with focus on offline and coop gameplay. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better yquake2 8.10 Renderer: Software CPU - AF: On - MSAA: On - Resolution: 1920 x 1080 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 40 80 120 160 200 SE +/- 0.15, N = 3 SE +/- 0.91, N = 3 SE +/- 0.43, N = 3 196.3 185.1 191.4 1. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC Renderer: Software CPU - AF: On - MSAA: On - Resolution: 1920 x 1080
LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
OpenBenchmarking.org Frames Per Second, More Is Better yquake2 8.10 Renderer: Software CPU - AF: Off - MSAA: On - Resolution: 1920 x 1080 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 50 100 150 200 250 SE +/- 1.56, N = 3 SE +/- 0.15, N = 3 SE +/- 0.64, N = 3 205.6 192.0 199.0 1. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC Renderer: Software CPU - AF: Off - MSAA: On - Resolution: 1920 x 1080
LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
OpenBenchmarking.org Frames Per Second, More Is Better yquake2 8.10 Renderer: Software CPU - AF: On - MSAA: Off - Resolution: 1920 x 1080 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 40 80 120 160 200 SE +/- 2.06, N = 3 SE +/- 1.68, N = 3 SE +/- 0.30, N = 3 198.1 185.9 192.3 1. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC Renderer: Software CPU - AF: On - MSAA: Off - Resolution: 1920 x 1080
LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
OpenBenchmarking.org Frames Per Second, More Is Better yquake2 8.10 Renderer: Software CPU - AF: Off - MSAA: Off - Resolution: 1920 x 1080 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 50 100 150 200 250 SE +/- 1.48, N = 3 SE +/- 0.24, N = 3 SE +/- 0.63, N = 3 205.7 192.2 201.0 1. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC Renderer: Software CPU - AF: Off - MSAA: Off - Resolution: 1920 x 1080
LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
OpenBenchmarking.org Frames Per Second, More Is Better yquake2 8.10 Renderer: Software CPU Color Light - AF: On - MSAA: On - Resolution: 1920 x 1080 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 40 80 120 160 200 SE +/- 0.68, N = 3 SE +/- 0.93, N = 3 SE +/- 0.33, N = 3 185.3 171.2 181.3 1. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC Renderer: Software CPU Color Light - AF: On - MSAA: On - Resolution: 1920 x 1080
LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
OpenBenchmarking.org Frames Per Second, More Is Better yquake2 8.10 Renderer: Software CPU Color Light - AF: Off - MSAA: On - Resolution: 1920 x 1080 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 40 80 120 160 200 SE +/- 0.54, N = 3 SE +/- 0.37, N = 3 SE +/- 0.78, N = 3 189.9 176.2 184.4 1. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC Renderer: Software CPU Color Light - AF: Off - MSAA: On - Resolution: 1920 x 1080
LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
OpenBenchmarking.org Frames Per Second, More Is Better yquake2 8.10 Renderer: Software CPU Color Light - AF: On - MSAA: Off - Resolution: 1920 x 1080 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 40 80 120 160 200 SE +/- 1.77, N = 3 SE +/- 0.26, N = 3 SE +/- 0.50, N = 3 188.2 172.5 178.5 1. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC Renderer: Software CPU Color Light - AF: On - MSAA: Off - Resolution: 1920 x 1080
LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
OpenBenchmarking.org Frames Per Second, More Is Better yquake2 8.10 Renderer: Software CPU Color Light - AF: Off - MSAA: Off - Resolution: 1920 x 1080 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 40 80 120 160 200 SE +/- 2.19, N = 4 SE +/- 0.52, N = 3 SE +/- 0.95, N = 3 191.8 176.4 185.3 1. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC Renderer: Software CPU Color Light - AF: Off - MSAA: Off - Resolution: 1920 x 1080
LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found
AOM AV1 This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.5 Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 5 10 15 20 25 SE +/- 0.11, N = 3 SE +/- 0.04, N = 3 SE +/- 0.08, N = 3 SE +/- 0.11, N = 3 SE +/- 0.02, N = 3 20.75 19.32 20.72 17.76 17.92 1. (CXX) g++ options: -O3 -march=native -flto -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.5 Encoder Mode: Speed 10 Realtime - Input: Bosphorus 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 15 30 45 60 75 SE +/- 0.19, N = 3 SE +/- 0.03, N = 3 SE +/- 0.12, N = 3 SE +/- 0.19, N = 3 SE +/- 0.21, N = 3 65.44 68.24 68.30 64.82 68.71 1. (CXX) g++ options: -O3 -march=native -flto -std=c++11 -U_FORTIFY_SOURCE -lm
Kvazaar This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.1 Video Input: Bosphorus 4K - Video Preset: Very Fast AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 11 22 33 44 55 SE +/- 0.42, N = 3 SE +/- 0.42, N = 3 SE +/- 0.61, N = 3 SE +/- 0.17, N = 3 SE +/- 0.22, N = 3 47.33 41.78 44.12 44.62 45.31 -lpthread -lpthread 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O3 -march=native -flto -lm -lrt
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.1 Video Input: Bosphorus 4K - Video Preset: Ultra Fast AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 20 40 60 80 100 SE +/- 0.03, N = 3 SE +/- 0.29, N = 3 SE +/- 0.22, N = 3 SE +/- 0.28, N = 3 SE +/- 0.30, N = 3 85.35 80.31 82.56 82.97 84.80 -lpthread -lpthread 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O3 -march=native -flto -lm -lrt
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.1 Video Input: Bosphorus 1080p - Video Preset: Very Fast AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 30 60 90 120 150 SE +/- 0.04, N = 3 SE +/- 0.39, N = 3 SE +/- 0.32, N = 3 SE +/- 0.13, N = 3 SE +/- 0.27, N = 3 113.72 102.77 108.11 107.85 109.33 -lpthread -lpthread 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O3 -march=native -flto -lm -lrt
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.1 Video Input: Bosphorus 1080p - Video Preset: Ultra Fast AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 50 100 150 200 250 SE +/- 0.25, N = 3 SE +/- 0.80, N = 3 SE +/- 0.51, N = 3 SE +/- 1.14, N = 3 SE +/- 0.89, N = 3 219.33 204.74 210.39 210.45 213.39 -lpthread -lpthread 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O3 -march=native -flto -lm -lrt
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.2 Encoder Mode: Preset 4 - Input: Bosphorus 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.6995 1.399 2.0985 2.798 3.4975 SE +/- 0.006, N = 3 SE +/- 0.008, N = 3 SE +/- 0.012, N = 3 SE +/- 0.008, N = 3 SE +/- 0.012, N = 3 3.078 3.067 3.109 3.020 3.058 1. (CXX) g++ options: -O3 -march=native -flto -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.2 Encoder Mode: Preset 8 - Input: Bosphorus 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 20 40 60 80 100 SE +/- 0.54, N = 3 SE +/- 0.25, N = 3 SE +/- 0.70, N = 3 SE +/- 0.17, N = 3 SE +/- 0.42, N = 3 76.51 76.40 76.36 74.61 74.82 1. (CXX) g++ options: -O3 -march=native -flto -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.2 Encoder Mode: Preset 10 - Input: Bosphorus 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 30 60 90 120 150 SE +/- 0.71, N = 3 SE +/- 1.43, N = 3 SE +/- 1.28, N = 3 SE +/- 0.71, N = 3 SE +/- 1.02, N = 3 131.29 131.04 132.75 130.11 129.16 1. (CXX) g++ options: -O3 -march=native -flto -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.2 Encoder Mode: Preset 12 - Input: Bosphorus 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 40 80 120 160 200 SE +/- 0.95, N = 3 SE +/- 0.19, N = 3 SE +/- 0.64, N = 3 SE +/- 0.91, N = 3 SE +/- 0.49, N = 3 188.40 187.37 188.62 183.19 184.15 1. (CXX) g++ options: -O3 -march=native -flto -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
SVT-HEVC This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better SVT-HEVC 1.5.0 Tuning: 1 - Input: Bosphorus 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 2 4 6 8 10 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 6.68 6.64 6.63 6.64 6.63 1. (CC) gcc options: -O3 -march=native -flto -fPIE -fPIC -O2 -pie -rdynamic -lpthread -lrt
OpenBenchmarking.org Frames Per Second, More Is Better SVT-HEVC 1.5.0 Tuning: 7 - Input: Bosphorus 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 20 40 60 80 100 SE +/- 0.09, N = 3 SE +/- 0.16, N = 3 SE +/- 0.19, N = 3 SE +/- 0.35, N = 3 SE +/- 0.07, N = 3 108.32 106.57 107.16 105.55 105.69 1. (CC) gcc options: -O3 -march=native -flto -fPIE -fPIC -O2 -pie -rdynamic -lpthread -lrt
OpenBenchmarking.org Frames Per Second, More Is Better SVT-HEVC 1.5.0 Tuning: 10 - Input: Bosphorus 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 40 80 120 160 200 SE +/- 0.26, N = 3 SE +/- 0.31, N = 3 SE +/- 0.18, N = 3 SE +/- 0.19, N = 3 SE +/- 0.29, N = 3 171.09 172.53 173.36 170.79 171.02 1. (CC) gcc options: -O3 -march=native -flto -fPIE -fPIC -O2 -pie -rdynamic -lpthread -lrt
SVT-VP9 This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: VMAF Optimized - Input: Bosphorus 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 30 60 90 120 150 SE +/- 1.06, N = 15 SE +/- 1.87, N = 15 SE +/- 1.85, N = 15 SE +/- 1.13, N = 15 SE +/- 1.81, N = 15 112.44 109.95 110.49 111.35 110.81 1. (CC) gcc options: -O3 -fcommon -march=native -flto -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: PSNR/SSIM Optimized - Input: Bosphorus 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 30 60 90 120 150 SE +/- 0.09, N = 3 SE +/- 0.11, N = 3 SE +/- 0.13, N = 3 SE +/- 0.15, N = 3 SE +/- 0.22, N = 3 121.78 122.32 122.36 120.82 121.26 1. (CC) gcc options: -O3 -fcommon -march=native -flto -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: Visual Quality Optimized - Input: Bosphorus 4K AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 30 60 90 120 150 SE +/- 0.03, N = 3 SE +/- 0.07, N = 3 SE +/- 0.23, N = 3 SE +/- 0.26, N = 3 SE +/- 0.17, N = 3 111.78 111.60 111.61 110.25 111.24 1. (CC) gcc options: -O3 -fcommon -march=native -flto -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
simdjson This is a benchmark of SIMDJSON, a high performance JSON parser. SIMDJSON aims to be the fastest JSON parser and is used by projects like Microsoft FishStore, Yandex ClickHouse, Shopify, and others. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org GB/s, More Is Better simdjson 2.0 Throughput Test: Kostya AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 2 4 6 8 10 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.00, N = 3 5.85 5.89 6.03 5.62 1. (CXX) g++ options: -O3 -march=native -flto Throughput Test: Kostya
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./simdjson: 3: ./bench_ondemand: not found
OpenBenchmarking.org GB/s, More Is Better simdjson 2.0 Throughput Test: TopTweet AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 3 6 9 12 15 SE +/- 0.06, N = 3 SE +/- 0.07, N = 3 SE +/- 0.09, N = 3 SE +/- 0.02, N = 3 10.85 9.90 10.22 10.60 1. (CXX) g++ options: -O3 -march=native -flto Throughput Test: TopTweet
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./simdjson: 3: ./bench_ondemand: not found
OpenBenchmarking.org GB/s, More Is Better simdjson 2.0 Throughput Test: LargeRandom AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 0.4118 0.8236 1.2354 1.6472 2.059 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 1.59 1.83 1.80 1.56 1. (CXX) g++ options: -O3 -march=native -flto Throughput Test: LargeRandom
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./simdjson: 3: ./bench_ondemand: not found
OpenBenchmarking.org GB/s, More Is Better simdjson 2.0 Throughput Test: PartialTweets AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 3 6 9 12 15 SE +/- 0.06, N = 3 SE +/- 0.05, N = 3 SE +/- 0.03, N = 3 SE +/- 0.01, N = 3 11.82 9.85 9.63 11.15 1. (CXX) g++ options: -O3 -march=native -flto Throughput Test: PartialTweets
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./simdjson: 3: ./bench_ondemand: not found
OpenBenchmarking.org GB/s, More Is Better simdjson 2.0 Throughput Test: DistinctUserID AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 3 6 9 12 15 SE +/- 0.10, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 11.97 9.80 9.28 10.77 1. (CXX) g++ options: -O3 -march=native -flto Throughput Test: DistinctUserID
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./simdjson: 3: ./bench_ondemand: not found
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Zoo. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: GPT-2 - Device: CPU - Executor: Parallel AOCC 4.0 GCC 12.2 LLVM Clang 14 1500 3000 4500 6000 7500 SE +/- 19.11, N = 3 SE +/- 34.87, N = 3 SE +/- 21.71, N = 3 7135 7126 7062 -flto=thin -flto=auto -fno-fat-lto-objects -flto=thin 1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt Model: GPT-2 - Device: CPU - Executor: Parallel
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: GPT-2 - Device: CPU - Executor: Standard AOCC 4.0 GCC 12.2 LLVM Clang 14 2K 4K 6K 8K 10K SE +/- 98.49, N = 5 SE +/- 107.39, N = 12 SE +/- 123.19, N = 3 9147 8940 8980 -flto=thin -flto=auto -fno-fat-lto-objects -flto=thin 1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt Model: GPT-2 - Device: CPU - Executor: Standard
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: yolov4 - Device: CPU - Executor: Parallel AOCC 4.0 GCC 12.2 LLVM Clang 14 120 240 360 480 600 SE +/- 0.73, N = 3 SE +/- 0.29, N = 3 SE +/- 0.44, N = 3 569 568 563 -flto=thin -flto=auto -fno-fat-lto-objects -flto=thin 1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt Model: yolov4 - Device: CPU - Executor: Parallel
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: yolov4 - Device: CPU - Executor: Standard AOCC 4.0 GCC 12.2 LLVM Clang 14 120 240 360 480 600 SE +/- 3.33, N = 3 SE +/- 21.40, N = 9 SE +/- 5.06, N = 3 576 564 566 -flto=thin -flto=auto -fno-fat-lto-objects -flto=thin 1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt Model: yolov4 - Device: CPU - Executor: Standard
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: bertsquad-12 - Device: CPU - Executor: Parallel AOCC 4.0 GCC 12.2 LLVM Clang 14 200 400 600 800 1000 SE +/- 3.50, N = 3 SE +/- 1.09, N = 3 SE +/- 2.52, N = 3 917 918 923 -flto=thin -flto=auto -fno-fat-lto-objects -flto=thin 1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt Model: bertsquad-12 - Device: CPU - Executor: Parallel
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: fcn-resnet101-11 - Device: CPU - Executor: Parallel AOCC 4.0 GCC 12.2 LLVM Clang 14 30 60 90 120 150 SE +/- 0.58, N = 3 SE +/- 0.17, N = 3 SE +/- 0.50, N = 3 135 134 135 -flto=thin -flto=auto -fno-fat-lto-objects -flto=thin 1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt Model: fcn-resnet101-11 - Device: CPU - Executor: Parallel
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: ArcFace ResNet-100 - Device: CPU - Executor: Parallel AOCC 4.0 GCC 12.2 LLVM Clang 14 500 1000 1500 2000 2500 SE +/- 4.18, N = 3 SE +/- 3.09, N = 3 SE +/- 2.92, N = 3 2356 2361 2353 -flto=thin -flto=auto -fno-fat-lto-objects -flto=thin 1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt Model: ArcFace ResNet-100 - Device: CPU - Executor: Parallel
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard AOCC 4.0 GCC 12.2 LLVM Clang 14 600 1200 1800 2400 3000 SE +/- 23.09, N = 12 SE +/- 81.96, N = 12 SE +/- 34.72, N = 12 2207 2222 2781 -flto=thin -flto=auto -fno-fat-lto-objects -flto=thin 1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: super-resolution-10 - Device: CPU - Executor: Parallel AOCC 4.0 GCC 12.2 LLVM Clang 14 2K 4K 6K 8K 10K SE +/- 69.59, N = 3 SE +/- 88.15, N = 6 SE +/- 110.94, N = 3 9113 9395 8949 -flto=thin -flto=auto -fno-fat-lto-objects -flto=thin 1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt Model: super-resolution-10 - Device: CPU - Executor: Parallel
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory
GraphicsMagick This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample 6000x4000 pixel JPEG image. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.38 Operation: Swirl AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 300 600 900 1200 1500 SE +/- 2.96, N = 3 SE +/- 4.18, N = 3 SE +/- 2.19, N = 3 SE +/- 4.91, N = 3 SE +/- 2.33, N = 3 1413 1473 1454 1382 1390 1. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.38 Operation: Rotate AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 200 400 600 800 1000 SE +/- 1.20, N = 3 SE +/- 3.18, N = 3 SE +/- 2.89, N = 3 SE +/- 7.17, N = 3 SE +/- 0.88, N = 3 952 988 912 896 876 1. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.38 Operation: Sharpen AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 100 200 300 400 500 SE +/- 0.67, N = 3 SE +/- 0.58, N = 3 SE +/- 0.67, N = 3 SE +/- 2.73, N = 3 SE +/- 0.88, N = 3 470 343 341 466 454 1. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.38 Operation: Enhanced AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 140 280 420 560 700 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 SE +/- 0.67, N = 3 SE +/- 0.33, N = 3 612 639 627 582 583 1. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.38 Operation: Resizing AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 600 1200 1800 2400 3000 SE +/- 5.17, N = 3 SE +/- 2.52, N = 3 SE +/- 8.67, N = 3 SE +/- 3.79, N = 3 SE +/- 4.67, N = 3 2609 2741 2625 2705 2655 1. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.38 Operation: Noise-Gaussian AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 140 280 420 560 700 SE +/- 0.67, N = 3 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 SE +/- 2.03, N = 3 SE +/- 1.20, N = 3 607 664 658 587 583 1. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.38 Operation: HWB Color Space AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 400 800 1200 1600 2000 SE +/- 12.35, N = 3 SE +/- 1.33, N = 3 SE +/- 4.93, N = 3 SE +/- 9.17, N = 3 SE +/- 5.36, N = 3 1710 1533 1438 1580 1584 1. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread
Coremark This is a test of EEMBC CoreMark processor benchmark. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Iterations/Sec, More Is Better Coremark 1.0 CoreMark Size 666 - Iterations Per Second AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 200K 400K 600K 800K 1000K SE +/- 973.61, N = 3 SE +/- 799.86, N = 3 SE +/- 1075.50, N = 3 SE +/- 521.77, N = 3 SE +/- 368.05, N = 3 946032.29 1066150.95 1117817.25 981244.49 1004815.01 1. (CC) gcc options: -O2 -O3 -march=native -flto -lrt" -lrt
Xsbench XSBench is a mini-app representing a key computational kernel of the Monte Carlo neutronics application OpenMC. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Lookups/s, More Is Better Xsbench 2017-07-06 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 1.5M 3M 4.5M 6M 7.5M SE +/- 4600.85, N = 3 SE +/- 1443.86, N = 3 SE +/- 4489.29, N = 3 SE +/- 7500.25, N = 3 SE +/- 2015.53, N = 3 7081346 6234508 6318262 6756001 6752238 1. (CC) gcc options: -std=gnu99 -fopenmp -O3 -lm
SecureMark SecureMark is an objective, standardized benchmarking framework for measuring the efficiency of cryptographic processing solutions developed by EEMBC. SecureMark-TLS is benchmarking Transport Layer Security performance with a focus on IoT/edge computing. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org marks, More Is Better SecureMark 1.0.4 Benchmark: SecureMark-TLS AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 100K 200K 300K 400K 500K SE +/- 454.53, N = 3 SE +/- 2377.52, N = 3 SE +/- 1407.45, N = 3 SE +/- 535.99, N = 3 SE +/- 2486.77, N = 3 475941 432810 432896 472624 454957 1. (CC) gcc options: -pedantic -O3
C-Blosc C-Blosc (c-blosc2) simple, compressed, fast and persistent data store library for C. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MB/s, More Is Better C-Blosc 2.3 Test: blosclz shuffle AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 5K 10K 15K 20K 25K SE +/- 238.69, N = 3 SE +/- 90.86, N = 3 SE +/- 203.38, N = 3 SE +/- 81.78, N = 3 SE +/- 173.15, N = 3 23837.2 23526.4 24105.0 24796.8 24197.7
OpenBenchmarking.org MB/s, More Is Better C-Blosc 2.3 Test: blosclz bitshuffle AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 3K 6K 9K 12K 15K SE +/- 81.98, N = 3 SE +/- 76.49, N = 3 SE +/- 38.45, N = 3 SE +/- 152.49, N = 3 SE +/- 150.84, N = 3 13508.9 13017.3 13286.5 13989.2 13694.2
Zstd Compression This test measures the time needed to compress/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 3 - Compression Speed AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 1100 2200 3300 4400 5500 SE +/- 10.48, N = 3 SE +/- 14.32, N = 3 SE +/- 10.83, N = 3 SE +/- 13.86, N = 3 SE +/- 28.62, N = 3 5092.2 4978.6 4976.2 5045.2 5030.8 1. (CC) gcc options: -O3 -march=native -flto -pthread -lz -llzma
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 3 - Decompression Speed AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 1200 2400 3600 4800 6000 SE +/- 57.07, N = 3 SE +/- 8.08, N = 3 SE +/- 8.47, N = 3 SE +/- 59.43, N = 3 SE +/- 4.13, N = 3 5461.2 5432.1 5485.8 5463.5 5458.6 1. (CC) gcc options: -O3 -march=native -flto -pthread -lz -llzma
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 19 - Compression Speed AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 20 40 60 80 100 SE +/- 0.27, N = 3 SE +/- 0.52, N = 3 SE +/- 0.50, N = 3 SE +/- 0.07, N = 3 SE +/- 0.52, N = 3 79.1 80.3 82.2 78.9 78.8 1. (CC) gcc options: -O3 -march=native -flto -pthread -lz -llzma
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 19 - Decompression Speed AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 1100 2200 3300 4400 5500 SE +/- 41.19, N = 3 SE +/- 20.04, N = 3 SE +/- 72.80, N = 3 SE +/- 21.62, N = 3 SE +/- 29.56, N = 3 5018.6 4932.0 4988.9 4913.2 4894.6 1. (CC) gcc options: -O3 -march=native -flto -pthread -lz -llzma
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 19, Long Mode - Compression Speed AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 12 24 36 48 60 SE +/- 0.15, N = 3 SE +/- 0.17, N = 3 SE +/- 0.09, N = 3 SE +/- 0.12, N = 3 SE +/- 0.03, N = 3 54.6 54.6 54.3 54.6 55.3 1. (CC) gcc options: -O3 -march=native -flto -pthread -lz -llzma
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 19, Long Mode - Decompression Speed AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 1100 2200 3300 4400 5500 SE +/- 24.94, N = 3 SE +/- 49.55, N = 3 SE +/- 81.06, N = 3 SE +/- 33.82, N = 3 SE +/- 45.42, N = 3 5117.8 4974.4 5073.8 5055.8 4928.9 1. (CC) gcc options: -O3 -march=native -flto -pthread -lz -llzma
Nettle GNU Nettle is a low-level cryptographic library used by GnuTLS and other software. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Mbyte/s, More Is Better Nettle 3.8 Test: aes256 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 2K 4K 6K 8K 10K SE +/- 116.25, N = 3 SE +/- 53.19, N = 3 SE +/- 1.22, N = 3 SE +/- 47.83, N = 3 SE +/- 0.32, N = 3 8564.64 8521.41 8490.46 8566.17 8553.15 MIN: 6283.86 / MAX: 12924.64 -lhogweed - MIN: 6434.82 / MAX: 12801.5 -lhogweed - MIN: 6473.53 / MAX: 12523.15 MIN: 6405.01 / MAX: 12818.56 MIN: 6494.58 / MAX: 12673.87 1. (CC) gcc options: -O3 -march=native -flto -ggdb3 -lnettle -lgmp -lm -lcrypto
OpenBenchmarking.org Mbyte/s, More Is Better Nettle 3.8 Test: chacha AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 300 600 900 1200 1500 SE +/- 2.93, N = 3 SE +/- 9.98, N = 3 SE +/- 0.01, N = 3 SE +/- 10.11, N = 3 SE +/- 0.55, N = 3 1528.82 1515.12 1507.34 1536.84 1523.44 MIN: 773.94 / MAX: 4229.15 -lhogweed - MIN: 754.38 / MAX: 4226.47 -lhogweed - MIN: 757.05 / MAX: 4174.74 MIN: 773.42 / MAX: 4324.28 MIN: 773.19 / MAX: 4193.52 1. (CC) gcc options: -O3 -march=native -flto -ggdb3 -lnettle -lgmp -lm -lcrypto
OpenBenchmarking.org Mbyte/s, More Is Better Nettle 3.8 Test: sha512 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 200 400 600 800 1000 SE +/- 0.77, N = 3 SE +/- 0.38, N = 3 SE +/- 0.35, N = 3 SE +/- 0.60, N = 3 SE +/- 0.75, N = 3 864.29 858.30 859.96 864.87 863.73 -lhogweed -lhogweed 1. (CC) gcc options: -O3 -march=native -flto -ggdb3 -lnettle -lgmp -lm -lcrypto
OpenBenchmarking.org Mbyte/s, More Is Better Nettle 3.8 Test: poly1305-aes AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 900 1800 2700 3600 4500 SE +/- 5.55, N = 3 SE +/- 6.07, N = 3 SE +/- 18.65, N = 3 SE +/- 4.96, N = 3 SE +/- 1.66, N = 3 4255.91 4182.31 4188.14 4231.63 4214.74 -lhogweed -lhogweed 1. (CC) gcc options: -O3 -march=native -flto -ggdb3 -lnettle -lgmp -lm -lcrypto
libjpeg-turbo tjbench tjbench is a JPEG decompression/compression benchmark that is part of libjpeg-turbo, a JPEG image codec library optimized for SIMD instructions on modern CPU architectures. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Megapixels/sec, More Is Better libjpeg-turbo tjbench 2.1.0 Test: Decompression Throughput AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 70 140 210 280 350 SE +/- 0.79, N = 3 SE +/- 0.85, N = 3 SE +/- 0.46, N = 3 SE +/- 0.31, N = 3 SE +/- 0.23, N = 3 318.29 314.76 322.49 320.16 321.22 1. (CC) gcc options: -O3 -march=native -flto -rdynamic -lm
Sockperf This is a network socket API performance benchmark developed by Mellanox. This test profile runs both the client and server on the local host for evaluating individual system performance. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Messages Per Second, More Is Better Sockperf 3.7 Test: Throughput AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 200K 400K 600K 800K 1000K SE +/- 1831.17, N = 5 SE +/- 3167.67, N = 5 SE +/- 2684.21, N = 5 SE +/- 4602.76, N = 5 SE +/- 3955.76, N = 5 879095 889103 900310 883756 897661 1. (CXX) g++ options: --param -O3 -march=native -flto -rdynamic
QuantLib QuantLib is an open-source library/framework around quantitative finance for modeling, trading and risk management scenarios. QuantLib is written in C++ with Boost and its built-in benchmark used reports the QuantLib Benchmark Index benchmark score. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MFLOPS, More Is Better QuantLib 1.21 AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 1000 2000 3000 4000 5000 SE +/- 7.07, N = 3 SE +/- 52.72, N = 12 SE +/- 66.71, N = 3 SE +/- 83.36, N = 12 4532.3 4501.1 4669.5 4783.9 1. (CXX) g++ options: -O3 -march=native -rdynamic
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: unknown location(0): [4;31;49mfatal error: in "QuantLib benchmark suite/ShortRateModel__Swaps": QuantLib::Error: cannot roll the asset back to0 (it is already at t = 1.75068)[0;39;49m
Crypto++ Crypto++ is a C++ class library of cryptographic algorithms. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MiB/second, More Is Better Crypto++ 8.2 Test: Keyed Algorithms AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 200 400 600 800 1000 SE +/- 2.11, N = 3 SE +/- 1.11, N = 3 SE +/- 0.29, N = 3 SE +/- 3.44, N = 3 1048.65 881.35 1086.85 1075.79 1. (CXX) g++ options: -O3 -march=native -flto -fPIC -pthread -pipe Test: Keyed Algorithms
GCC 13.0 14 Nov: The test quit with a non-zero exit status.
OpenBenchmarking.org MiB/second, More Is Better Crypto++ 8.2 Test: Unkeyed Algorithms AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 130 260 390 520 650 SE +/- 0.53, N = 3 SE +/- 3.68, N = 3 SE +/- 5.89, N = 3 SE +/- 4.73, N = 3 591.18 481.76 593.21 596.87 1. (CXX) g++ options: -O3 -march=native -flto -fPIC -pthread -pipe Test: Unkeyed Algorithms
GCC 13.0 14 Nov: The test quit with a non-zero exit status.
JPEG XL libjxl The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MP/s, More Is Better JPEG XL libjxl 0.7 Input: PNG - Quality: 90 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 4 8 12 16 20 SE +/- 0.04, N = 3 SE +/- 0.05, N = 3 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 13.50 12.98 13.01 13.19 17.98 -Xclang -mrelax-all -Xclang -mrelax-all -Xclang -mrelax-all 1. (CXX) g++ options: -O3 -march=native -flto -fno-rtti -funwind-tables -O2 -fPIE -pie -latomic
OpenBenchmarking.org MP/s, More Is Better JPEG XL libjxl 0.7 Input: JPEG - Quality: 90 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 4 8 12 16 20 SE +/- 0.08, N = 3 SE +/- 0.04, N = 3 SE +/- 0.09, N = 3 SE +/- 0.11, N = 3 SE +/- 0.11, N = 3 13.13 12.54 12.76 13.10 17.46 -Xclang -mrelax-all -Xclang -mrelax-all -Xclang -mrelax-all 1. (CXX) g++ options: -O3 -march=native -flto -fno-rtti -funwind-tables -O2 -fPIE -pie -latomic
OpenBenchmarking.org MP/s, More Is Better JPEG XL libjxl 0.7 Input: PNG - Quality: 100 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.2498 0.4996 0.7494 0.9992 1.249 SE +/- 0.01, N = 3 SE +/- 0.01, N = 5 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 1.08 1.04 1.03 1.07 1.11 -Xclang -mrelax-all -Xclang -mrelax-all -Xclang -mrelax-all 1. (CXX) g++ options: -O3 -march=native -flto -fno-rtti -funwind-tables -O2 -fPIE -pie -latomic
OpenBenchmarking.org MP/s, More Is Better JPEG XL libjxl 0.7 Input: JPEG - Quality: 100 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.2115 0.423 0.6345 0.846 1.0575 SE +/- 0.01, N = 6 SE +/- 0.01, N = 3 SE +/- 0.02, N = 6 SE +/- 0.01, N = 3 SE +/- 0.01, N = 9 0.85 0.85 0.85 0.94 0.87 -Xclang -mrelax-all -Xclang -mrelax-all -Xclang -mrelax-all 1. (CXX) g++ options: -O3 -march=native -flto -fno-rtti -funwind-tables -O2 -fPIE -pie -latomic
JPEG XL Decoding libjxl The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is suited for JPEG XL decode performance testing to PNG output file, the pts/jpexl test is for encode performance. The JPEG XL encoding/decoding is done using the libjxl codebase. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MP/s, More Is Better JPEG XL Decoding libjxl 0.7 CPU Threads: 1 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 20 40 60 80 100 SE +/- 0.16, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 SE +/- 0.16, N = 3 SE +/- 0.04, N = 3 77.27 68.44 66.30 62.29 63.32
OpenBenchmarking.org MP/s, More Is Better JPEG XL Decoding libjxl 0.7 CPU Threads: All AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 60 120 180 240 300 SE +/- 2.69, N = 3 SE +/- 3.47, N = 3 SE +/- 1.42, N = 3 SE +/- 1.06, N = 3 SE +/- 1.97, N = 3 286.44 263.90 245.54 264.48 266.89
WebP Image Encode This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MP/s, More Is Better WebP Image Encode 1.2.4 Encode Settings: Default AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 7 14 21 28 35 SE +/- 0.02, N = 3 SE +/- 0.09, N = 3 SE +/- 0.07, N = 3 SE +/- 0.40, N = 3 SE +/- 0.05, N = 3 29.65 28.02 28.18 29.45 29.23 -lpng16 -ljpeg -lpng16 -ljpeg -ltiff -lpng16 -ljpeg -ltiff 1. (CC) gcc options: -fvisibility=hidden -O3 -march=native -flto -lm
OpenBenchmarking.org MP/s, More Is Better WebP Image Encode 1.2.4 Encode Settings: Quality 100 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 4 8 12 16 20 SE +/- 0.01, N = 3 SE +/- 0.11, N = 3 SE +/- 0.06, N = 3 SE +/- 0.09, N = 3 SE +/- 0.24, N = 3 17.67 16.61 16.80 17.51 17.00 -lpng16 -ljpeg -lpng16 -ljpeg -ltiff -lpng16 -ljpeg -ltiff 1. (CC) gcc options: -fvisibility=hidden -O3 -march=native -flto -lm
OpenBenchmarking.org MP/s, More Is Better WebP Image Encode 1.2.4 Encode Settings: Quality 100, Lossless AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.513 1.026 1.539 2.052 2.565 SE +/- 0.00, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 2.17 2.23 2.28 2.06 2.15 -lpng16 -ljpeg -lpng16 -ljpeg -ltiff -lpng16 -ljpeg -ltiff 1. (CC) gcc options: -fvisibility=hidden -O3 -march=native -flto -lm
OpenBenchmarking.org MP/s, More Is Better WebP Image Encode 1.2.4 Encode Settings: Quality 100, Highest Compression AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 2 4 6 8 10 SE +/- 0.00, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 6.33 4.48 4.60 6.16 6.21 -lpng16 -ljpeg -lpng16 -ljpeg -ltiff -lpng16 -ljpeg -ltiff 1. (CC) gcc options: -fvisibility=hidden -O3 -march=native -flto -lm
OpenBenchmarking.org MP/s, More Is Better WebP Image Encode 1.2.4 Encode Settings: Quality 100, Lossless, Highest Compression AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.2025 0.405 0.6075 0.81 1.0125 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 0.89 0.88 0.89 0.83 0.90 -lpng16 -ljpeg -lpng16 -ljpeg -ltiff -lpng16 -ljpeg -ltiff 1. (CC) gcc options: -fvisibility=hidden -O3 -march=native -flto -lm
ASTC Encoder ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 4.0 Preset: Fast AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 90 180 270 360 450 SE +/- 0.03, N = 3 SE +/- 1.00, N = 3 SE +/- 0.15, N = 3 SE +/- 0.36, N = 3 SE +/- 0.42, N = 3 427.33 289.29 294.57 404.45 412.21 1. (CXX) g++ options: -O3 -march=native -flto -pthread
OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 4.0 Preset: Medium AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 40 80 120 160 200 SE +/- 0.27, N = 3 SE +/- 0.92, N = 3 SE +/- 0.51, N = 3 SE +/- 0.21, N = 3 SE +/- 0.11, N = 3 160.96 132.59 135.19 155.09 156.45 1. (CXX) g++ options: -O3 -march=native -flto -pthread
OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 4.0 Preset: Thorough AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 5 10 15 20 25 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 19.91 19.45 19.77 19.28 19.43 1. (CXX) g++ options: -O3 -march=native -flto -pthread
OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 4.0 Preset: Exhaustive AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.4562 0.9124 1.3686 1.8248 2.281 SE +/- 0.0024, N = 3 SE +/- 0.0035, N = 3 SE +/- 0.0024, N = 3 SE +/- 0.0029, N = 3 SE +/- 0.0022, N = 3 2.0160 2.0071 2.0276 1.9814 1.9886 1. (CXX) g++ options: -O3 -march=native -flto -pthread
LeelaChessZero LeelaChessZero (lc0 / lczero) is a chess engine automated vian neural networks. This test profile can be used for OpenCL, CUDA + cuDNN, and BLAS (CPU-based) benchmarking. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Nodes Per Second, More Is Better LeelaChessZero 0.28 Backend: BLAS AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 400 800 1200 1600 2000 SE +/- 24.06, N = 3 SE +/- 18.37, N = 4 SE +/- 14.33, N = 3 SE +/- 18.88, N = 4 1748 1697 1702 1686 1. (CXX) g++ options: -flto -O3 -march=native -pthread Backend: BLAS
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./lczero: line 4: ./lc0: No such file or directory
OpenBenchmarking.org Nodes Per Second, More Is Better LeelaChessZero 0.28 Backend: Eigen AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 400 800 1200 1600 2000 SE +/- 19.60, N = 3 SE +/- 19.08, N = 3 SE +/- 17.46, N = 3 SE +/- 11.93, N = 3 1945 1725 1861 1863 1. (CXX) g++ options: -flto -O3 -march=native -pthread Backend: Eigen
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./lczero: line 4: ./lc0: No such file or directory
TSCP This is a performance test of TSCP, Tom Kerrigan's Simple Chess Program, which has a built-in performance benchmark. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Nodes Per Second, More Is Better TSCP 1.81 AI Chess Performance AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 600K 1200K 1800K 2400K 3000K SE +/- 18536.62, N = 5 SE +/- 7641.62, N = 5 SE +/- 3566.28, N = 5 SE +/- 17050.36, N = 5 SE +/- 14422.21, N = 5 2477063 2496821 2492228 2942579 2857022 1. (CC) gcc options: -O3 -march=native -flto
Dragonflydb Dragonfly is an open-source database server that is a "modern Redis replacement" that aims to be the fastest memory store while being compliant with the Redis and Memcached protocols. For benchmarking Dragonfly, Memtier_benchmark is used as a NoSQL Redis/Memcache traffic generation plus benchmarking tool developed by Redis Labs. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Ops/sec, More Is Better Dragonflydb 0.6 Clients: 50 - Set To Get Ratio: 1:5 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 1.2M 2.4M 3.6M 4.8M 6M SE +/- 24619.72, N = 3 SE +/- 28018.78, N = 3 SE +/- 37756.01, N = 3 5728724.37 5707956.74 5757620.86 1. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre Clients: 50 - Set To Get Ratio: 1:5
LLVM Clang 14: The test run did not produce a result. E: ./dragonflydb: 7: ./memtier_benchmark: not found
LLVM Clang 15.0.2: The test run did not produce a result. E: ./dragonflydb: 7: ./memtier_benchmark: not found
OpenBenchmarking.org Ops/sec, More Is Better Dragonflydb 0.6 Clients: 50 - Set To Get Ratio: 5:1 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 1.1M 2.2M 3.3M 4.4M 5.5M SE +/- 18465.52, N = 3 SE +/- 11820.30, N = 3 SE +/- 10081.64, N = 3 5242954.59 5275879.83 5271140.49 1. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre Clients: 50 - Set To Get Ratio: 5:1
LLVM Clang 14: The test run did not produce a result. E: ./dragonflydb: 7: ./memtier_benchmark: not found
LLVM Clang 15.0.2: The test run did not produce a result. E: ./dragonflydb: 7: ./memtier_benchmark: not found
Redis Redis is an open-source in-memory data structure store, used as a database, cache, and message broker. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Requests Per Second, More Is Better Redis 7.0.4 Test: SET - Parallel Connections: 50 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 1000K 2000K 3000K 4000K 5000K SE +/- 43555.06, N = 15 SE +/- 56367.98, N = 15 SE +/- 63650.32, N = 15 4701871.50 4498344.35 4675697.00 1. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3 -march=native -flto Test: SET - Parallel Connections: 50
LLVM Clang 14: The test run did not produce a result. E: ./redis: 12: ./src/redis-benchmark: not found
LLVM Clang 15.0.2: The test run did not produce a result. E: ./redis: 12: ./src/redis-benchmark: not found
PJSIP PJSIP is a free and open source multimedia communication library written in C language implementing standard based protocols such as SIP, SDP, RTP, STUN, TURN, and ICE. It combines signaling protocol (SIP) with rich multimedia framework and NAT traversal functionality into high level API that is portable and suitable for almost any type of systems ranging from desktops, embedded systems, to mobile handsets. This test profile is making use of pjsip-perf with both the client/server on teh system. More details on the PJSIP benchmark at https://www.pjsip.org/high-performance-sip.htm Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Responses Per Second, More Is Better PJSIP 2.11 Method: INVITE AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 1100 2200 3300 4400 5500 SE +/- 64.01, N = 3 SE +/- 5.24, N = 3 SE +/- 57.10, N = 5 5344 5279 5338 1. (CC) gcc options: -lSDL2 -lavformat -lavcodec -lswscale -lavutil -lstdc++ -lopus -lssl -lcrypto -luuid -lm -lrt -lpthread -lasound -O3 -march=native -flto Method: INVITE
LLVM Clang 14: The test quit with a non-zero exit status. E: ./pjsip: line 5: ./pjsip-apps/bin/samples/*/pjsip-perf: No such file or directory
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./pjsip: line 5: ./pjsip-apps/bin/samples/*/pjsip-perf: No such file or directory
OpenBenchmarking.org Responses Per Second, More Is Better PJSIP 2.11 Method: OPTIONS, Stateful AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 2K 4K 6K 8K 10K SE +/- 19.14, N = 3 SE +/- 17.01, N = 3 SE +/- 37.24, N = 3 9370 9330 9389 1. (CC) gcc options: -lSDL2 -lavformat -lavcodec -lswscale -lavutil -lstdc++ -lopus -lssl -lcrypto -luuid -lm -lrt -lpthread -lasound -O3 -march=native -flto Method: OPTIONS, Stateful
LLVM Clang 14: The test quit with a non-zero exit status. E: ./pjsip: line 5: ./pjsip-apps/bin/samples/*/pjsip-perf: No such file or directory
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./pjsip: line 5: ./pjsip-apps/bin/samples/*/pjsip-perf: No such file or directory
OpenBenchmarking.org Responses Per Second, More Is Better PJSIP 2.11 Method: OPTIONS, Stateless AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov 30K 60K 90K 120K 150K SE +/- 515.14, N = 3 SE +/- 794.35, N = 3 SE +/- 236.11, N = 3 124523 125825 124028 1. (CC) gcc options: -lSDL2 -lavformat -lavcodec -lswscale -lavutil -lstdc++ -lopus -lssl -lcrypto -luuid -lm -lrt -lpthread -lasound -O3 -march=native -flto Method: OPTIONS, Stateless
LLVM Clang 14: The test quit with a non-zero exit status. E: ./pjsip: line 5: ./pjsip-apps/bin/samples/*/pjsip-perf: No such file or directory
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./pjsip: line 5: ./pjsip-apps/bin/samples/*/pjsip-perf: No such file or directory
Liquid-DSP LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 2021.01.31 Threads: 8 - Buffer Length: 256 - Filter Length: 57 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 200M 400M 600M 800M 1000M SE +/- 6969075.34, N = 3 SE +/- 1506567.11, N = 3 SE +/- 2781883.85, N = 3 SE +/- 475966.85, N = 3 SE +/- 84936093.08, N = 12 896243333 834506667 650876667 809823333 566352750 1. (CC) gcc options: -O3 -march=native -flto -pthread -lm -lc -lliquid
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 2021.01.31 Threads: 16 - Buffer Length: 256 - Filter Length: 57 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 400M 800M 1200M 1600M 2000M SE +/- 5356097.25, N = 3 SE +/- 6222093.25, N = 3 SE +/- 2273274.68, N = 3 SE +/- 1993600.87, N = 3 SE +/- 883804.91, N = 3 1754833333 1627233333 1274433333 1584066667 1544266667 1. (CC) gcc options: -O3 -march=native -flto -pthread -lm -lc -lliquid
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 2021.01.31 Threads: 32 - Buffer Length: 256 - Filter Length: 57 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 400M 800M 1200M 1600M 2000M SE +/- 2450170.06, N = 3 SE +/- 1289702.81, N = 3 SE +/- 4147824.06, N = 3 SE +/- 721110.26, N = 3 SE +/- 983756.97, N = 3 1886600000 1680600000 1633166667 1723600000 1654933333 1. (CC) gcc options: -O3 -march=native -flto -pthread -lm -lc -lliquid
OpenSSL OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org sign/s, More Is Better OpenSSL 3.0 Algorithm: RSA4096 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 1300 2600 3900 5200 6500 SE +/- 4.01, N = 3 SE +/- 1.50, N = 3 SE +/- 0.27, N = 3 SE +/- 10.70, N = 3 SE +/- 0.55, N = 3 6026.2 6014.6 6029.1 6016.6 6030.9 -Qunused-arguments -Qunused-arguments -Qunused-arguments 1. (CC) gcc options: -pthread -m64 -O3 -march=native -flto -lssl -lcrypto -ldl
CLOMP CLOMP is the C version of the Livermore OpenMP benchmark developed to measure OpenMP overheads and other performance impacts due to threading in order to influence future system designs. This particular test profile configuration is currently set to look at the OpenMP static schedule speed-up across all available CPU cores using the recommended test configuration. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Speedup, More Is Better CLOMP 1.2 Static OMP Speedup AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 6 12 18 24 30 SE +/- 0.27, N = 5 SE +/- 0.27, N = 3 SE +/- 0.03, N = 3 SE +/- 0.23, N = 6 SE +/- 0.15, N = 3 25.0 22.8 22.6 23.7 23.6 1. (CC) gcc options: -fopenmp -O3 -march=native -flto -lm
Kripke Kripke is a simple, scalable, 3D Sn deterministic particle transport code. Its primary purpose is to research how data layout, programming paradigms and architectures effect the implementation and performance of Sn transport. Kripke is developed by LLNL. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Throughput FoM, More Is Better Kripke 1.2.4 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 30M 60M 90M 120M 150M SE +/- 1766612.42, N = 3 SE +/- 1385510.65, N = 15 SE +/- 1504986.40, N = 4 SE +/- 1260500.21, N = 4 SE +/- 1079837.67, N = 3 125045767 133327173 132057600 109794900 108977267 -fopenmp=libomp -fopenmp -fopenmp -fopenmp=libomp -fopenmp=libomp 1. (CXX) g++ options: -O3 -march=native -flto
OpenSSL OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org verify/s, More Is Better OpenSSL 3.0 Algorithm: RSA4096 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 80K 160K 240K 320K 400K SE +/- 12.58, N = 3 SE +/- 54.65, N = 3 SE +/- 44.62, N = 3 SE +/- 38.98, N = 3 SE +/- 41.04, N = 3 394139.1 393865.8 394339.8 395258.1 395636.4 -Qunused-arguments -Qunused-arguments -Qunused-arguments 1. (CC) gcc options: -pthread -m64 -O3 -march=native -flto -lssl -lcrypto -ldl
Caffe This is a benchmark of the Caffe deep learning framework and currently supports the AlexNet and Googlenet model and execution on both CPUs and NVIDIA GPUs. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Milli-Seconds, Fewer Is Better Caffe 2020-02-13 Model: AlexNet - Acceleration: CPU - Iterations: 100 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 10K 20K 30K 40K 50K SE +/- 59.47, N = 3 SE +/- 77.62, N = 3 SE +/- 57.54, N = 3 SE +/- 157.80, N = 3 SE +/- 113.58, N = 3 27653 23685 23696 46976 46800 1. (CXX) g++ options: -O3 -march=native -flto -fPIC -rdynamic -lglog -lgflags -lprotobuf -lcrypto -lcurl -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas
OpenBenchmarking.org Milli-Seconds, Fewer Is Better Caffe 2020-02-13 Model: GoogleNet - Acceleration: CPU - Iterations: 100 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 20K 40K 60K 80K 100K SE +/- 160.54, N = 3 SE +/- 206.69, N = 3 SE +/- 221.71, N = 3 SE +/- 77.99, N = 3 SE +/- 326.67, N = 3 73934 65742 64987 104942 107414 1. (CXX) g++ options: -O3 -march=native -flto -fPIC -rdynamic -lglog -lgflags -lprotobuf -lcrypto -lcurl -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas
oneDNN OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.118 0.236 0.354 0.472 0.59 SE +/- 0.000214, N = 3 SE +/- 0.002131, N = 3 SE +/- 0.029304, N = 15 SE +/- 0.000115, N = 3 SE +/- 0.000600, N = 3 0.338581 0.479906 0.524283 0.343595 0.343749 -fopenmp=libomp - MIN: 0.33 -fopenmp - MIN: 0.38 -fopenmp -lpthread - MIN: 0.35 -fopenmp=libomp - MIN: 0.33 -fopenmp=libomp - MIN: 0.33 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.0815 0.163 0.2445 0.326 0.4075 SE +/- 0.002915, N = 3 SE +/- 0.003679, N = 15 SE +/- 0.003020, N = 15 SE +/- 0.004019, N = 3 SE +/- 0.001605, N = 3 0.347017 0.356546 0.356407 0.337760 0.362285 -fopenmp=libomp - MIN: 0.31 -fopenmp - MIN: 0.3 -fopenmp -lpthread - MIN: 0.3 -fopenmp=libomp - MIN: 0.31 -fopenmp=libomp - MIN: 0.32 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.1633 0.3266 0.4899 0.6532 0.8165 SE +/- 0.000272, N = 3 SE +/- 0.034180, N = 12 SE +/- 0.016807, N = 15 SE +/- 0.000668, N = 3 SE +/- 0.000306, N = 3 0.577834 0.725978 0.662905 0.583377 0.582381 -fopenmp=libomp - MIN: 0.56 -fopenmp - MIN: 0.58 -fopenmp -lpthread - MIN: 0.58 -fopenmp=libomp - MIN: 0.57 -fopenmp=libomp - MIN: 0.57 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.3474 0.6948 1.0422 1.3896 1.737 SE +/- 0.00757, N = 3 SE +/- 0.00876, N = 3 SE +/- 0.00799, N = 3 SE +/- 0.01741, N = 3 SE +/- 0.01491, N = 3 1.50016 1.54379 1.53305 1.51176 1.51369 -fopenmp=libomp - MIN: 1.45 -fopenmp - MIN: 1.46 -fopenmp -lpthread - MIN: 1.46 -fopenmp=libomp - MIN: 1.45 -fopenmp=libomp - MIN: 1.45 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 1.1988 2.3976 3.5964 4.7952 5.994 SE +/- 0.00469, N = 3 SE +/- 0.00364, N = 3 SE +/- 0.00740, N = 3 SE +/- 0.01758, N = 3 SE +/- 0.01647, N = 3 5.28785 5.32780 5.30501 5.31930 5.30337 -fopenmp=libomp - MIN: 5.23 -fopenmp - MIN: 5.25 -fopenmp -lpthread - MIN: 5.23 -fopenmp=libomp - MIN: 5.24 -fopenmp=libomp - MIN: 5.24 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.0948 0.1896 0.2844 0.3792 0.474 SE +/- 0.000864, N = 3 SE +/- 0.000200, N = 3 SE +/- 0.000894, N = 3 SE +/- 0.000027, N = 3 SE +/- 0.000120, N = 3 0.405097 0.421419 0.418860 0.408432 0.408594 -fopenmp=libomp - MIN: 0.39 -fopenmp - MIN: 0.41 -fopenmp -lpthread - MIN: 0.4 -fopenmp=libomp - MIN: 0.39 -fopenmp=libomp - MIN: 0.4 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 200 400 600 800 1000 SE +/- 0.74, N = 3 SE +/- 2.28, N = 3 SE +/- 1.14, N = 3 SE +/- 0.92, N = 3 SE +/- 0.82, N = 3 1138.83 1140.34 1139.64 1138.26 1137.49 -fopenmp=libomp - MIN: 1134.32 -fopenmp - MIN: 1132.99 -fopenmp -lpthread - MIN: 1133.46 -fopenmp=libomp - MIN: 1134.02 -fopenmp=libomp - MIN: 1132.69 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.7795 1.559 2.3385 3.118 3.8975 SE +/- 0.01463, N = 3 SE +/- 0.00041, N = 3 SE +/- 0.00889, N = 3 SE +/- 0.00364, N = 3 SE +/- 0.03887, N = 3 3.27367 3.36412 3.42686 3.39581 3.46444 -fopenmp=libomp - MIN: 3.13 -fopenmp - MIN: 3.21 -fopenmp -lpthread - MIN: 3.24 -fopenmp=libomp - MIN: 3.3 -fopenmp=libomp - MIN: 3.26 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.3167 0.6334 0.9501 1.2668 1.5835 SE +/- 0.00291, N = 3 SE +/- 0.00253, N = 3 SE +/- 0.01473, N = 13 SE +/- 0.00660, N = 3 SE +/- 0.00184, N = 3 1.38455 1.38696 1.40503 1.40766 1.39579 -fopenmp=libomp - MIN: 1.33 -fopenmp - MIN: 1.35 -fopenmp -lpthread - MIN: 1.34 -fopenmp=libomp - MIN: 1.35 -fopenmp=libomp - MIN: 1.35 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 130 260 390 520 650 SE +/- 0.21, N = 3 SE +/- 2.60, N = 3 SE +/- 1.15, N = 3 SE +/- 0.44, N = 3 SE +/- 0.73, N = 3 574.01 580.33 582.00 575.60 575.13 -fopenmp=libomp - MIN: 571.16 -fopenmp - MIN: 573.07 -fopenmp -lpthread - MIN: 575.92 -fopenmp=libomp - MIN: 571.93 -fopenmp=libomp - MIN: 571.75 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 200 400 600 800 1000 SE +/- 1.07, N = 3 SE +/- 0.51, N = 3 SE +/- 0.77, N = 3 SE +/- 1.57, N = 3 SE +/- 1.59, N = 3 1139.44 1139.55 1142.22 1142.95 1142.00 -fopenmp=libomp - MIN: 1134.21 -fopenmp - MIN: 1134.18 -fopenmp -lpthread - MIN: 1135.79 -fopenmp=libomp - MIN: 1136.51 -fopenmp=libomp - MIN: 1135.13 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 130 260 390 520 650 SE +/- 0.14, N = 3 SE +/- 1.50, N = 3 SE +/- 2.68, N = 3 SE +/- 0.11, N = 3 SE +/- 0.79, N = 3 574.91 578.82 578.96 575.94 576.22 -fopenmp=libomp - MIN: 572.07 -fopenmp - MIN: 572.91 -fopenmp -lpthread - MIN: 571.53 -fopenmp=libomp - MIN: 573.26 -fopenmp=libomp - MIN: 572.91 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.0316 0.0632 0.0948 0.1264 0.158 SE +/- 0.000006, N = 3 SE +/- 0.000173, N = 3 SE +/- 0.000031, N = 3 SE +/- 0.000110, N = 3 SE +/- 0.000185, N = 3 0.133169 0.139862 0.140349 0.136343 0.136310 -fopenmp=libomp - MIN: 0.13 -fopenmp - MIN: 0.13 -fopenmp -lpthread - MIN: 0.13 -fopenmp=libomp - MIN: 0.13 -fopenmp=libomp - MIN: 0.13 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.7 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.0514 0.1028 0.1542 0.2056 0.257 SE +/- 0.000436, N = 3 SE +/- 0.000273, N = 3 SE +/- 0.000270, N = 3 SE +/- 0.000409, N = 3 SE +/- 0.000141, N = 3 0.222378 0.228514 0.227918 0.224101 0.224449 -fopenmp=libomp - MIN: 0.21 -fopenmp - MIN: 0.21 -fopenmp -lpthread - MIN: 0.22 -fopenmp=libomp - MIN: 0.22 -fopenmp=libomp - MIN: 0.22 1. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl
OpenJPEG OpenJPEG is an open-source JPEG 2000 codec written in the C programming language. The default input for this test profile is the NASA/JPL-Caltech/MSSS Curiosity panorama 717MB TIFF image file converting to JPEG2000 format. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better OpenJPEG 2.4 Encode: NASA Curiosity Panorama M34 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 16K 32K 48K 64K 80K SE +/- 115.52, N = 3 SE +/- 575.75, N = 15 SE +/- 357.45, N = 3 SE +/- 566.51, N = 7 SE +/- 79.04, N = 3 73991 71222 69608 64416 62689 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic
Google Draco Draco is a library developed by Google for compressing/decompressing 3D geometric meshes and point clouds. This test profile uses some Artec3D PLY models as the sample 3D model input formats for Draco compression/decompression. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better Google Draco 1.5.0 Model: Lion AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 800 1600 2400 3200 4000 SE +/- 5.51, N = 3 SE +/- 30.51, N = 3 SE +/- 52.92, N = 3 SE +/- 22.07, N = 3 3373 3486 3763 3700 1. (CXX) g++ options: -O3 -march=native -flto Model: Lion
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./draco: 3: ./draco_encoder: not found
OpenBenchmarking.org ms, Fewer Is Better Google Draco 1.5.0 Model: Church Facade AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 1000 2000 3000 4000 5000 SE +/- 7.54, N = 3 SE +/- 1.00, N = 3 SE +/- 8.51, N = 3 SE +/- 47.16, N = 6 4324 4424 4782 4649 1. (CXX) g++ options: -O3 -march=native -flto Model: Church Facade
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./draco: 3: ./draco_encoder: not found
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 20220729 Target: CPU - Model: mobilenet AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 3 6 9 12 15 SE +/- 0.11, N = 3 SE +/- 0.08, N = 3 SE +/- 0.11, N = 3 SE +/- 0.09, N = 6 SE +/- 0.10, N = 5 8.57 8.68 8.68 9.09 9.24 -lomp - MIN: 8.29 / MAX: 9.54 -lgomp - MIN: 8.53 / MAX: 9.96 -lgomp - MIN: 8.49 / MAX: 11.83 MIN: 8.74 / MAX: 10.31 MIN: 8.92 / MAX: 30.18 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU-v2-v2 - Model: mobilenet-v2 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.837 1.674 2.511 3.348 4.185 SE +/- 0.07, N = 3 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 6 SE +/- 0.03, N = 5 3.17 3.71 3.72 3.44 3.53 -lomp - MIN: 3.04 / MAX: 4.89 -lgomp - MIN: 3.64 / MAX: 5.07 -lgomp - MIN: 3.64 / MAX: 4.37 MIN: 3.31 / MAX: 4.22 MIN: 3.4 / MAX: 4.36 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU-v3-v3 - Model: mobilenet-v3 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.7335 1.467 2.2005 2.934 3.6675 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 6 SE +/- 0.05, N = 5 2.69 3.24 3.26 2.99 2.99 -lomp - MIN: 2.6 / MAX: 3.34 -lgomp - MIN: 3.18 / MAX: 4.64 -lgomp - MIN: 3.19 / MAX: 3.89 MIN: 2.9 / MAX: 3.96 MIN: 2.85 / MAX: 4.22 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: shufflenet-v2 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.846 1.692 2.538 3.384 4.23 SE +/- 0.07, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.05, N = 6 SE +/- 0.03, N = 5 2.96 3.76 3.74 3.25 3.31 -lomp - MIN: 2.81 / MAX: 3.55 -lgomp - MIN: 3.69 / MAX: 4.45 -lgomp - MIN: 3.68 / MAX: 12.3 MIN: 3.05 / MAX: 3.95 MIN: 3.2 / MAX: 4.88 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: mnasnet AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.747 1.494 2.241 2.988 3.735 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 6 SE +/- 0.04, N = 5 2.72 3.30 3.32 3.02 3.07 -lomp - MIN: 2.65 / MAX: 5.37 -lgomp - MIN: 3.26 / MAX: 3.92 -lgomp - MIN: 3.26 / MAX: 4.03 MIN: 2.88 / MAX: 3.77 MIN: 2.98 / MAX: 3.82 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: efficientnet-b0 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 1.044 2.088 3.132 4.176 5.22 SE +/- 0.07, N = 3 SE +/- 0.11, N = 3 SE +/- 0.07, N = 3 SE +/- 0.03, N = 6 SE +/- 0.05, N = 5 3.85 4.64 4.57 4.25 4.35 -lomp - MIN: 3.73 / MAX: 4.7 -lgomp - MIN: 4.46 / MAX: 16.81 -lgomp - MIN: 4.45 / MAX: 5.39 MIN: 4.09 / MAX: 5.43 MIN: 4.17 / MAX: 17.27 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: blazeface AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.3578 0.7156 1.0734 1.4312 1.789 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 6 SE +/- 0.02, N = 5 1.21 1.59 1.58 1.37 1.41 -lomp - MIN: 1.18 / MAX: 1.94 -lgomp - MIN: 1.55 / MAX: 2.26 -lgomp - MIN: 1.55 / MAX: 2.27 MIN: 1.32 / MAX: 2.09 MIN: 1.33 / MAX: 1.84 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: googlenet AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 2 4 6 8 10 SE +/- 0.15, N = 3 SE +/- 0.13, N = 3 SE +/- 0.14, N = 3 SE +/- 0.08, N = 6 SE +/- 0.15, N = 5 7.85 8.66 8.65 8.54 8.63 -lomp - MIN: 7.47 / MAX: 9.41 -lgomp - MIN: 8.41 / MAX: 9.8 -lgomp - MIN: 8.39 / MAX: 10.25 MIN: 8.23 / MAX: 15.46 MIN: 8.24 / MAX: 9.92 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: vgg16 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 6 12 18 24 30 SE +/- 0.31, N = 3 SE +/- 0.26, N = 3 SE +/- 0.36, N = 3 SE +/- 0.42, N = 6 SE +/- 0.34, N = 5 24.83 24.56 24.39 25.96 25.52 -lomp - MIN: 24.13 / MAX: 31.13 -lgomp - MIN: 24.02 / MAX: 31.29 -lgomp - MIN: 23.8 / MAX: 30.07 MIN: 24.5 / MAX: 34.87 MIN: 24.44 / MAX: 33.71 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: resnet18 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 2 4 6 8 10 SE +/- 0.05, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 SE +/- 0.08, N = 6 SE +/- 0.20, N = 5 7.18 7.49 7.49 7.36 7.41 -lomp - MIN: 6.98 / MAX: 8.34 -lgomp - MIN: 7.3 / MAX: 8.43 -lgomp - MIN: 7.32 / MAX: 8.85 MIN: 7.12 / MAX: 8.7 MIN: 7.08 / MAX: 10.22 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: alexnet AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 1.1678 2.3356 3.5034 4.6712 5.839 SE +/- 0.08, N = 3 SE +/- 0.15, N = 3 SE +/- 0.13, N = 3 SE +/- 0.01, N = 6 SE +/- 0.04, N = 5 4.79 4.99 4.98 5.19 5.16 -lomp - MIN: 4.59 / MAX: 5.88 -lgomp - MIN: 4.76 / MAX: 12.84 -lgomp - MIN: 4.76 / MAX: 5.88 MIN: 5.05 / MAX: 6.8 MIN: 4.97 / MAX: 7.15 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: resnet50 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 3 6 9 12 15 SE +/- 0.11, N = 3 SE +/- 0.19, N = 3 SE +/- 0.23, N = 3 SE +/- 0.18, N = 6 SE +/- 0.24, N = 5 11.62 12.25 12.26 12.32 12.32 -lomp - MIN: 11.29 / MAX: 13.29 -lgomp - MIN: 11.82 / MAX: 13.43 -lgomp - MIN: 11.85 / MAX: 14.07 MIN: 11.77 / MAX: 22.76 MIN: 11.77 / MAX: 14.07 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: yolov4-tiny AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 4 8 12 16 20 SE +/- 0.06, N = 3 SE +/- 0.06, N = 3 SE +/- 0.07, N = 3 SE +/- 0.08, N = 6 SE +/- 0.22, N = 5 14.38 14.03 14.12 14.68 14.94 -lomp - MIN: 14.16 / MAX: 15.32 -lgomp - MIN: 13.79 / MAX: 15.31 -lgomp - MIN: 13.91 / MAX: 15.48 MIN: 14.35 / MAX: 24.68 MIN: 14.52 / MAX: 17.6 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: squeezenet_ssd AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 3 6 9 12 15 SE +/- 0.21, N = 3 SE +/- 0.09, N = 3 SE +/- 0.02, N = 3 SE +/- 0.10, N = 6 SE +/- 0.11, N = 5 11.27 12.43 12.27 11.80 11.97 -lomp - MIN: 10.85 / MAX: 12.53 -lgomp - MIN: 12.2 / MAX: 13.43 -lgomp - MIN: 12.11 / MAX: 13.52 MIN: 11.31 / MAX: 28.92 MIN: 11.52 / MAX: 21.23 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
OpenBenchmarking.org ms, Fewer Is Better NCNN 20220729 Target: CPU - Model: regnety_400m AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 3 6 9 12 15 SE +/- 0.13, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.09, N = 6 SE +/- 0.12, N = 5 7.55 12.01 11.84 9.41 9.62 -lomp - MIN: 7.32 / MAX: 8.83 -lgomp - MIN: 11.85 / MAX: 18.93 -lgomp - MIN: 11.71 / MAX: 18.24 MIN: 9.11 / MAX: 27.31 MIN: 9.1 / MAX: 10.59 1. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread
TNN TNN is an open-source deep learning reasoning framework developed by Tencent. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better TNN 0.3 Target: CPU - Model: DenseNet AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 1000 2000 3000 4000 5000 SE +/- 0.57, N = 3 SE +/- 50.48, N = 3 SE +/- 1.92, N = 3 SE +/- 6.75, N = 3 2455.64 4513.71 3634.33 2499.82 -fopenmp=libomp - MIN: 2405.98 / MAX: 2515.95 -fopenmp - MIN: 4338.8 / MAX: 4643.28 -fopenmp=libomp - MIN: 3566.47 / MAX: 3716.42 -fopenmp=libomp - MIN: 2420.99 / MAX: 2590.77 1. (CXX) g++ options: -O3 -march=native -flto -pthread -fvisibility=hidden -fvisibility=default -rdynamic -ldl Target: CPU - Model: DenseNet
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./tnn: 3: ./test/TNNTest: not found
OpenBenchmarking.org ms, Fewer Is Better TNN 0.3 Target: CPU - Model: MobileNet v2 AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 90 180 270 360 450 SE +/- 0.24, N = 3 SE +/- 0.97, N = 3 SE +/- 1.19, N = 3 SE +/- 0.07, N = 3 244.05 412.94 431.56 337.48 -fopenmp=libomp - MIN: 243.09 / MAX: 248.96 -fopenmp - MIN: 410.59 / MAX: 431.69 -fopenmp=libomp - MIN: 428.32 / MAX: 434.62 -fopenmp=libomp - MIN: 336.45 / MAX: 345.73 1. (CXX) g++ options: -O3 -march=native -flto -pthread -fvisibility=hidden -fvisibility=default -rdynamic -ldl Target: CPU - Model: MobileNet v2
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./tnn: 3: ./test/TNNTest: not found
OpenBenchmarking.org ms, Fewer Is Better TNN 0.3 Target: CPU - Model: SqueezeNet v2 AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 16 32 48 64 80 SE +/- 0.41, N = 12 SE +/- 0.89, N = 3 SE +/- 0.37, N = 15 SE +/- 0.31, N = 15 44.59 69.92 61.11 49.43 -fopenmp=libomp - MIN: 40.15 / MAX: 45.73 -fopenmp - MIN: 68.11 / MAX: 71.55 -fopenmp=libomp - MIN: 56.49 / MAX: 63.12 -fopenmp=libomp - MIN: 45.34 / MAX: 50.65 1. (CXX) g++ options: -O3 -march=native -flto -pthread -fvisibility=hidden -fvisibility=default -rdynamic -ldl Target: CPU - Model: SqueezeNet v2
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./tnn: 3: ./test/TNNTest: not found
OpenBenchmarking.org ms, Fewer Is Better TNN 0.3 Target: CPU - Model: SqueezeNet v1.1 AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 80 160 240 320 400 SE +/- 1.12, N = 3 SE +/- 0.23, N = 3 SE +/- 0.06, N = 3 SE +/- 1.42, N = 3 217.84 388.44 302.02 232.44 -fopenmp=libomp - MIN: 215.42 / MAX: 219.12 -fopenmp - MIN: 387.68 / MAX: 389.12 -fopenmp=libomp - MIN: 301.67 / MAX: 304.24 -fopenmp=libomp - MIN: 230.61 / MAX: 241.13 1. (CXX) g++ options: -O3 -march=native -flto -pthread -fvisibility=hidden -fvisibility=default -rdynamic -ldl Target: CPU - Model: SqueezeNet v1.1
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./tnn: 3: ./test/TNNTest: not found
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.2.dev Model: Face Detection FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 120 240 360 480 600 SE +/- 1.12, N = 3 SE +/- 1.81, N = 3 SE +/- 2.19, N = 3 SE +/- 2.14, N = 3 557.01 557.36 555.00 555.05 MIN: 283.75 / MAX: 603.22 -fno-strict-overflow -fwrapv - MIN: 527.31 / MAX: 601 MIN: 537.69 / MAX: 616.06 MIN: 522.63 / MAX: 589.63 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.2.dev Model: Person Detection FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 200 400 600 800 1000 SE +/- 2.64, N = 3 SE +/- 1.87, N = 3 SE +/- 0.94, N = 3 SE +/- 4.13, N = 3 1055.29 1060.87 1066.07 1065.31 MIN: 875.98 / MAX: 1296.48 -fno-strict-overflow -fwrapv - MIN: 672.13 / MAX: 1319.18 MIN: 718.92 / MAX: 1326.23 MIN: 636.05 / MAX: 1337.86 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.2.dev Model: Person Detection FP32 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 200 400 600 800 1000 SE +/- 3.44, N = 3 SE +/- 7.82, N = 3 SE +/- 5.67, N = 3 SE +/- 5.64, N = 3 1081.01 1085.33 1092.87 1071.50 MIN: 580.39 / MAX: 1284.59 -fno-strict-overflow -fwrapv - MIN: 731.35 / MAX: 1302.96 MIN: 753 / MAX: 1274.14 MIN: 726.76 / MAX: 1318.28 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.2.dev Model: Vehicle Detection FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 3 6 9 12 15 SE +/- 0.09, N = 3 SE +/- 0.02, N = 3 SE +/- 0.06, N = 3 SE +/- 0.04, N = 3 10.62 10.55 10.62 10.76 MIN: 4 / MAX: 26.8 -fno-strict-overflow -fwrapv - MIN: 4.88 / MAX: 26.32 MIN: 3.86 / MAX: 26.67 MIN: 5.08 / MAX: 25.94 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.2.dev Model: Face Detection FP16-INT8 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 60 120 180 240 300 SE +/- 0.22, N = 3 SE +/- 0.11, N = 3 SE +/- 0.93, N = 3 SE +/- 1.08, N = 3 283.07 284.35 283.72 282.59 MIN: 145.95 / MAX: 325.14 -fno-strict-overflow -fwrapv - MIN: 250.18 / MAX: 318.64 MIN: 146.73 / MAX: 325.51 MIN: 216.72 / MAX: 375.59 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.2.dev Model: Vehicle Detection FP16-INT8 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 0.9788 1.9576 2.9364 3.9152 4.894 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 4.32 4.35 4.32 4.33 MIN: 2.62 / MAX: 13.23 -fno-strict-overflow -fwrapv - MIN: 2.69 / MAX: 14.15 MIN: 2.61 / MAX: 13.39 MIN: 2.65 / MAX: 13.04 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.2.dev Model: Weld Porosity Detection FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 1.2578 2.5156 3.7734 5.0312 6.289 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 5.55 5.59 5.57 5.56 MIN: 2.87 / MAX: 13.9 -fno-strict-overflow -fwrapv - MIN: 2.91 / MAX: 14.45 MIN: 2.89 / MAX: 15.21 MIN: 2.88 / MAX: 14.86 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.2.dev Model: Machine Translation EN To DE FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 13 26 39 52 65 SE +/- 0.40, N = 3 SE +/- 0.64, N = 3 SE +/- 0.34, N = 3 SE +/- 0.27, N = 3 59.95 59.47 59.83 59.20 MIN: 28.33 / MAX: 87.44 -fno-strict-overflow -fwrapv - MIN: 27.89 / MAX: 86.56 MIN: 27.32 / MAX: 85.28 MIN: 30.14 / MAX: 87.47 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.2.dev Model: Weld Porosity Detection FP16-INT8 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 1.2488 2.4976 3.7464 4.9952 6.244 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 5.52 5.55 5.54 5.52 MIN: 2.93 / MAX: 12.86 -fno-strict-overflow -fwrapv - MIN: 2.9 / MAX: 18.91 MIN: 2.86 / MAX: 13.37 MIN: 2.86 / MAX: 14.5 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.2.dev Model: Person Vehicle Bike Detection FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 1.0913 2.1826 3.2739 4.3652 5.4565 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.05, N = 4 4.80 4.82 4.83 4.85 MIN: 3.12 / MAX: 13.53 -fno-strict-overflow -fwrapv - MIN: 3.66 / MAX: 14.3 MIN: 3.7 / MAX: 13.95 MIN: 3.25 / MAX: 15.02 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.2.dev Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 0.081 0.162 0.243 0.324 0.405 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 0.36 0.36 0.36 0.36 MIN: 0.21 / MAX: 7.72 -fno-strict-overflow -fwrapv - MIN: 0.21 / MAX: 9.03 MIN: 0.21 / MAX: 8.45 MIN: 0.21 / MAX: 8.85 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.2.dev Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 0.0563 0.1126 0.1689 0.2252 0.2815 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 0.25 0.25 0.25 0.25 MIN: 0.15 / MAX: 7.46 -fno-strict-overflow -fwrapv - MIN: 0.15 / MAX: 8.94 MIN: 0.15 / MAX: 8.4 MIN: 0.15 / MAX: 21.4 1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared
Dolfyn Dolfyn is a Computational Fluid Dynamics (CFD) code of modern numerical simulation techniques. The Dolfyn test profile measures the execution time of the bundled computational fluid dynamics demos that are bundled with Dolfyn. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Dolfyn 0.527 Computational Fluid Dynamics AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 3 6 9 12 15 SE +/- 0.04, N = 3 SE +/- 0.06, N = 3 SE +/- 0.05, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 10.77 10.79 10.51 10.77 10.82
Monte Carlo Simulations of Ionised Nebulae Mocassin is the Monte Carlo Simulations of Ionised Nebulae. MOCASSIN is a fully 3D or 2D photoionisation and dust radiative transfer code which employs a Monte Carlo approach to the transfer of radiation through media of arbitrary geometry and density distribution. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Monte Carlo Simulations of Ionised Nebulae 2019-03-24 Input: Dust 2D tau100.0 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 15 30 45 60 75 SE +/- 0.58, N = 3 SE +/- 0.88, N = 3 SE +/- 0.33, N = 3 SE +/- 0.71, N = 4 SE +/- 0.88, N = 3 64 64 63 65 64 1. (F9X) gfortran options: -cpp -Jsource/ -ffree-line-length-0 -lm -std=legacy -O3 -O2 -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lz
libavif avifenc This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 0.11 Encoder Speed: 0 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 15 30 45 60 75 SE +/- 0.18, N = 3 SE +/- 0.11, N = 3 SE +/- 0.17, N = 3 SE +/- 0.45, N = 3 SE +/- 0.23, N = 3 63.11 68.52 66.27 65.23 65.15 1. (CXX) g++ options: -O3 -fPIC -march=native -flto -lm
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 0.11 Encoder Speed: 2 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 8 16 24 32 40 SE +/- 0.08, N = 3 SE +/- 0.24, N = 3 SE +/- 0.16, N = 3 SE +/- 0.12, N = 3 SE +/- 0.04, N = 3 31.38 34.85 33.75 32.38 32.31 1. (CXX) g++ options: -O3 -fPIC -march=native -flto -lm
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 0.11 Encoder Speed: 6 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.7205 1.441 2.1615 2.882 3.6025 SE +/- 0.022, N = 3 SE +/- 0.011, N = 3 SE +/- 0.006, N = 3 SE +/- 0.018, N = 3 SE +/- 0.005, N = 3 2.943 3.202 3.105 3.079 3.044 1. (CXX) g++ options: -O3 -fPIC -march=native -flto -lm
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 0.11 Encoder Speed: 6, Lossless AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 1.1313 2.2626 3.3939 4.5252 5.6565 SE +/- 0.019, N = 3 SE +/- 0.027, N = 3 SE +/- 0.002, N = 3 SE +/- 0.002, N = 3 SE +/- 0.024, N = 3 4.229 5.028 4.675 4.377 4.363 1. (CXX) g++ options: -O3 -fPIC -march=native -flto -lm
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 0.11 Encoder Speed: 10, Lossless AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 0.6923 1.3846 2.0769 2.7692 3.4615 SE +/- 0.033, N = 3 SE +/- 0.021, N = 3 SE +/- 0.004, N = 3 SE +/- 0.009, N = 3 SE +/- 0.018, N = 3 2.773 3.077 2.947 2.761 2.725 1. (CXX) g++ options: -O3 -fPIC -march=native -flto -lm
C-Ray This is a test of C-Ray, a simple raytracer designed to test the floating-point CPU performance. This test is multi-threaded (16 threads per core), will shoot 8 rays per pixel for anti-aliasing, and will generate a 1600 x 1200 image. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better C-Ray 1.1 Total Time - 4K, 16 Rays Per Pixel AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 6 12 18 24 30 SE +/- 0.06, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 27.45 18.99 19.33 20.13 20.15 1. (CC) gcc options: -lm -lpthread -O3 -march=native -flto
POV-Ray This is a test of POV-Ray, the Persistence of Vision Raytracer. POV-Ray is used to create 3D graphics using ray-tracing. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better POV-Ray 3.7.0.7 Trace Time AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 4 8 12 16 20 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 16.10 15.96 16.57 14.83 -R/usr/lib 1. (CXX) g++ options: -pipe -O3 -ffast-math -march=native -flto -lXpm -lSM -lICE -lX11 -ltiff -ljpeg -lpng -lz -lrt -lm -lboost_thread -lboost_system Trace Time
GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./povray: 3: ./unix/povray: not found
Primesieve Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Primesieve 8.0 Length: 1e12 AOCC 4.0 GCC 12.2 LLVM Clang 14 LLVM Clang 15.0.2 2 4 6 8 10 SE +/- 0.011, N = 3 SE +/- 0.010, N = 3 SE +/- 0.008, N = 3 SE +/- 0.016, N = 3 6.245 6.187 6.172 6.243 1. (CXX) g++ options: -O3 -march=native -flto Length: 1e12
GCC 13.0 14 Nov: The test quit with a non-zero exit status.
AOBench AOBench is a lightweight ambient occlusion renderer, written in C. The test profile is using a size of 2048 x 2048. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better AOBench Size: 2048 x 2048 - Total Time AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 5 10 15 20 25 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.28, N = 3 SE +/- 0.02, N = 3 21.58 19.76 19.90 21.70 21.39 1. (CC) gcc options: -lm -O3 -march=native -flto
FLAC Audio Encoding This test times how long it takes to encode a sample WAV file to FLAC audio format ten times using the --best preset settings. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better FLAC Audio Encoding 1.4 WAV To FLAC AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 3 6 9 12 15 SE +/- 0.018, N = 5 SE +/- 0.073, N = 5 SE +/- 0.021, N = 5 SE +/- 0.025, N = 5 SE +/- 0.007, N = 5 9.928 10.520 10.810 9.280 9.435 1. (CXX) g++ options: -O3 -march=native -flto -fvisibility=hidden -logg -lm
LAME MP3 Encoding LAME is an MP3 encoder licensed under the LGPL. This test measures the time required to encode a WAV file to MP3 format. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better LAME MP3 Encoding 3.100 WAV To MP3 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 1.0647 2.1294 3.1941 4.2588 5.3235 SE +/- 0.014, N = 3 SE +/- 0.012, N = 3 SE +/- 0.007, N = 3 SE +/- 0.005, N = 3 SE +/- 0.010, N = 3 4.732 4.409 4.387 4.620 4.529 -ffast-math -funroll-loops -fschedule-insns2 -fbranch-count-reg -fforce-addr -ffast-math -funroll-loops -fschedule-insns2 -fbranch-count-reg -fforce-addr 1. (CC) gcc options: -O3 -pipe -march=native -flto -lncurses -lm
eSpeak-NG Speech Engine This test times how long it takes the eSpeak speech synthesizer to read Project Gutenberg's The Outline of Science and output to a WAV file. This test profile is now tracking the eSpeak-NG version of eSpeak. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better eSpeak-NG Speech Engine 20200907 Text-To-Speech Synthesis AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 5 10 15 20 25 SE +/- 0.11, N = 4 SE +/- 0.11, N = 4 SE +/- 0.14, N = 4 SE +/- 0.11, N = 4 14.54 17.83 17.77 20.54 1. (CC) gcc options: -O3 -march=native -flto -std=c99 Text-To-Speech Synthesis
LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./espeak: 3: ./espeak-ng: not found
Ngspice Ngspice is an open-source SPICE circuit simulator. Ngspice was originally based on the Berkeley SPICE electronic circuit simulator. Ngspice supports basic threading using OpenMP. This test profile is making use of the ISCAS 85 benchmark circuits. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Ngspice 34 Circuit: C2670 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 14 28 42 56 70 SE +/- 0.47, N = 3 SE +/- 0.14, N = 3 SE +/- 0.41, N = 3 SE +/- 0.14, N = 3 SE +/- 0.46, N = 3 58.26 63.75 62.68 57.71 58.96 -lstdc++ -lstdc++ -lstdc++ -lstdc++ 1. (CC) gcc options: -O3 -march=native -flto -fopenmp -lm -lfftw3 -lXaw -lXmu -lXt -lXext -lX11 -lSM -lICE
OpenBenchmarking.org Seconds, Fewer Is Better Ngspice 34 Circuit: C7552 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 13 26 39 52 65 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.09, N = 3 SE +/- 0.06, N = 3 SE +/- 0.17, N = 3 56.31 58.28 60.29 56.22 56.48 -lstdc++ -lstdc++ -lstdc++ -lstdc++ 1. (CC) gcc options: -O3 -march=native -flto -fopenmp -lm -lfftw3 -lXaw -lXmu -lXt -lXext -lX11 -lSM -lICE
Tachyon This is a test of the threaded Tachyon, a parallel ray-tracing system, measuring the time to ray-trace a sample scene. The sample scene used is the Teapot scene ray-traced to 8K x 8K with 32 samples. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Tachyon 0.99.2 Total Time AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 12 24 36 48 60 SE +/- 0.19, N = 3 SE +/- 0.22, N = 3 SE +/- 0.20, N = 3 SE +/- 0.19, N = 3 SE +/- 0.16, N = 3 55.49 55.48 55.28 54.97 55.08 1. (CC) gcc options: -m64 -O3 -fomit-frame-pointer -ffast-math -ltachyon -lm -lpthread
OpenBenchmarking.org Seconds, Fewer Is Better CppPerformanceBenchmarks 9 Test: Ctype AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 8 16 24 32 40 SE +/- 0.44, N = 3 SE +/- 0.22, N = 3 SE +/- 0.36, N = 3 SE +/- 0.01, N = 3 SE +/- 0.20, N = 3 31.36 32.28 33.77 31.63 31.78 1. (CXX) g++ options: -O3 -march=native -flto -std=c++11
OpenBenchmarking.org Seconds, Fewer Is Better CppPerformanceBenchmarks 9 Test: Math Library AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 40 80 120 160 200 SE +/- 0.09, N = 3 SE +/- 0.39, N = 3 SE +/- 0.41, N = 3 SE +/- 0.38, N = 3 SE +/- 0.74, N = 3 164.80 166.58 169.48 164.39 165.55 1. (CXX) g++ options: -O3 -march=native -flto -std=c++11
OpenBenchmarking.org Seconds, Fewer Is Better CppPerformanceBenchmarks 9 Test: Stepanov Vector AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 9 18 27 36 45 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.06, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 39.38 35.20 35.51 40.18 40.28 1. (CXX) g++ options: -O3 -march=native -flto -std=c++11
OpenBenchmarking.org Seconds, Fewer Is Better CppPerformanceBenchmarks 9 Test: Function Objects AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 3 6 9 12 15 SE +/- 0.031, N = 3 SE +/- 0.086, N = 3 SE +/- 0.094, N = 3 SE +/- 0.014, N = 3 SE +/- 0.006, N = 3 9.023 9.251 9.211 9.246 9.333 1. (CXX) g++ options: -O3 -march=native -flto -std=c++11
OpenBenchmarking.org Seconds, Fewer Is Better CppPerformanceBenchmarks 9 Test: Stepanov Abstraction AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 4 8 12 16 20 SE +/- 0.15, N = 3 SE +/- 0.14, N = 3 SE +/- 0.11, N = 3 SE +/- 0.10, N = 3 SE +/- 0.03, N = 3 16.93 15.73 15.73 17.03 17.18 1. (CXX) g++ options: -O3 -march=native -flto -std=c++11
SQLite Speedtest This is a benchmark of SQLite's speedtest1 benchmark program with an increased problem size of 1,000. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better SQLite Speedtest 3.30 Timed Time - Size 1,000 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 8 16 24 32 40 SE +/- 0.17, N = 3 SE +/- 0.15, N = 3 SE +/- 0.22, N = 3 SE +/- 0.08, N = 3 SE +/- 0.30, N = 3 34.77 34.23 34.73 36.07 36.02 1. (CC) gcc options: -O3 -march=native -flto -lz
KTX-Software toktx This is a benchmark of The Khronos Group's KTX-Software library and tools. KTX-Software provides "toktx" for converting/creating in the KTX container format for image textures. This benchmark times how long it takes to convert to KTX 2.0 format with various settings using a reference PNG sample input. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better KTX-Software toktx 4.0 Settings: UASTC 3 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 1.1212 2.2424 3.3636 4.4848 5.606 SE +/- 0.004, N = 3 SE +/- 0.005, N = 3 SE +/- 0.006, N = 3 SE +/- 0.025, N = 3 SE +/- 0.005, N = 3 4.957 4.958 4.961 4.982 4.983
OpenBenchmarking.org Seconds, Fewer Is Better KTX-Software toktx 4.0 Settings: Zstd Compression 19 AOCC 4.0 GCC 12.2 GCC 13.0 14 Nov LLVM Clang 14 LLVM Clang 15.0.2 3 6 9 12 15 SE +/- 0.02, N = 3 SE +/- 0.07, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 11.35 11.69 11.62 11.64 11.46
AOCC 4.0 Processor: AMD Ryzen 9 7950X 16-Core @ 5.88GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR X670E HERO (0703 BIOS), Chipset: AMD Device 14d8, Memory: 32GB, Disk: 1000GB Sabrent Rocket 4.0 Plus, Graphics: AMD Radeon RX 6800 16GB (2475/1000MHz), Audio: AMD Navi 21/23, Monitor: ASUS MG28U, Network: Intel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411
OS: Ubuntu 22.10, Kernel: 6.1.0-060100rc3daily20221103-generic (x86_64), Desktop: GNOME Shell 43.0, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.2.1 (LLVM 15.0.2 DRM 3.49), Vulkan: 1.3.224, Compiler: Clang 14.0.6, File-System: ext4, Screen Resolution: 3840x2160
Kernel Notes: Transparent Huge Pages: madviseEnvironment Notes: CXXFLAGS="-O3 -march=native -flto" CFLAGS="-O3 -march=native -flto"Compiler Notes: Optimized build with assertions; Default target: x86_64-unknown-linux-gnu; Host CPU: znver4Processor Notes: Scaling Governor: amd-pstate schedutil (Boost: Enabled) - CPU Microcode: 0xa601203Graphics Notes: BAR1 / Visible vRAM Size: 16368 MB - vBIOS Version: 113-D4120900-101Python Notes: Python 3.10.7Security 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: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 12 November 2022 04:55 by user pts.
GCC 12.2 Processor: AMD Ryzen 9 7950X 16-Core @ 5.88GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR X670E HERO (0703 BIOS), Chipset: AMD Device 14d8, Memory: 32GB, Disk: 1000GB Sabrent Rocket 4.0 Plus, Graphics: AMD Radeon RX 6800 16GB (2475/1000MHz), Audio: AMD Navi 21/23, Monitor: ASUS MG28U, Network: Intel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411
OS: Ubuntu 22.10, Kernel: 6.1.0-060100rc3daily20221103-generic (x86_64), Desktop: GNOME Shell 43.0, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.2.1 (LLVM 15.0.2 DRM 3.49), Vulkan: 1.3.224, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 3840x2160
Kernel Notes: Transparent Huge Pages: madviseEnvironment Notes: CXXFLAGS="-O3 -march=native -flto" CFLAGS="-O3 -march=native -flto"Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --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-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-gcn/usr --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 -vProcessor Notes: Scaling Governor: amd-pstate schedutil (Boost: Enabled) - CPU Microcode: 0xa601203Graphics Notes: BAR1 / Visible vRAM Size: 16368 MB - vBIOS Version: 113-D4120900-101Python Notes: Python 3.10.7Security 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: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 13 November 2022 16:24 by user pts.
GCC 13.0 14 Nov Processor: AMD Ryzen 9 7950X 16-Core @ 5.88GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR X670E HERO (0703 BIOS), Chipset: AMD Device 14d8, Memory: 32GB, Disk: 1000GB Sabrent Rocket 4.0 Plus, Graphics: AMD Radeon RX 6800 16GB (2475/1000MHz), Audio: AMD Navi 21/23, Monitor: ASUS MG28U, Network: Intel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411
OS: Ubuntu 22.10, Kernel: 6.1.0-060100rc3daily20221103-generic (x86_64), Desktop: GNOME Shell 43.0, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.2.1 (LLVM 15.0.2 DRM 3.49), Vulkan: 1.3.224, Compiler: GCC 13.0.0 20221114 + clang (GCC) 13.0.0 20221114 (experimental), File-System: ext4, Screen Resolution: 3840x2160
Kernel Notes: Transparent Huge Pages: madviseEnvironment Notes: CXXFLAGS="-O3 -march=native -flto" CFLAGS="-O3 -march=native -flto"Compiler Notes: --disable-multilibProcessor Notes: Scaling Governor: amd-pstate schedutil (Boost: Enabled) - CPU Microcode: 0xa601203Graphics Notes: BAR1 / Visible vRAM Size: 16368 MB - vBIOS Version: 113-D4120900-101Python Notes: Python 3.10.7Security 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: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 14 November 2022 09:05 by user pts.
LLVM Clang 14 Processor: AMD Ryzen 9 7950X 16-Core @ 5.88GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR X670E HERO (0703 BIOS), Chipset: AMD Device 14d8, Memory: 32GB, Disk: 1000GB Sabrent Rocket 4.0 Plus, Graphics: AMD Radeon RX 6800 16GB (2475/1000MHz), Audio: AMD Navi 21/23, Monitor: ASUS MG28U, Network: Intel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411
OS: Ubuntu 22.10, Kernel: 6.1.0-060100rc3daily20221103-generic (x86_64), Desktop: GNOME Shell 43.0, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.2.1 (LLVM 15.0.2 DRM 3.49), Vulkan: 1.3.224, Compiler: Clang 14.0.6-2, File-System: ext4, Screen Resolution: 3840x2160
Kernel Notes: Transparent Huge Pages: madviseEnvironment Notes: CXXFLAGS="-O3 -march=native -flto" CFLAGS="-O3 -march=native -flto"Processor Notes: Scaling Governor: amd-pstate schedutil (Boost: Enabled) - CPU Microcode: 0xa601203Graphics Notes: BAR1 / Visible vRAM Size: 16368 MB - vBIOS Version: 113-D4120900-101Python Notes: Python 3.10.7Security 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: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 13 November 2022 04:35 by user pts.
LLVM Clang 15.0.2 Processor: AMD Ryzen 9 7950X 16-Core @ 5.88GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR X670E HERO (0703 BIOS), Chipset: AMD Device 14d8, Memory: 32GB, Disk: 1000GB Sabrent Rocket 4.0 Plus, Graphics: AMD Radeon RX 6800 16GB (2475/1000MHz), Audio: AMD Navi 21/23, Monitor: ASUS MG28U, Network: Intel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411
OS: Ubuntu 22.10, Kernel: 6.1.0-060100rc3daily20221103-generic (x86_64), Desktop: GNOME Shell 43.0, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.2.1 (LLVM 15.0.2 DRM 3.49), Vulkan: 1.3.224, Compiler: Clang 15.0.2-1, File-System: ext4, Screen Resolution: 3840x2160
Kernel Notes: Transparent Huge Pages: madviseEnvironment Notes: CXXFLAGS="-O3 -march=native -flto" CFLAGS="-O3 -march=native -flto"Processor Notes: Scaling Governor: amd-pstate schedutil (Boost: Enabled) - CPU Microcode: 0xa601203Graphics Notes: BAR1 / Visible vRAM Size: 16368 MB - vBIOS Version: 113-D4120900-101Python Notes: Python 3.10.7Security 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: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 14 November 2022 16:49 by user pts.