Intel Xeon Platinum 8380 Ice Lake P-State CPU frequency scaling Linux benchmarks by Michael Larabel for a future article.
P-State powersave Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --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: intel_pstate powersave - CPU Microcode: 0xd000270Python Notes: Python 3.9.4Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected
P-State performance Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --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: intel_pstate performance - CPU Microcode: 0xd000270Python Notes: Python 3.9.4Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected
intel_cpufreq schedutil Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --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: intel_cpufreq schedutil - CPU Microcode: 0xd000270Python Notes: Python 3.9.4Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected
intel_cpufreq performance Processor: 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads), Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS), Chipset: Intel Device 0998, Memory: 16 x 32 GB DDR4-3200MT/s Hynix HMA84GR7CJR4N-XN, Disk: 800GB INTEL SSDPF21Q800GB, Graphics: ASPEED, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP
OS: Ubuntu 21.04, Kernel: 5.11.0-17-generic (x86_64), Desktop: GNOME Shell 3.38.4, Display Server: X Server, Vulkan: 1.0.2, Compiler: GCC 10.3.0, File-System: ext4, Screen Resolution: 1024x768
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --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: intel_cpufreq performance - CPU Microcode: 0xd000270Python Notes: Python 3.9.4Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected
Xeon Scalable Ice Lake P-State Governor OpenBenchmarking.org Phoronix Test Suite 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads) Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS) Intel Device 0998 16 x 32 GB DDR4-3200MT/s Hynix HMA84GR7CJR4N-XN 800GB INTEL SSDPF21Q800GB ASPEED 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP Ubuntu 21.04 5.11.0-17-generic (x86_64) GNOME Shell 3.38.4 X Server 1.0.2 GCC 10.3.0 ext4 1024x768 Processor Motherboard Chipset Memory Disk Graphics Network OS Kernel Desktop Display Server Vulkan Compiler File-System Screen Resolution Xeon Scalable Ice Lake P-State Governor Benchmarks System Logs - Transparent Huge Pages: madvise - --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v - P-State powersave: Scaling Governor: intel_pstate powersave - CPU Microcode: 0xd000270 - P-State performance: Scaling Governor: intel_pstate performance - CPU Microcode: 0xd000270 - intel_cpufreq schedutil: Scaling Governor: intel_cpufreq schedutil - CPU Microcode: 0xd000270 - intel_cpufreq performance: Scaling Governor: intel_cpufreq performance - CPU Microcode: 0xd000270 - Python 3.9.4 - itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected
P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance Result Overview Phoronix Test Suite 100% 148% 195% 243% 290% AOM AV1 Kvazaar PostgreSQL pgbench SVT-VP9 SVT-HEVC SVT-AV1 PJSIP Timed Godot Game Engine Compilation Darmstadt Automotive Parallel Heterogeneous Suite Zstd Compression ArrayFire OpenVKL libavif avifenc Timed Linux Kernel Compilation Timed Node.js Compilation Rodinia TensorFlow Lite NAS Parallel Benchmarks Quantum ESPRESSO Embree OSPray Blender Timed LLVM Compilation Intel Open Image Denoise Timed Mesa Compilation OpenFOAM ONNX Runtime High Performance Conjugate Gradient Stockfish GROMACS LAMMPS Molecular Dynamics Simulator Liquid-DSP Xcompact3d Incompact3d RELION NAMD WRF NWChem
P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance Per Watt Result Overview Phoronix Test Suite 100% 107% 113% 120% 126% AOM AV1 NAS Parallel Benchmarks SVT-HEVC Intel Open Image Denoise OSPray Zstd Compression Kvazaar OpenVKL SVT-AV1 Stockfish Embree ArrayFire Darmstadt Automotive Parallel Heterogeneous Suite PJSIP SVT-VP9 ONNX Runtime Liquid-DSP High Performance Conjugate Gradient GROMACS LAMMPS Molecular Dynamics Simulator P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M P.W.G.M
Xeon Scalable Ice Lake P-State Governor ospray: Magnetic Reconnection - SciVis ospray: XFrog Forest - SciVis ospray: XFrog Forest - Path Tracer ospray: NASA Streamlines - SciVis ospray: San Miguel - SciVis ospray: NASA Streamlines - Path Tracer ospray: San Miguel - Path Tracer svt-av1: Preset 8 - Bosphorus 4K svt-hevc: 7 - Bosphorus 1080p svt-hevc: 10 - Bosphorus 1080p svt-vp9: PSNR/SSIM Optimized - Bosphorus 1080p svt-vp9: VMAF Optimized - Bosphorus 1080p aom-av1: Speed 9 Realtime - Bosphorus 4K aom-av1: Speed 8 Realtime - Bosphorus 4K aom-av1: Speed 6 Realtime - Bosphorus 4K aom-av1: Speed 6 Two-Pass - Bosphorus 4K embree: Pathtracer ISPC - Asian Dragon embree: Pathtracer ISPC - Asian Dragon Obj embree: Pathtracer ISPC - Crown kvazaar: Bosphorus 4K - Very Fast kvazaar: Bosphorus 4K - Ultra Fast hpcg: arrayfire: BLAS CPU oidn: RT.hdr_alb_nrm.3840x2160 oidn: RT.ldr_alb_nrm.3840x2160 oidn: RTLightmap.hdr.4096x4096 onnx: yolov4 - OpenMP CPU onnx: fcn-resnet101-11 - OpenMP CPU onnx: shufflenet-v2-10 - OpenMP CPU onnx: super-resolution-10 - OpenMP CPU onnx: bertsquad-10 - OpenMP CPU openvkl: vklBenchmark openvkl: vklBenchmarkStructuredVolume openvkl: vklBenchmarkUnstructuredVolume openvkl: vklBenchmarkVdbVolume compress-zstd: 8, Long Mode - Compression Speed compress-zstd: 19 - Compression Speed compress-zstd: 19 - Decompression Speed compress-zstd: 19, Long Mode - Compression Speed compress-zstd: 19, Long Mode - Decompression Speed stockfish: Total Time gromacs: MPI CPU - water_GMX50_bare lammps: 20k Atoms pjsip: OPTIONS, Stateful pjsip: OPTIONS, Stateless pjsip: INVITE liquid-dsp: 64 - 256 - 57 liquid-dsp: 128 - 256 - 57 liquid-dsp: 160 - 256 - 57 daphne: OpenMP - Euclidean Cluster daphne: OpenMP - NDT Mapping daphne: OpenMP - Points2Image npb: BT.C npb: EP.C npb: EP.D npb: FT.C npb: LU.C npb: SP.B npb: SP.C npb: IS.D npb: MG.C npb: CG.C pgbench: 1000 - 100 - Read Write pgbench: 1000 - 100 - Read Only pgbench: 1000 - 250 - Read Write pgbench: 1000 - 250 - Read Only namd: ATPase Simulation - 327,506 Atoms tensorflow-lite: Mobilenet Quant tensorflow-lite: Inception ResNet V2 tensorflow-lite: Inception V4 arrayfire: Conjugate Gradient CPU pgbench: 1000 - 100 - Read Write - Average Latency pgbench: 1000 - 100 - Read Only - Average Latency pgbench: 1000 - 250 - Read Write - Average Latency pgbench: 1000 - 250 - Read Only - Average Latency openfoam: Motorbike 30M openfoam: Motorbike 60M wrf: conus 2.5km incompact3d: input.i3d 193 Cells Per Direction incompact3d: X3D-benchmarking input.i3d qe: AUSURF112 relion: Basic - CPU rodinia: OpenMP CFD Solver rodinia: OpenMP LavaMD rodinia: OpenMP Leukocyte rodinia: OpenMP Streamcluster rodinia: OpenMP HotSpot3D avifenc: 6, Lossless avifenc: 10 avifenc: 10, Lossless build-godot: Time To Compile build-linux-kernel: Time To Compile build-mesa: Time To Compile build-llvm: Ninja build-llvm: Unix Makefiles build-nodejs: Time To Compile blender: BMW27 - CPU-Only blender: Classroom - CPU-Only blender: Barbershop - CPU-Only nwchem: C240 Buckyball P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 111.11 18.87 10.38 125 90.91 27.78 10.42 30.546 168.38 258.65 206.54 210.92 23.59 17.07 7.38 3.94 106.6219 87.2092 66.5251 13.26 23.03 40.1359 4538.69 2.85 2.87 1.38 480 196 8212 6898 509 670 71974438 1757269 21722326 296.0 82.7 2563.2 46.7 2711.6 177031122 8.999 35.495 3769 41421 2548 3029900000 3270133333 3073933333 831.33 742.30 6634.58 191166.41 6051.90 8620.97 90989.73 174911.02 114946.46 91313.60 2999.72 116754.16 39485.72 23383 393329 23411 906434 0.27093 41126.8 572065 682153 4.292 4.283 0.256 10.723 0.277 14.92 105.03 9892.42 11.1916566 290.411102 1106.69 350.266 4.827 39.815 69.049 7.710 104.339 37.469 5.205 8.833 75.612 24.881 18.983 129.272 199.398 102.197 29.72 72.17 107.73 1882.3 114.81 18.87 10.31 142.86 100 27.78 10.45 56.887 305.77 576.99 453.67 460.44 61.06 48.62 19.09 9.69 107.6307 89.0895 74.0551 38.21 46.31 40.6471 4539.99 2.93 2.93 1.40 483 194 8392 6626 477 811 106056458 1803840 29996539 1011.9 86.7 2561.8 49.1 2677.9 177131428 9.052 35.825 10362 42389 5393 3066133333 3293000000 3085100000 1020.77 924.84 13527.238996430 197815.31 7610.18 8920.10 100543.72 187300.47 123544.09 91817.76 3014.77 120084.27 40261.26 63049 1236713 66493 924190 0.27168 33656.3 572149 663707 6.131 1.588 0.081 3.774 0.271 14.79 104.76 9885.556 11.0554509 290.143331 1178.52 349.993 4.746 39.423 46.358 7.816 104.185 30.345 4.341 7.248 48.157 21.486 18.497 127.321 190.685 90.850 28.40 71.27 105.10 1888.2 111.11 18.87 10.31 125 90.91 27.78 10.38 35.480 193.80 345.48 260.25 262.78 21.72 15.90 5.86 4.00 102.7043 87.9927 67.7540 11.30 20.53 40.0659 4510.29 2.78 2.88 1.38 483 194 8334 6756 504 718 77997628 1748017 23599249 311.7 81.5 2594.7 46.1 2693.4 175446182 9.042 35.792 3797 41743 2540 3003125000 3291466667 3082533333 884.75 743.01 7077.76 191705.02 6157.58 8634.30 91970.99 177878.68 115501.41 91642.33 2926.26 117869.49 39430.66 36577 1228570 65726 885379 0.27227 35239.0 571632 672254 6.454 2.740 0.081 3.814 0.285 15.07 105.10 9889.907 11.2054179 290.416697 1156.49 349.436 4.710 39.199 63.286 7.712 104.184 31.987 5.134 8.486 53.001 22.190 18.754 127.688 191.173 91.857 29.25 72.24 107.39 1880 112.96 18.87 10.31 142.86 100 27.78 10.42 56.750 309.13 574.88 457.02 464.71 61.27 48.91 19.20 10.01 106.1736 89.4905 74.2420 38.26 46.34 40.8311 4574.83 2.93 2.93 1.40 476 197 8380 6955 503 821 106704641 1809667 30016318 1018.6 83.1 2609.7 48.3 2713.1 178629474 9.102 35.849 10409 42338 5416 3072600000 3293400000 3096700000 1021.62 914.70 13249.678695875 197821.88 7705.44 8938.09 100763.11 186973.06 123176.72 91830.56 2978.52 120124.05 40226.02 62826 1231983 66186 876462 0.27051 33581.3 572679 677421 4.005 1.594 0.081 3.791 0.288 14.49 104.27 9875.264 11.0561748 289.975586 1175.69 349.028 4.729 39.746 47.105 7.727 104.088 30.014 4.340 7.276 48.651 21.410 18.517 127.332 191.801 91.081 28.28 71.22 104.51 1882.2 OpenBenchmarking.org
CPU Temperature Monitor OpenBenchmarking.org Celsius CPU Temperature Monitor Phoronix Test Suite System Monitoring P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 15 30 45 60 75 Min: 28 / Avg: 58.72 / Max: 79 Min: 36 / Avg: 62.03 / Max: 80 Min: 28 / Avg: 59.65 / Max: 78 Min: 36 / Avg: 63.47 / Max: 78
CPU Peak Freq (Highest CPU Core Frequency) Monitor OpenBenchmarking.org Megahertz CPU Peak Freq (Highest CPU Core Frequency) Monitor Phoronix Test Suite System Monitoring P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 600 1200 1800 2400 3000 Min: 801 / Avg: 2883.69 / Max: 3537 Min: 2195 / Avg: 3186.64 / Max: 3518 Min: 801 / Avg: 3003.21 / Max: 3442 Min: 2162 / Avg: 3195.19 / Max: 3483
CPU Power Consumption Monitor OpenBenchmarking.org Watts CPU Power Consumption Monitor Phoronix Test Suite System Monitoring P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 130 260 390 520 650 Min: 53.58 / Avg: 371.79 / Max: 722.02 Min: 70.83 / Avg: 413.76 / Max: 639.8 Min: 53.9 / Avg: 368.68 / Max: 634.98 Min: 137.84 / Avg: 417.23 / Max: 641.13
OSPray OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: Magnetic Reconnection - Renderer: SciVis P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.1285 0.257 0.3855 0.514 0.6425 0.571 0.365 0.570 0.370
OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: XFrog Forest - Renderer: SciVis P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0115 0.023 0.0345 0.046 0.0575 0.051 0.047 0.048 0.046
OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: XFrog Forest - Renderer: Path Tracer P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0054 0.0108 0.0162 0.0216 0.027 0.024 0.023 0.023 0.023
Result
OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: Magnetic Reconnection - Renderer: SciVis P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 30 60 90 120 150 SE +/- 0.00, N = 8 SE +/- 1.64, N = 15 SE +/- 0.00, N = 8 SE +/- 1.26, N = 15 111.11 114.81 111.11 112.96 MIN: 22.22 MIN: 24.39 / MAX: 125 MIN: 22.73 MIN: 26.32 / MAX: 125 FPS Per Watt
OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: Magnetic Reconnection - Renderer: SciVis P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.1285 0.257 0.3855 0.514 0.6425 0.571 0.365 0.570 0.370 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 1986 2628 P-State performance 2968 3363 3404 intel_cpufreq schedutil 1019 2269 3267 intel_cpufreq performance 2209 3321 3403 OpenBenchmarking.org Megahertz, More Is Better OSPray 1.8.5 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 69.6 194.7 334.6 P-State performance 174.6 314.4 422.6 intel_cpufreq schedutil 70.1 194.9 328.7 intel_cpufreq performance 154.4 305.6 423.6 OpenBenchmarking.org Watts, Fewer Is Better OSPray 1.8.5 CPU Power Consumption Monitor 110 220 330 440 550 1. P-State powersave: Approximate power consumption of 365 Joules per run. 2. P-State performance: Approximate power consumption of 503 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 390 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 489 Joules per run. CPU Temp
Min Avg Max P-State powersave 41.0 45.6 54.0 P-State performance 46.0 54.0 64.0 intel_cpufreq schedutil 43.0 47.7 57.0 intel_cpufreq performance 45.0 54.1 65.0 OpenBenchmarking.org Celsius, Fewer Is Better OSPray 1.8.5 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: Magnetic Reconnection - Renderer: SciVis P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 20 40 60 80 100 Min: 111.11 / Avg: 111.11 / Max: 111.11 Min: 111.11 / Avg: 114.81 / Max: 125 Min: 111.11 / Avg: 111.11 / Max: 111.11 Min: 111.11 / Avg: 112.96 / Max: 125
OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: NASA Streamlines - Renderer: SciVis P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.1375 0.275 0.4125 0.55 0.6875 0.611 0.520 0.528 0.513
Result
OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: XFrog Forest - Renderer: SciVis P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 5 10 15 20 25 SE +/- 0.00, N = 4 SE +/- 0.00, N = 4 SE +/- 0.00, N = 4 SE +/- 0.00, N = 4 18.87 18.87 18.87 18.87 MIN: 13.7 / MAX: 19.23 MIN: 13.89 / MAX: 19.61 MIN: 12.2 / MAX: 19.23 MIN: 12.99 / MAX: 19.61 FPS Per Watt
OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: XFrog Forest - Renderer: SciVis P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0115 0.023 0.0345 0.046 0.0575 0.051 0.047 0.048 0.046 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2426 3400 P-State performance 2525 3000 3407 intel_cpufreq schedutil 2202 2949 3404 intel_cpufreq performance 2517 3070 3408 OpenBenchmarking.org Megahertz, More Is Better OSPray 1.8.5 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 67.5 367.6 594.2 P-State performance 151.3 403.8 607.2 intel_cpufreq schedutil 68.1 393.1 604.2 intel_cpufreq performance 206.0 414.6 603.2 OpenBenchmarking.org Watts, Fewer Is Better OSPray 1.8.5 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 1654 Joules per run. 2. P-State performance: Approximate power consumption of 1716 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1769 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1762 Joules per run. CPU Temp
Min Avg Max P-State powersave 33.0 53.4 66.0 P-State performance 48.0 60.0 70.0 intel_cpufreq schedutil 37.0 57.8 70.0 intel_cpufreq performance 48.0 61.7 72.0 OpenBenchmarking.org Celsius, Fewer Is Better OSPray 1.8.5 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: XFrog Forest - Renderer: SciVis P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 5 10 15 20 25 Min: 18.87 / Avg: 18.87 / Max: 18.87 Min: 18.87 / Avg: 18.87 / Max: 18.87 Min: 18.87 / Avg: 18.87 / Max: 18.87 Min: 18.87 / Avg: 18.87 / Max: 18.87
OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: NASA Streamlines - Renderer: Path Tracer P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0173 0.0346 0.0519 0.0692 0.0865 0.075 0.072 0.077 0.071
Result
OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: XFrog Forest - Renderer: Path Tracer P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 SE +/- 0.04, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 10.38 10.31 10.31 10.31 MIN: 8.06 / MAX: 10.53 MIN: 8.33 / MAX: 10.75 MIN: 7.75 / MAX: 10.53 MIN: 8.26 / MAX: 10.75 FPS Per Watt
OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: XFrog Forest - Renderer: Path Tracer P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0054 0.0108 0.0162 0.0216 0.027 0.024 0.023 0.023 0.023 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2604 3400 P-State performance 2511 2855 3403 intel_cpufreq schedutil 2175 2803 3403 intel_cpufreq performance 2510 2815 3404 OpenBenchmarking.org Megahertz, More Is Better OSPray 1.8.5 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 142.0 430.1 597.4 P-State performance 160.8 444.9 606.8 intel_cpufreq schedutil 164.3 440.0 605.9 intel_cpufreq performance 189.9 449.5 597.2 OpenBenchmarking.org Watts, Fewer Is Better OSPray 1.8.5 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 3011 Joules per run. 2. P-State performance: Approximate power consumption of 2966 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2933 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2847 Joules per run. CPU Temp
Min Avg Max P-State powersave 48.0 64.6 73.0 P-State performance 53.0 67.4 74.0 intel_cpufreq schedutil 53.0 67.7 75.0 intel_cpufreq performance 54.0 67.8 75.0 OpenBenchmarking.org Celsius, Fewer Is Better OSPray 1.8.5 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: XFrog Forest - Renderer: Path Tracer P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 Min: 10.31 / Avg: 10.38 / Max: 10.42 Min: 10.31 / Avg: 10.31 / Max: 10.31 Min: 10.31 / Avg: 10.31 / Max: 10.31 Min: 10.31 / Avg: 10.31 / Max: 10.31
Result
OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: NASA Streamlines - Renderer: Path Tracer P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 7 14 21 28 35 SE +/- 0.00, N = 5 SE +/- 0.00, N = 5 SE +/- 0.00, N = 5 SE +/- 0.00, N = 5 27.78 27.78 27.78 27.78 MIN: 15.87 / MAX: 29.41 MIN: 17.54 / MAX: 29.41 MIN: 16.95 / MAX: 28.57 MIN: 17.54 / MAX: 29.41 FPS Per Watt
OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: NASA Streamlines - Renderer: Path Tracer P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0173 0.0346 0.0519 0.0692 0.0865 0.075 0.072 0.077 0.071 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2370 3400 P-State performance 2473 2991 3404 intel_cpufreq schedutil 2514 2971 3403 intel_cpufreq performance 2508 3032 3405 OpenBenchmarking.org Megahertz, More Is Better OSPray 1.8.5 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 67.6 368.0 608.4 P-State performance 189.3 387.6 594.0 intel_cpufreq schedutil 81.8 361.3 543.3 intel_cpufreq performance 152.7 394.0 593.4 OpenBenchmarking.org Watts, Fewer Is Better OSPray 1.8.5 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 1251 Joules per run. 2. P-State performance: Approximate power consumption of 1240 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1156 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1261 Joules per run. CPU Temp
Min Avg Max P-State powersave 45.0 58.9 69.0 P-State performance 52.0 63.3 71.0 intel_cpufreq schedutil 47.0 60.8 70.0 intel_cpufreq performance 54.0 63.8 72.0 OpenBenchmarking.org Celsius, Fewer Is Better OSPray 1.8.5 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: NASA Streamlines - Renderer: Path Tracer P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 6 12 18 24 30 Min: 27.78 / Avg: 27.78 / Max: 27.78 Min: 27.78 / Avg: 27.78 / Max: 27.78 Min: 27.78 / Avg: 27.78 / Max: 27.78 Min: 27.78 / Avg: 27.78 / Max: 27.78
OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: San Miguel - Renderer: Path Tracer P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0072 0.0144 0.0216 0.0288 0.036 0.032 0.031 0.031 0.030
Result
OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: San Miguel - Renderer: Path Tracer P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 SE +/- 0.00, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.00, N = 3 10.42 10.45 10.38 10.42 MIN: 7.69 / MAX: 10.64 MIN: 7.94 / MAX: 10.75 MIN: 7.69 / MAX: 10.53 MIN: 7.14 / MAX: 10.64 FPS Per Watt
OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: San Miguel - Renderer: Path Tracer P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0072 0.0144 0.0216 0.0288 0.036 0.032 0.031 0.031 0.030 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 802 2893 3405 P-State performance 2525 3111 3407 intel_cpufreq schedutil 2302 2987 3406 intel_cpufreq performance 2522 3101 3404 OpenBenchmarking.org Megahertz, More Is Better OSPray 1.8.5 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 59.2 323.6 599.8 P-State performance 149.2 340.5 585.7 intel_cpufreq schedutil 73.2 331.0 607.4 intel_cpufreq performance 155.3 347.7 589.2 OpenBenchmarking.org Watts, Fewer Is Better OSPray 1.8.5 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 3775 Joules per run. 2. P-State performance: Approximate power consumption of 3859 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 3752 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 4057 Joules per run. CPU Temp
Min Avg Max P-State powersave 43.0 55.8 69.0 P-State performance 48.0 57.8 70.0 intel_cpufreq schedutil 45.0 58.9 72.0 intel_cpufreq performance 50.0 59.5 72.0 OpenBenchmarking.org Celsius, Fewer Is Better OSPray 1.8.5 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: San Miguel - Renderer: Path Tracer P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 Min: 10.42 / Avg: 10.42 / Max: 10.42 Min: 10.42 / Avg: 10.45 / Max: 10.53 Min: 10.31 / Avg: 10.38 / Max: 10.42 Min: 10.42 / Avg: 10.42 / Max: 10.42
SVT-AV1 This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 0.8.7 Encoder Mode: Preset 8 - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 13 26 39 52 65 SE +/- 0.20, N = 3 SE +/- 0.24, N = 4 SE +/- 0.47, N = 3 SE +/- 0.31, N = 4 30.55 56.89 35.48 56.75 1. (CXX) g++ options: -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq -pie Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better SVT-AV1 0.8.7 Encoder Mode: Preset 8 - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0439 0.0878 0.1317 0.1756 0.2195 0.195 0.186 0.183 0.185 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2169 2307 P-State performance 3400 3403 3408 intel_cpufreq schedutil 2300 2742 3405 intel_cpufreq performance 3400 3403 3406 OpenBenchmarking.org Megahertz, More Is Better SVT-AV1 0.8.7 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 56.4 156.5 236.6 P-State performance 185.5 305.7 490.5 intel_cpufreq schedutil 74.6 193.4 332.0 intel_cpufreq performance 162.8 306.0 496.1 OpenBenchmarking.org Watts, Fewer Is Better SVT-AV1 0.8.7 CPU Power Consumption Monitor 130 260 390 520 650 1. P-State powersave: Approximate power consumption of 1200 Joules per run. 2. P-State performance: Approximate power consumption of 1375 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1289 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1377 Joules per run. CPU Temp
Min Avg Max P-State powersave 39.0 42.4 47.0 P-State performance 48.0 53.2 63.0 intel_cpufreq schedutil 44.0 48.3 53.0 intel_cpufreq performance 47.0 53.8 64.0 OpenBenchmarking.org Celsius, Fewer Is Better SVT-AV1 0.8.7 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 0.8.7 Encoder Mode: Preset 8 - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 11 22 33 44 55 Min: 30.15 / Avg: 30.55 / Max: 30.76 Min: 56.23 / Avg: 56.89 / Max: 57.39 Min: 34.67 / Avg: 35.48 / Max: 36.29 Min: 56.15 / Avg: 56.75 / Max: 57.62 1. (CXX) g++ options: -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq -pie
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 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better SVT-HEVC 1.5.0 Tuning: 7 - Input: Bosphorus 1080p P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 70 140 210 280 350 SE +/- 1.49, N = 8 SE +/- 0.58, N = 10 SE +/- 1.34, N = 15 SE +/- 1.23, N = 10 168.38 305.77 193.80 309.13 1. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better SVT-HEVC 1.5.0 Tuning: 7 - Input: Bosphorus 1080p P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.3287 0.6574 0.9861 1.3148 1.6435 1.461 1.198 1.449 1.216 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 1599 2308 P-State performance 3403 3403 3403 intel_cpufreq schedutil 1429 2450 3400 intel_cpufreq performance 3400 3403 3405 OpenBenchmarking.org Megahertz, More Is Better SVT-HEVC 1.5.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 54.7 115.2 180.8 P-State performance 193.9 255.3 332.3 intel_cpufreq schedutil 68.6 133.7 182.9 intel_cpufreq performance 175.6 254.1 330.5 OpenBenchmarking.org Watts, Fewer Is Better SVT-HEVC 1.5.0 CPU Power Consumption Monitor 80 160 240 320 400 1. P-State powersave: Approximate power consumption of 245 Joules per run. 2. P-State performance: Approximate power consumption of 383 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 267 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 407 Joules per run. CPU Temp
Min Avg Max P-State powersave 33.0 36.8 40.0 P-State performance 49.0 51.3 58.0 intel_cpufreq schedutil 39.0 43.2 48.0 intel_cpufreq performance 50.0 53.1 60.0 OpenBenchmarking.org Celsius, Fewer Is Better SVT-HEVC 1.5.0 CPU Temperature Monitor 16 32 48 64 80 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better SVT-HEVC 1.5.0 Tuning: 7 - Input: Bosphorus 1080p P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 60 120 180 240 300 Min: 160.6 / Avg: 168.38 / Max: 173.56 Min: 301.05 / Avg: 305.77 / Max: 307.38 Min: 186.68 / Avg: 193.8 / Max: 203.18 Min: 301.96 / Avg: 309.13 / Max: 314.47 1. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt
Result
OpenBenchmarking.org Frames Per Second, More Is Better SVT-HEVC 1.5.0 Tuning: 10 - Input: Bosphorus 1080p P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 120 240 360 480 600 SE +/- 1.12, N = 9 SE +/- 0.97, N = 11 SE +/- 3.58, N = 15 SE +/- 1.99, N = 11 258.65 576.99 345.48 574.88 1. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better SVT-HEVC 1.5.0 Tuning: 10 - Input: Bosphorus 1080p P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.6611 1.3222 1.9833 2.6444 3.3055 2.638 2.541 2.938 2.517 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 1922 2302 P-State performance 3400 3403 3405 intel_cpufreq schedutil 2300 2697 3403 intel_cpufreq performance 3400 3403 3404 OpenBenchmarking.org Megahertz, More Is Better SVT-HEVC 1.5.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 53.6 98.1 138.3 P-State performance 149.6 227.0 275.3 intel_cpufreq schedutil 56.9 117.6 166.9 intel_cpufreq performance 161.5 228.4 274.5 OpenBenchmarking.org Watts, Fewer Is Better SVT-HEVC 1.5.0 CPU Power Consumption Monitor 70 140 210 280 350 1. P-State powersave: Approximate power consumption of 174 Joules per run. 2. P-State performance: Approximate power consumption of 310 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 196 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 311 Joules per run. CPU Temp
Min Avg Max P-State powersave 31.0 33.1 36.0 P-State performance 47.0 48.9 57.0 intel_cpufreq schedutil 37.0 39.0 44.0 intel_cpufreq performance 48.0 50.3 60.0 OpenBenchmarking.org Celsius, Fewer Is Better SVT-HEVC 1.5.0 CPU Temperature Monitor 16 32 48 64 80 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better SVT-HEVC 1.5.0 Tuning: 10 - Input: Bosphorus 1080p P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 100 200 300 400 500 Min: 254.45 / Avg: 258.65 / Max: 262.93 Min: 572.52 / Avg: 576.99 / Max: 581.96 Min: 319.83 / Avg: 345.48 / Max: 366.75 Min: 559.18 / Avg: 574.88 / Max: 581.96 1. (CC) gcc options: -fPIE -fPIC -O3 -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.
Result
OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 100 200 300 400 500 SE +/- 1.53, N = 7 SE +/- 4.77, N = 15 SE +/- 2.47, N = 15 SE +/- 3.11, N = 13 206.54 453.67 260.25 457.02 1. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better SVT-VP9 0.3 Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.4671 0.9342 1.4013 1.8684 2.3355 1.939 1.922 2.076 1.919 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 1714 2305 P-State performance 3403 3403 3405 intel_cpufreq schedutil 801 2430 3396 intel_cpufreq performance 3400 3403 3405 OpenBenchmarking.org Megahertz, More Is Better SVT-VP9 0.3 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 53.7 106.5 160.7 P-State performance 171.6 236.0 292.6 intel_cpufreq schedutil 64.8 125.4 194.5 intel_cpufreq performance 158.6 238.2 290.6 OpenBenchmarking.org Watts, Fewer Is Better SVT-VP9 0.3 CPU Power Consumption Monitor 70 140 210 280 350 1. P-State powersave: Approximate power consumption of 243 Joules per run. 2. P-State performance: Approximate power consumption of 362 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 251 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 385 Joules per run. CPU Temp
Min Avg Max P-State powersave 30.0 32.1 36.0 P-State performance 46.0 48.0 55.0 intel_cpufreq schedutil 36.0 39.3 43.0 intel_cpufreq performance 47.0 49.0 55.0 OpenBenchmarking.org Celsius, Fewer Is Better SVT-VP9 0.3 CPU Temperature Monitor 15 30 45 60 75 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 80 160 240 320 400 Min: 200.57 / Avg: 206.54 / Max: 211.55 Min: 390.58 / Avg: 453.67 / Max: 468.13 Min: 245.36 / Avg: 260.25 / Max: 278.05 Min: 425.37 / Avg: 457.02 / Max: 470.89 1. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
Result
OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: VMAF Optimized - Input: Bosphorus 1080p P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 100 200 300 400 500 SE +/- 1.16, N = 7 SE +/- 2.03, N = 10 SE +/- 2.52, N = 15 SE +/- 2.50, N = 10 210.92 460.44 262.78 464.71 1. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better SVT-VP9 0.3 Tuning: VMAF Optimized - Input: Bosphorus 1080p P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.4572 0.9144 1.3716 1.8288 2.286 2.029 1.992 2.032 1.999 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 1800 2300 P-State performance 3403 3403 3404 intel_cpufreq schedutil 2300 2556 3404 intel_cpufreq performance 3403 3403 3405 OpenBenchmarking.org Megahertz, More Is Better SVT-VP9 0.3 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 53.7 104.0 158.2 P-State performance 152.0 231.2 283.3 intel_cpufreq schedutil 61.7 129.3 201.5 intel_cpufreq performance 176.9 232.4 289.1 OpenBenchmarking.org Watts, Fewer Is Better SVT-VP9 0.3 CPU Power Consumption Monitor 70 140 210 280 350 1. P-State powersave: Approximate power consumption of 223 Joules per run. 2. P-State performance: Approximate power consumption of 347 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 259 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 372 Joules per run. CPU Temp
Min Avg Max P-State powersave 30.0 31.3 33.0 P-State performance 46.0 47.9 53.0 intel_cpufreq schedutil 36.0 38.5 43.0 intel_cpufreq performance 47.0 48.6 55.0 OpenBenchmarking.org Celsius, Fewer Is Better SVT-VP9 0.3 CPU Temperature Monitor 15 30 45 60 75 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: VMAF Optimized - Input: Bosphorus 1080p P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 80 160 240 320 400 Min: 207.18 / Avg: 210.92 / Max: 216.06 Min: 453.87 / Avg: 460.44 / Max: 471.1 Min: 251.1 / Avg: 262.78 / Max: 280.26 Min: 451.72 / Avg: 464.71 / Max: 476.38 1. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
AOM AV1 OpenBenchmarking.org Frames Per Second Per Watt, More Is Better AOM AV1 3.1 Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.056 0.112 0.168 0.224 0.28 0.225 0.245 0.208 0.249
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better AOM AV1 3.1 Encoder Mode: Speed 6 Realtime - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0151 0.0302 0.0453 0.0604 0.0755 0.065 0.067 0.052 0.067
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better AOM AV1 3.1 Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0074 0.0148 0.0222 0.0296 0.037 0.031 0.032 0.025 0.033
Embree OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Crown P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0448 0.0896 0.1344 0.1792 0.224 0.183 0.185 0.199 0.181
Kvazaar OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Kvazaar 2.0 Video Input: Bosphorus 4K - Video Preset: Ultra Fast P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0389 0.0778 0.1167 0.1556 0.1945 0.173 0.135 0.155 0.138
AOM AV1 This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.1 Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 14 28 42 56 70 SE +/- 0.22, N = 3 SE +/- 0.27, N = 5 SE +/- 0.16, N = 15 SE +/- 0.43, N = 5 23.59 61.06 21.72 61.27 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better AOM AV1 3.1 Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.056 0.112 0.168 0.224 0.28 0.225 0.245 0.208 0.249 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 1833 2301 P-State performance 3403 3403 3404 intel_cpufreq schedutil 801 1528 2303 intel_cpufreq performance 3400 3403 3404 OpenBenchmarking.org Megahertz, More Is Better AOM AV1 3.1 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 53.6 105.0 116.9 P-State performance 190.4 249.4 290.0 intel_cpufreq schedutil 62.7 104.3 119.1 intel_cpufreq performance 148.3 246.5 288.9 OpenBenchmarking.org Watts, Fewer Is Better AOM AV1 3.1 CPU Power Consumption Monitor 70 140 210 280 350 1. P-State powersave: Approximate power consumption of 805 Joules per run. 2. P-State performance: Approximate power consumption of 898 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 849 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 887 Joules per run. CPU Temp
Min Avg Max P-State powersave 29.0 30.3 31.0 P-State performance 45.0 45.2 46.0 intel_cpufreq schedutil 35.0 36.6 38.0 intel_cpufreq performance 46.0 47.1 53.0 OpenBenchmarking.org Celsius, Fewer Is Better AOM AV1 3.1 CPU Temperature Monitor 15 30 45 60 75 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.1 Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 12 24 36 48 60 Min: 23.35 / Avg: 23.59 / Max: 24.02 Min: 60.3 / Avg: 61.06 / Max: 61.76 Min: 20.3 / Avg: 21.72 / Max: 22.82 Min: 60.29 / Avg: 61.27 / Max: 62.74 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
Embree OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Asian Dragon P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.065 0.13 0.195 0.26 0.325 0.289 0.274 0.274 0.264
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Asian Dragon Obj P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0691 0.1382 0.2073 0.2764 0.3455 0.307 0.291 0.300 0.289
AOM AV1 This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.1 Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 11 22 33 44 55 SE +/- 0.02, N = 3 SE +/- 0.30, N = 4 SE +/- 0.03, N = 3 SE +/- 0.52, N = 4 17.07 48.62 15.90 48.91 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better AOM AV1 3.1 Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0428 0.0856 0.1284 0.1712 0.214 0.163 0.190 0.151 0.190 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 1794 2287 P-State performance 3403 3404 3407 intel_cpufreq schedutil 810 1372 3202 intel_cpufreq performance 3400 3403 3404 OpenBenchmarking.org Megahertz, More Is Better AOM AV1 3.1 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 67.7 104.9 115.9 P-State performance 186.2 255.5 299.2 intel_cpufreq schedutil 67.0 105.1 114.8 intel_cpufreq performance 181.7 257.1 310.9 OpenBenchmarking.org Watts, Fewer Is Better AOM AV1 3.1 CPU Power Consumption Monitor 80 160 240 320 400 1. P-State powersave: Approximate power consumption of 1049 Joules per run. 2. P-State performance: Approximate power consumption of 1086 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1121 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1093 Joules per run. CPU Temp
Min Avg Max P-State powersave 29.0 31.2 32.0 P-State performance 44.0 44.3 45.0 intel_cpufreq schedutil 34.0 35.9 37.0 intel_cpufreq performance 45.0 45.7 47.0 OpenBenchmarking.org Celsius, Fewer Is Better AOM AV1 3.1 CPU Temperature Monitor 14 28 42 56 70 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.1 Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 10 20 30 40 50 Min: 17.05 / Avg: 17.07 / Max: 17.11 Min: 47.91 / Avg: 48.62 / Max: 49.31 Min: 15.86 / Avg: 15.9 / Max: 15.96 Min: 47.43 / Avg: 48.91 / Max: 49.75 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
Result
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.1 Encoder Mode: Speed 6 Realtime - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 5 10 15 20 25 SE +/- 0.07, N = 3 SE +/- 0.11, N = 3 SE +/- 0.02, N = 3 SE +/- 0.23, N = 3 7.38 19.09 5.86 19.20 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better AOM AV1 3.1 Encoder Mode: Speed 6 Realtime - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0151 0.0302 0.0453 0.0604 0.0755 0.065 0.067 0.052 0.067 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2130 2305 P-State performance 3400 3403 3404 intel_cpufreq schedutil 1001 1788 3403 intel_cpufreq performance 3403 3404 3422 OpenBenchmarking.org Megahertz, More Is Better AOM AV1 3.1 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 54.3 114.4 126.6 P-State performance 148.8 283.3 319.8 intel_cpufreq schedutil 60.7 111.7 120.3 intel_cpufreq performance 153.7 285.8 321.6 OpenBenchmarking.org Watts, Fewer Is Better AOM AV1 3.1 CPU Power Consumption Monitor 80 160 240 320 400 1. P-State powersave: Approximate power consumption of 2402 Joules per run. 2. P-State performance: Approximate power consumption of 2456 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2905 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2477 Joules per run. CPU Temp
Min Avg Max P-State powersave 30.0 32.8 34.0 P-State performance 43.0 44.8 46.0 intel_cpufreq schedutil 34.0 36.5 38.0 intel_cpufreq performance 45.0 48.7 56.0 OpenBenchmarking.org Celsius, Fewer Is Better AOM AV1 3.1 CPU Temperature Monitor 16 32 48 64 80 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.1 Encoder Mode: Speed 6 Realtime - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 5 10 15 20 25 Min: 7.26 / Avg: 7.38 / Max: 7.49 Min: 18.93 / Avg: 19.09 / Max: 19.29 Min: 5.84 / Avg: 5.86 / Max: 5.91 Min: 18.77 / Avg: 19.2 / Max: 19.55 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
Result
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.1 Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 SE +/- 0.05, N = 3 SE +/- 0.03, N = 3 SE +/- 0.08, N = 12 SE +/- 0.09, N = 3 3.94 9.69 4.00 10.01 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better AOM AV1 3.1 Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0074 0.0148 0.0222 0.0296 0.037 0.031 0.032 0.025 0.033 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2263 3318 P-State performance 3400 3404 3434 intel_cpufreq schedutil 908 2963 3409 intel_cpufreq performance 3403 3404 3409 OpenBenchmarking.org Megahertz, More Is Better AOM AV1 3.1 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 54.6 125.5 223.4 P-State performance 203.7 301.4 346.5 intel_cpufreq schedutil 64.3 160.5 249.6 intel_cpufreq performance 188.6 303.1 339.2 OpenBenchmarking.org Watts, Fewer Is Better AOM AV1 3.1 CPU Power Consumption Monitor 100 200 300 400 500 1. P-State powersave: Approximate power consumption of 6106 Joules per run. 2. P-State performance: Approximate power consumption of 6531 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 7958 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 6467 Joules per run. CPU Temp
Min Avg Max P-State powersave 31.0 35.5 41.0 P-State performance 45.0 51.6 62.0 intel_cpufreq schedutil 34.0 42.8 58.0 intel_cpufreq performance 45.0 47.2 53.0 OpenBenchmarking.org Celsius, Fewer Is Better AOM AV1 3.1 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.1 Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4K P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 Min: 3.88 / Avg: 3.94 / Max: 4.04 Min: 9.64 / Avg: 9.69 / Max: 9.75 Min: 3.65 / Avg: 4 / Max: 4.57 Min: 9.83 / Avg: 10.01 / Max: 10.13 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm -lpthread
Embree Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Asian Dragon P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 20 40 60 80 100 SE +/- 0.79, N = 6 SE +/- 0.42, N = 6 SE +/- 0.80, N = 6 SE +/- 0.56, N = 6 106.62 107.63 102.70 106.17 MIN: 84.26 / MAX: 112.87 MIN: 88.63 / MAX: 112.78 MIN: 81.43 / MAX: 111.73 MIN: 84.09 / MAX: 112.29 Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Asian Dragon P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.065 0.13 0.195 0.26 0.325 0.289 0.274 0.274 0.264 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2377 3012 P-State performance 2666 3138 3403 intel_cpufreq schedutil 1381 2603 3400 intel_cpufreq performance 2651 3144 3405 OpenBenchmarking.org Megahertz, More Is Better Embree 3.13 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 141.2 368.4 592.4 P-State performance 154.3 392.3 597.7 intel_cpufreq schedutil 157.4 374.6 575.1 intel_cpufreq performance 166.4 402.7 597.3 OpenBenchmarking.org Watts, Fewer Is Better Embree 3.13 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 982 Joules per run. 2. P-State performance: Approximate power consumption of 981 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 999 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1007 Joules per run. CPU Temp
Min Avg Max P-State powersave 52.0 62.4 70.0 P-State performance 55.0 63.7 71.0 intel_cpufreq schedutil 57.0 65.4 71.0 intel_cpufreq performance 56.0 65.3 72.0 OpenBenchmarking.org Celsius, Fewer Is Better Embree 3.13 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Asian Dragon P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 20 40 60 80 100 Min: 103.41 / Avg: 106.62 / Max: 108.64 Min: 106.18 / Avg: 107.63 / Max: 108.97 Min: 100.92 / Avg: 102.7 / Max: 105.51 Min: 104.91 / Avg: 106.17 / Max: 108.11
Result
OpenBenchmarking.org Frames Per Second, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Asian Dragon Obj P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 20 40 60 80 100 SE +/- 0.41, N = 3 SE +/- 0.48, N = 3 SE +/- 0.38, N = 3 SE +/- 0.14, N = 3 87.21 89.09 87.99 89.49 MIN: 69.09 / MAX: 97.86 MIN: 70.73 / MAX: 97.89 MIN: 70.3 / MAX: 98.24 MIN: 71.05 / MAX: 98.17 Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Asian Dragon Obj P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0691 0.1382 0.2073 0.2764 0.3455 0.307 0.291 0.300 0.289 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 3030 3404 P-State performance 2668 3215 3405 intel_cpufreq schedutil 1100 2995 3404 intel_cpufreq performance 2691 3233 3410 OpenBenchmarking.org Megahertz, More Is Better Embree 3.13 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 102.3 284.4 554.8 P-State performance 163.0 305.7 571.3 intel_cpufreq schedutil 128.2 293.2 576.1 intel_cpufreq performance 151.5 310.0 582.8 OpenBenchmarking.org Watts, Fewer Is Better Embree 3.13 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 1801 Joules per run. 2. P-State performance: Approximate power consumption of 1834 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1857 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1963 Joules per run. CPU Temp
Min Avg Max P-State powersave 52.0 56.4 66.0 P-State performance 54.0 57.5 67.0 intel_cpufreq schedutil 54.0 58.8 70.0 intel_cpufreq performance 55.0 58.8 68.0 OpenBenchmarking.org Celsius, Fewer Is Better Embree 3.13 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Asian Dragon Obj P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 20 40 60 80 100 Min: 86.76 / Avg: 87.21 / Max: 88.02 Min: 88.34 / Avg: 89.09 / Max: 89.97 Min: 87.27 / Avg: 87.99 / Max: 88.53 Min: 89.23 / Avg: 89.49 / Max: 89.67
Result
OpenBenchmarking.org Frames Per Second, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Crown P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 16 32 48 64 80 SE +/- 0.21, N = 5 SE +/- 0.18, N = 5 SE +/- 0.18, N = 5 SE +/- 0.34, N = 5 66.53 74.06 67.75 74.24 MIN: 59.75 / MAX: 87.7 MIN: 64.5 / MAX: 94.28 MIN: 60.74 / MAX: 88.34 MIN: 64.44 / MAX: 93.9 Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Crown P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0448 0.0896 0.1344 0.1792 0.224 0.183 0.185 0.199 0.181 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2384 2984 P-State performance 2685 3094 3404 intel_cpufreq schedutil 1800 2659 3400 intel_cpufreq performance 2723 3098 3409 OpenBenchmarking.org Megahertz, More Is Better Embree 3.13 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 109.8 362.7 513.8 P-State performance 198.9 399.5 489.3 intel_cpufreq schedutil 121.3 340.0 510.0 intel_cpufreq performance 191.2 410.3 575.3 OpenBenchmarking.org Watts, Fewer Is Better Embree 3.13 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 1088 Joules per run. 2. P-State performance: Approximate power consumption of 1199 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1088 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1231 Joules per run. CPU Temp
Min Avg Max P-State powersave 50.0 60.5 67.0 P-State performance 53.0 64.4 71.0 intel_cpufreq schedutil 53.0 62.5 69.0 intel_cpufreq performance 54.0 65.5 71.0 OpenBenchmarking.org Celsius, Fewer Is Better Embree 3.13 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Crown P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 14 28 42 56 70 Min: 65.76 / Avg: 66.53 / Max: 66.96 Min: 73.67 / Avg: 74.06 / Max: 74.72 Min: 67.38 / Avg: 67.75 / Max: 68.43 Min: 73.64 / Avg: 74.24 / Max: 75.52
Kvazaar This is a test of Kvazaar as a CPU-based H.265 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.
Result
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.0 Video Input: Bosphorus 4K - Video Preset: Very Fast P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 9 18 27 36 45 SE +/- 0.04, N = 3 SE +/- 0.07, N = 4 SE +/- 0.02, N = 3 SE +/- 0.07, N = 4 13.26 38.21 11.30 38.26 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Kvazaar 2.0 Video Input: Bosphorus 4K - Video Preset: Very Fast P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0225 0.045 0.0675 0.09 0.1125 0.095 0.098 0.084 0.100 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2163 2304 P-State performance 3187 3327 3403 intel_cpufreq schedutil 1104 2065 3400 intel_cpufreq performance 2206 3266 3403 OpenBenchmarking.org Megahertz, More Is Better Kvazaar 2.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 67.1 139.8 156.9 P-State performance 211.4 388.5 491.7 intel_cpufreq schedutil 70.9 134.1 148.8 intel_cpufreq performance 169.5 382.7 492.0 OpenBenchmarking.org Watts, Fewer Is Better Kvazaar 2.0 CPU Power Consumption Monitor 130 260 390 520 650 1. P-State powersave: Approximate power consumption of 1631 Joules per run. 2. P-State performance: Approximate power consumption of 1748 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1787 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1722 Joules per run. CPU Temp
Min Avg Max P-State powersave 38.0 40.6 43.0 P-State performance 48.0 58.6 65.0 intel_cpufreq schedutil 38.0 40.9 43.0 intel_cpufreq performance 46.0 58.2 66.0 OpenBenchmarking.org Celsius, Fewer Is Better Kvazaar 2.0 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.0 Video Input: Bosphorus 4K - Video Preset: Very Fast P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 8 16 24 32 40 Min: 13.19 / Avg: 13.26 / Max: 13.32 Min: 38.06 / Avg: 38.21 / Max: 38.39 Min: 11.28 / Avg: 11.3 / Max: 11.34 Min: 38.1 / Avg: 38.26 / Max: 38.41 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
Result
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.0 Video Input: Bosphorus 4K - Video Preset: Ultra Fast P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 11 22 33 44 55 SE +/- 0.14, N = 3 SE +/- 0.31, N = 15 SE +/- 0.06, N = 3 SE +/- 0.56, N = 4 23.03 46.31 20.53 46.34 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Kvazaar 2.0 Video Input: Bosphorus 4K - Video Preset: Ultra Fast P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0389 0.0778 0.1167 0.1556 0.1945 0.173 0.135 0.155 0.138 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2175 2300 P-State performance 3117 3298 3403 intel_cpufreq schedutil 1141 2222 2303 intel_cpufreq performance 3101 3294 3404 OpenBenchmarking.org Megahertz, More Is Better Kvazaar 2.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 57.1 133.2 155.1 P-State performance 183.3 343.8 438.5 intel_cpufreq schedutil 55.9 132.0 150.2 intel_cpufreq performance 191.2 335.5 439.8 OpenBenchmarking.org Watts, Fewer Is Better Kvazaar 2.0 CPU Power Consumption Monitor 120 240 360 480 600 1. P-State powersave: Approximate power consumption of 933 Joules per run. 2. P-State performance: Approximate power consumption of 1329 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1012 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1342 Joules per run. CPU Temp
Min Avg Max P-State powersave 36.0 38.3 39.0 P-State performance 54.0 60.7 65.0 intel_cpufreq schedutil 37.0 38.7 39.0 intel_cpufreq performance 54.0 60.6 63.0 OpenBenchmarking.org Celsius, Fewer Is Better Kvazaar 2.0 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.0 Video Input: Bosphorus 4K - Video Preset: Ultra Fast P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 9 18 27 36 45 Min: 22.85 / Avg: 23.03 / Max: 23.3 Min: 44.32 / Avg: 46.31 / Max: 48.17 Min: 20.42 / Avg: 20.53 / Max: 20.6 Min: 45.34 / Avg: 46.34 / Max: 47.67 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
High Performance Conjugate Gradient HPCG is the High Performance Conjugate Gradient and is a new scientific benchmark from Sandia National Lans focused for super-computer testing with modern real-world workloads compared to HPCC. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org GFLOP/s, More Is Better High Performance Conjugate Gradient 3.1 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 9 18 27 36 45 SE +/- 0.17, N = 3 SE +/- 0.20, N = 3 SE +/- 0.12, N = 3 SE +/- 0.21, N = 3 40.14 40.65 40.07 40.83 1. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -pthread -lmpi_cxx -lmpi GFLOP/s Per Watt
OpenBenchmarking.org GFLOP/s Per Watt, More Is Better High Performance Conjugate Gradient 3.1 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0173 0.0346 0.0519 0.0692 0.0865 0.076 0.076 0.076 0.077 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 802 2988 3390 P-State performance 3002 3025 3403 intel_cpufreq schedutil 2200 3003 3403 intel_cpufreq performance 3002 3023 3403 OpenBenchmarking.org Megahertz, More Is Better High Performance Conjugate Gradient 3.1 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 101.0 525.4 566.6 P-State performance 192.8 531.4 567.4 intel_cpufreq schedutil 148.5 526.9 567.0 intel_cpufreq performance 185.0 529.9 567.5 OpenBenchmarking.org Watts, Fewer Is Better High Performance Conjugate Gradient 3.1 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 24345 Joules per run. 2. P-State performance: Approximate power consumption of 23737 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 24587 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 24023 Joules per run. CPU Temp
Min Avg Max P-State powersave 37.0 66.6 75.0 P-State performance 46.0 66.4 74.0 intel_cpufreq schedutil 43.0 67.0 76.0 intel_cpufreq performance 45.0 66.5 75.0 OpenBenchmarking.org Celsius, Fewer Is Better High Performance Conjugate Gradient 3.1 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org GFLOP/s, More Is Better High Performance Conjugate Gradient 3.1 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 8 16 24 32 40 Min: 39.95 / Avg: 40.14 / Max: 40.47 Min: 40.35 / Avg: 40.65 / Max: 41.03 Min: 39.92 / Avg: 40.07 / Max: 40.3 Min: 40.42 / Avg: 40.83 / Max: 41.04 1. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -pthread -lmpi_cxx -lmpi
ArrayFire ArrayFire is an GPU and CPU numeric processing library, this test uses the built-in CPU and OpenCL ArrayFire benchmarks. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org GFLOPS, More Is Better ArrayFire 3.7 Test: BLAS CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1000 2000 3000 4000 5000 SE +/- 10.80, N = 4 SE +/- 24.52, N = 4 SE +/- 36.10, N = 4 SE +/- 11.74, N = 4 4538.69 4539.99 4510.29 4574.83 1. (CXX) g++ options: -rdynamic GFLOPS Per Watt
OpenBenchmarking.org GFLOPS Per Watt, More Is Better ArrayFire 3.7 Test: BLAS CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 12.87 12.17 12.46 12.33 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 802 3051 3408 P-State performance 3403 3407 3437 intel_cpufreq schedutil 3300 3386 3408 intel_cpufreq performance 2233 3338 3433 OpenBenchmarking.org Megahertz, More Is Better ArrayFire 3.7 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 151.4 352.8 502.8 P-State performance 193.5 373.0 497.4 intel_cpufreq schedutil 142.1 362.1 489.2 intel_cpufreq performance 162.2 371.0 504.1 OpenBenchmarking.org Watts, Fewer Is Better ArrayFire 3.7 CPU Power Consumption Monitor 130 260 390 520 650 1. P-State powersave: Approximate power consumption of 1499 Joules per run. 2. P-State performance: Approximate power consumption of 1492 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1539 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1577 Joules per run. CPU Temp
Min Avg Max P-State powersave 44.0 52.1 61.0 P-State performance 47.0 52.1 63.0 intel_cpufreq schedutil 46.0 50.9 62.0 intel_cpufreq performance 47.0 54.6 63.0 OpenBenchmarking.org Celsius, Fewer Is Better ArrayFire 3.7 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org GFLOPS, More Is Better ArrayFire 3.7 Test: BLAS CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 800 1600 2400 3200 4000 Min: 4506.93 / Avg: 4538.69 / Max: 4554.9 Min: 4478.24 / Avg: 4539.99 / Max: 4597.81 Min: 4426.25 / Avg: 4510.29 / Max: 4580.19 Min: 4545.38 / Avg: 4574.83 / Max: 4602.84 1. (CXX) g++ options: -rdynamic
Intel Open Image Denoise Open Image Denoise is a denoising library for ray-tracing and part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 1.4.0 Run: RT.hdr_alb_nrm.3840x2160 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.6593 1.3186 1.9779 2.6372 3.2965 SE +/- 0.00, N = 4 SE +/- 0.01, N = 5 SE +/- 0.03, N = 15 SE +/- 0.00, N = 5 2.85 2.93 2.78 2.93 Images / Sec Per Watt
OpenBenchmarking.org Images / Sec Per Watt, More Is Better Intel Open Image Denoise 1.4.0 Run: RT.hdr_alb_nrm.3840x2160 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.002 0.004 0.006 0.008 0.01 0.009 0.008 0.009 0.008 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2026 3174 P-State performance 2636 3223 3404 intel_cpufreq schedutil 1088 2854 3403 intel_cpufreq performance 2931 3286 3405 OpenBenchmarking.org Megahertz, More Is Better Intel Open Image Denoise 1.4.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 84.7 307.4 558.3 P-State performance 154.8 346.3 560.7 intel_cpufreq schedutil 59.7 321.9 551.2 intel_cpufreq performance 158.5 349.7 562.6 OpenBenchmarking.org Watts, Fewer Is Better Intel Open Image Denoise 1.4.0 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 1153 Joules per run. 2. P-State performance: Approximate power consumption of 1247 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1202 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1259 Joules per run. CPU Temp
Min Avg Max P-State powersave 49.0 58.1 68.0 P-State performance 55.0 62.1 73.0 intel_cpufreq schedutil 50.0 60.1 72.0 intel_cpufreq performance 55.0 62.7 72.0 OpenBenchmarking.org Celsius, Fewer Is Better Intel Open Image Denoise 1.4.0 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 1.4.0 Run: RT.hdr_alb_nrm.3840x2160 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 2 4 6 8 10 Min: 2.85 / Avg: 2.85 / Max: 2.86 Min: 2.91 / Avg: 2.93 / Max: 2.94 Min: 2.54 / Avg: 2.78 / Max: 2.87 Min: 2.92 / Avg: 2.93 / Max: 2.94
OpenBenchmarking.org Images / Sec Per Watt, More Is Better Intel Open Image Denoise 1.4.0 Run: RT.ldr_alb_nrm.3840x2160 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.002 0.004 0.006 0.008 0.01 0.009 0.008 0.009 0.009
OpenBenchmarking.org Images / Sec Per Watt, More Is Better Intel Open Image Denoise 1.4.0 Run: RTLightmap.hdr.4096x4096 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0009 0.0018 0.0027 0.0036 0.0045 0.004 0.004 0.004 0.004
Result
OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 1.4.0 Run: RT.ldr_alb_nrm.3840x2160 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.6593 1.3186 1.9779 2.6372 3.2965 SE +/- 0.01, N = 4 SE +/- 0.01, N = 5 SE +/- 0.02, N = 4 SE +/- 0.01, N = 5 2.87 2.93 2.88 2.93 Images / Sec Per Watt
OpenBenchmarking.org Images / Sec Per Watt, More Is Better Intel Open Image Denoise 1.4.0 Run: RT.ldr_alb_nrm.3840x2160 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.002 0.004 0.006 0.008 0.01 0.009 0.008 0.009 0.009 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2120 3300 P-State performance 2861 3319 3404 intel_cpufreq schedutil 2125 3050 3403 intel_cpufreq performance 2900 3286 3404 OpenBenchmarking.org Megahertz, More Is Better Intel Open Image Denoise 1.4.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 71.7 336.6 551.4 P-State performance 159.5 346.4 561.4 intel_cpufreq schedutil 89.9 324.7 557.4 intel_cpufreq performance 179.4 341.8 560.5 OpenBenchmarking.org Watts, Fewer Is Better Intel Open Image Denoise 1.4.0 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 1094 Joules per run. 2. P-State performance: Approximate power consumption of 1247 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1218 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1230 Joules per run. CPU Temp
Min Avg Max P-State powersave 46.0 56.9 67.0 P-State performance 52.0 59.8 69.0 intel_cpufreq schedutil 50.0 59.8 72.0 intel_cpufreq performance 54.0 60.2 69.0 OpenBenchmarking.org Celsius, Fewer Is Better Intel Open Image Denoise 1.4.0 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 1.4.0 Run: RT.ldr_alb_nrm.3840x2160 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 2 4 6 8 10 Min: 2.84 / Avg: 2.87 / Max: 2.89 Min: 2.89 / Avg: 2.93 / Max: 2.95 Min: 2.84 / Avg: 2.88 / Max: 2.93 Min: 2.91 / Avg: 2.93 / Max: 2.95
Result
OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 1.4.0 Run: RTLightmap.hdr.4096x4096 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.315 0.63 0.945 1.26 1.575 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 1.38 1.40 1.38 1.40 Images / Sec Per Watt
OpenBenchmarking.org Images / Sec Per Watt, More Is Better Intel Open Image Denoise 1.4.0 Run: RTLightmap.hdr.4096x4096 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0009 0.0018 0.0027 0.0036 0.0045 0.004 0.004 0.004 0.004 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2187 3299 P-State performance 2851 3313 3411 intel_cpufreq schedutil 2301 3089 3404 intel_cpufreq performance 2804 3247 3403 OpenBenchmarking.org Megahertz, More Is Better Intel Open Image Denoise 1.4.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 75.2 357.4 558.0 P-State performance 151.6 384.8 559.5 intel_cpufreq schedutil 92.3 363.8 561.6 intel_cpufreq performance 186.9 380.7 559.6 OpenBenchmarking.org Watts, Fewer Is Better Intel Open Image Denoise 1.4.0 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 2145 Joules per run. 2. P-State performance: Approximate power consumption of 2309 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2304 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2284 Joules per run. CPU Temp
Min Avg Max P-State powersave 45.0 60.3 72.0 P-State performance 52.0 61.3 72.0 intel_cpufreq schedutil 48.0 63.0 72.0 intel_cpufreq performance 53.0 62.8 72.0 OpenBenchmarking.org Celsius, Fewer Is Better Intel Open Image Denoise 1.4.0 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 1.4.0 Run: RTLightmap.hdr.4096x4096 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 2 4 6 8 10 Min: 1.38 / Avg: 1.38 / Max: 1.39 Min: 1.39 / Avg: 1.4 / Max: 1.41 Min: 1.38 / Avg: 1.38 / Max: 1.39 Min: 1.4 / Avg: 1.4 / Max: 1.4
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.
Result
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.6 Model: yolov4 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 100 200 300 400 500 SE +/- 2.02, N = 3 SE +/- 5.43, N = 3 SE +/- 5.29, N = 3 SE +/- 0.73, N = 3 480 483 483 476 1. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt Inferences Per Minute Per Watt
OpenBenchmarking.org Inferences Per Minute Per Watt, More Is Better ONNX Runtime 1.6 Model: yolov4 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.2412 0.4824 0.7236 0.9648 1.206 1.072 1.063 1.068 1.054 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 802 2904 3007 P-State performance 2930 2992 3404 intel_cpufreq schedutil 2922 2975 3405 intel_cpufreq performance 2927 2996 3406 OpenBenchmarking.org Megahertz, More Is Better ONNX Runtime 1.6 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 154.5 447.7 550.4 P-State performance 179.5 454.5 548.4 intel_cpufreq schedutil 124.6 452.1 549.9 intel_cpufreq performance 207.9 451.6 544.9 OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.6 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 11939 Joules per run. 2. P-State performance: Approximate power consumption of 12271 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 11905 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 12193 Joules per run. CPU Temp
Min Avg Max P-State powersave 54.0 68.1 72.0 P-State performance 57.0 68.6 73.0 intel_cpufreq schedutil 55.0 69.0 72.0 intel_cpufreq performance 51.0 61.6 63.0 OpenBenchmarking.org Celsius, Fewer Is Better ONNX Runtime 1.6 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.6 Model: yolov4 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 90 180 270 360 450 Min: 476 / Avg: 480 / Max: 482.5 Min: 477 / Avg: 483.17 / Max: 494 Min: 474.5 / Avg: 482.5 / Max: 492.5 Min: 475 / Avg: 476.17 / Max: 477.5 1. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt
Result
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.6 Model: fcn-resnet101-11 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 40 80 120 160 200 SE +/- 2.36, N = 3 SE +/- 0.44, N = 3 SE +/- 1.89, N = 3 SE +/- 0.29, N = 3 196 194 194 197 1. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt Inferences Per Minute Per Watt
OpenBenchmarking.org Inferences Per Minute Per Watt, More Is Better ONNX Runtime 1.6 Model: fcn-resnet101-11 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.1004 0.2008 0.3012 0.4016 0.502 0.446 0.435 0.434 0.440 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2880 3400 P-State performance 2866 2951 3404 intel_cpufreq schedutil 2177 2919 3404 intel_cpufreq performance 2855 2940 3404 OpenBenchmarking.org Megahertz, More Is Better ONNX Runtime 1.6 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 99.4 439.7 539.2 P-State performance 156.5 446.0 541.2 intel_cpufreq schedutil 111.1 447.2 540.9 intel_cpufreq performance 204.1 447.9 541.1 OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.6 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 11580 Joules per run. 2. P-State performance: Approximate power consumption of 12042 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 11924 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 11943 Joules per run. CPU Temp
Min Avg Max P-State powersave 51.0 66.7 74.0 P-State performance 54.0 68.1 74.0 intel_cpufreq schedutil 52.0 66.4 71.0 intel_cpufreq performance 49.0 63.5 68.0 OpenBenchmarking.org Celsius, Fewer Is Better ONNX Runtime 1.6 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.6 Model: fcn-resnet101-11 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 40 80 120 160 200 Min: 192 / Avg: 195.5 / Max: 200 Min: 193.5 / Avg: 194.17 / Max: 195 Min: 191 / Avg: 194 / Max: 197.5 Min: 196 / Avg: 196.5 / Max: 197 1. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt
OpenBenchmarking.org Inferences Per Minute Per Watt, More Is Better ONNX Runtime 1.6 Model: shufflenet-v2-10 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 5 10 15 20 25 19.43 19.92 19.67 19.87
OpenBenchmarking.org Inferences Per Minute Per Watt, More Is Better ONNX Runtime 1.6 Model: super-resolution-10 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 4 8 12 16 20 14.99 14.43 14.62 15.03
OpenBenchmarking.org Inferences Per Minute Per Watt, More Is Better ONNX Runtime 1.6 Model: bertsquad-10 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.2567 0.5134 0.7701 1.0268 1.2835 1.141 1.064 1.117 1.113
Result
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.6 Model: shufflenet-v2-10 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 2K 4K 6K 8K 10K SE +/- 58.56, N = 12 SE +/- 25.12, N = 3 SE +/- 65.66, N = 3 SE +/- 24.06, N = 3 8212 8392 8334 8380 1. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt Inferences Per Minute Per Watt
OpenBenchmarking.org Inferences Per Minute Per Watt, More Is Better ONNX Runtime 1.6 Model: shufflenet-v2-10 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 5 10 15 20 25 19.43 19.92 19.67 19.87 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2974 3005 P-State performance 2943 3004 3403 intel_cpufreq schedutil 2943 3008 3403 intel_cpufreq performance 2944 3012 3403 OpenBenchmarking.org Megahertz, More Is Better ONNX Runtime 1.6 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 138.0 422.6 516.0 P-State performance 190.2 421.2 500.4 intel_cpufreq schedutil 185.5 423.7 515.2 intel_cpufreq performance 155.3 421.8 507.2 OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.6 CPU Power Consumption Monitor 130 260 390 520 650 1. P-State powersave: Approximate power consumption of 11269 Joules per run. 2. P-State performance: Approximate power consumption of 11654 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 11297 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 11529 Joules per run. CPU Temp
Min Avg Max P-State powersave 52.0 64.5 67.0 P-State performance 55.0 63.0 64.0 intel_cpufreq schedutil 54.0 63.8 65.0 intel_cpufreq performance 52.0 60.3 62.0 OpenBenchmarking.org Celsius, Fewer Is Better ONNX Runtime 1.6 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.6 Model: shufflenet-v2-10 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1500 3000 4500 6000 7500 Min: 7618 / Avg: 8212.08 / Max: 8342 Min: 8348.5 / Avg: 8392.33 / Max: 8435.5 Min: 8208.5 / Avg: 8334.17 / Max: 8430 Min: 8348.5 / Avg: 8380.33 / Max: 8427.5 1. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt
Result
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.6 Model: super-resolution-10 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1500 3000 4500 6000 7500 SE +/- 205.17, N = 12 SE +/- 200.91, N = 12 SE +/- 210.11, N = 12 SE +/- 156.45, N = 12 6898 6626 6756 6955 1. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt Inferences Per Minute Per Watt
OpenBenchmarking.org Inferences Per Minute Per Watt, More Is Better ONNX Runtime 1.6 Model: super-resolution-10 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 4 8 12 16 20 14.99 14.43 14.62 15.03 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2987 3006 P-State performance 3000 3022 3406 intel_cpufreq schedutil 3000 3013 3405 intel_cpufreq performance 3000 3023 3403 OpenBenchmarking.org Megahertz, More Is Better ONNX Runtime 1.6 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 116.2 460.3 567.1 P-State performance 187.4 459.1 563.8 intel_cpufreq schedutil 196.6 462.3 560.9 intel_cpufreq performance 175.7 462.8 568.1 OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.6 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 12351 Joules per run. 2. P-State performance: Approximate power consumption of 12663 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 12713 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 12727 Joules per run. CPU Temp
Min Avg Max P-State powersave 53.0 67.3 72.0 P-State performance 52.0 70.5 75.0 intel_cpufreq schedutil 52.0 70.6 75.0 intel_cpufreq performance 50.0 66.1 74.0 OpenBenchmarking.org Celsius, Fewer Is Better ONNX Runtime 1.6 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.6 Model: super-resolution-10 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1200 2400 3600 4800 6000 Min: 5184 / Avg: 6897.71 / Max: 7233 Min: 5152 / Avg: 6625.63 / Max: 7219.5 Min: 5207 / Avg: 6755.5 / Max: 7271 Min: 5786 / Avg: 6955 / Max: 7263 1. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt
Result
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.6 Model: bertsquad-10 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 110 220 330 440 550 SE +/- 10.98, N = 9 SE +/- 2.17, N = 3 SE +/- 8.53, N = 12 SE +/- 6.15, N = 12 509 477 504 503 1. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt Inferences Per Minute Per Watt
OpenBenchmarking.org Inferences Per Minute Per Watt, More Is Better ONNX Runtime 1.6 Model: bertsquad-10 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.2567 0.5134 0.7701 1.0268 1.2835 1.141 1.064 1.117 1.113 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2967 3401 P-State performance 3000 3025 3407 intel_cpufreq schedutil 2202 3018 3403 intel_cpufreq performance 3000 3028 3404 OpenBenchmarking.org Megahertz, More Is Better ONNX Runtime 1.6 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 134.9 446.2 551.9 P-State performance 154.6 448.2 538.5 intel_cpufreq schedutil 171.1 451.4 550.5 intel_cpufreq performance 182.9 451.8 550.2 OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.6 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 11899 Joules per run. 2. P-State performance: Approximate power consumption of 12400 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 12300 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 12312 Joules per run. CPU Temp
Min Avg Max P-State powersave 50.0 64.8 70.0 P-State performance 54.0 65.1 67.0 intel_cpufreq schedutil 53.0 66.6 70.0 intel_cpufreq performance 53.0 68.2 71.0 OpenBenchmarking.org Celsius, Fewer Is Better ONNX Runtime 1.6 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.6 Model: bertsquad-10 - Device: OpenMP CPU P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 90 180 270 360 450 Min: 461 / Avg: 508.5 / Max: 546 Min: 473.5 / Avg: 477.17 / Max: 481 Min: 472.5 / Avg: 503.96 / Max: 545 Min: 471 / Avg: 503 / Max: 534.5 1. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt
OpenVKL OpenBenchmarking.org Items / Sec Per Watt, More Is Better OpenVKL 0.9 Benchmark: vklBenchmark P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.8224 1.6448 2.4672 3.2896 4.112 3.536 2.896 3.655 2.914
OpenBenchmarking.org Items / Sec Per Watt, More Is Better OpenVKL 0.9 Benchmark: vklBenchmarkStructuredVolume P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 90K 180K 270K 360K 450K 424481.55 438562.32 418290.31 436355.85
OpenBenchmarking.org Items / Sec Per Watt, More Is Better OpenVKL 0.9 Benchmark: vklBenchmarkUnstructuredVolume P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1600 3200 4800 6400 8000 7267.58 6287.40 7409.47 6326.14
OpenBenchmarking.org Items / Sec Per Watt, More Is Better OpenVKL 0.9 Benchmark: vklBenchmarkVdbVolume P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 30K 60K 90K 120K 150K 123147.50 125723.06 118699.08 126322.91
Result
OpenBenchmarking.org Items / Sec, More Is Better OpenVKL 0.9 Benchmark: vklBenchmark P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 200 400 600 800 1000 SE +/- 2.19, N = 3 SE +/- 2.67, N = 3 SE +/- 8.72, N = 3 SE +/- 2.40, N = 3 670 811 718 821 MIN: 1 / MAX: 2830 MIN: 1 / MAX: 3166 MIN: 1 / MAX: 3134 MIN: 1 / MAX: 3226 Items / Sec Per Watt
OpenBenchmarking.org Items / Sec Per Watt, More Is Better OpenVKL 0.9 Benchmark: vklBenchmark P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.8224 1.6448 2.4672 3.2896 4.112 3.536 2.896 3.655 2.914 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2764 3436 P-State performance 3002 3379 3428 intel_cpufreq schedutil 801 2708 3426 intel_cpufreq performance 3002 3388 3424 OpenBenchmarking.org Megahertz, More Is Better OpenVKL 0.9 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 66.8 189.5 352.9 P-State performance 176.7 280.1 541.6 intel_cpufreq schedutil 72.2 196.4 369.2 intel_cpufreq performance 177.1 281.8 555.7 OpenBenchmarking.org Watts, Fewer Is Better OpenVKL 0.9 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 30316 Joules per run. 2. P-State performance: Approximate power consumption of 30530 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 30579 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 29679 Joules per run. CPU Temp
Min Avg Max P-State powersave 38.0 43.5 56.0 P-State performance 48.0 54.2 65.0 intel_cpufreq schedutil 42.0 47.9 61.0 intel_cpufreq performance 51.0 56.6 69.0 OpenBenchmarking.org Celsius, Fewer Is Better OpenVKL 0.9 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Items / Sec, More Is Better OpenVKL 0.9 Benchmark: vklBenchmark P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 140 280 420 560 700 Min: 667 / Avg: 669.67 / Max: 674 Min: 806 / Avg: 811.33 / Max: 814 Min: 704 / Avg: 718 / Max: 734 Min: 816 / Avg: 820.67 / Max: 824
Result
OpenBenchmarking.org Items / Sec, More Is Better OpenVKL 0.9 Benchmark: vklBenchmarkStructuredVolume P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 20M 40M 60M 80M 100M SE +/- 799905.44, N = 4 SE +/- 57856.16, N = 3 SE +/- 444572.22, N = 3 SE +/- 497134.18, N = 3 71974438 106056458 77997628 106704641 MIN: 1304623 / MAX: 672770484 MIN: 1396156 / MAX: 908416224 MIN: 1380198 / MAX: 693620172 MIN: 1343151 / MAX: 909146052 Items / Sec Per Watt
OpenBenchmarking.org Items / Sec Per Watt, More Is Better OpenVKL 0.9 Benchmark: vklBenchmarkStructuredVolume P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 90K 180K 270K 360K 450K 424481.55 438562.32 418290.31 436355.85 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2866 3407 P-State performance 3403 3403 3408 intel_cpufreq schedutil 1430 3076 3408 intel_cpufreq performance 3400 3405 3410 OpenBenchmarking.org Megahertz, More Is Better OpenVKL 0.9 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 70.7 169.6 254.7 P-State performance 150.8 241.8 326.7 intel_cpufreq schedutil 69.4 186.5 255.9 intel_cpufreq performance 171.6 244.5 315.4 OpenBenchmarking.org Watts, Fewer Is Better OpenVKL 0.9 CPU Power Consumption Monitor 80 160 240 320 400 1. P-State powersave: Approximate power consumption of 2204 Joules per run. 2. P-State performance: Approximate power consumption of 2983 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2424 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2934 Joules per run. CPU Temp
Min Avg Max P-State powersave 38.0 42.9 52.0 P-State performance 47.0 49.8 59.0 intel_cpufreq schedutil 41.0 44.9 55.0 intel_cpufreq performance 50.0 52.6 67.0 OpenBenchmarking.org Celsius, Fewer Is Better OpenVKL 0.9 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Items / Sec, More Is Better OpenVKL 0.9 Benchmark: vklBenchmarkStructuredVolume P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 20M 40M 60M 80M 100M Min: 69756114 / Avg: 71974437.75 / Max: 73272846 Min: 105941634 / Avg: 106056457.67 / Max: 106126265 Min: 77328702 / Avg: 77997628.33 / Max: 78839375 Min: 105733065 / Avg: 106704640.67 / Max: 107373343
Result
OpenBenchmarking.org Items / Sec, More Is Better OpenVKL 0.9 Benchmark: vklBenchmarkUnstructuredVolume P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 400K 800K 1200K 1600K 2000K SE +/- 19689.28, N = 3 SE +/- 8888.13, N = 3 SE +/- 4960.83, N = 3 SE +/- 4355.52, N = 3 1757269 1803840 1748017 1809667 MIN: 18102 / MAX: 5872318 MIN: 24013 / MAX: 5810552 MIN: 13879 / MAX: 5825120 MIN: 24362 / MAX: 5849692 Items / Sec Per Watt
OpenBenchmarking.org Items / Sec Per Watt, More Is Better OpenVKL 0.9 Benchmark: vklBenchmarkUnstructuredVolume P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1600 3200 4800 6400 8000 7267.58 6287.40 7409.47 6326.14 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 3262 3422 P-State performance 3403 3404 3424 intel_cpufreq schedutil 1433 3208 3416 intel_cpufreq performance 3403 3404 3432 OpenBenchmarking.org Megahertz, More Is Better OpenVKL 0.9 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 56.5 241.8 276.8 P-State performance 166.0 286.9 341.6 intel_cpufreq schedutil 61.0 235.9 281.7 intel_cpufreq performance 150.1 286.1 358.4 OpenBenchmarking.org Watts, Fewer Is Better OpenVKL 0.9 CPU Power Consumption Monitor 100 200 300 400 500 1. P-State powersave: Approximate power consumption of 51825 Joules per run. 2. P-State performance: Approximate power consumption of 56614 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 52924 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 56259 Joules per run. CPU Temp
Min Avg Max P-State powersave 39.0 50.0 56.0 P-State performance 46.0 53.9 60.0 intel_cpufreq schedutil 40.0 52.1 62.0 intel_cpufreq performance 48.0 55.7 62.0 OpenBenchmarking.org Celsius, Fewer Is Better OpenVKL 0.9 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Items / Sec, More Is Better OpenVKL 0.9 Benchmark: vklBenchmarkUnstructuredVolume P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 300K 600K 900K 1200K 1500K Min: 1733465 / Avg: 1757269.33 / Max: 1796338 Min: 1786789 / Avg: 1803840 / Max: 1816718 Min: 1738321 / Avg: 1748016.67 / Max: 1754688 Min: 1803971 / Avg: 1809667.33 / Max: 1818223
Result
OpenBenchmarking.org Items / Sec, More Is Better OpenVKL 0.9 Benchmark: vklBenchmarkVdbVolume P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 6M 12M 18M 24M 30M SE +/- 47580.68, N = 3 SE +/- 268896.33, N = 15 SE +/- 38895.86, N = 3 SE +/- 349026.46, N = 15 21722326 29996539 23599249 30016318 MIN: 1031810 / MAX: 151043112 MIN: 979600 / MAX: 247803840 MIN: 1060641 / MAX: 156949416 MIN: 979586 / MAX: 253907928 Items / Sec Per Watt
OpenBenchmarking.org Items / Sec Per Watt, More Is Better OpenVKL 0.9 Benchmark: vklBenchmarkVdbVolume P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 30K 60K 90K 120K 150K 123147.50 125723.06 118699.08 126322.91 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2993 3400 P-State performance 3402 3404 3411 intel_cpufreq schedutil 1702 3241 3405 intel_cpufreq performance 3400 3404 3410 OpenBenchmarking.org Megahertz, More Is Better OpenVKL 0.9 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 70.7 176.4 222.9 P-State performance 147.7 238.6 361.2 intel_cpufreq schedutil 71.3 198.8 249.7 intel_cpufreq performance 160.5 237.6 341.6 OpenBenchmarking.org Watts, Fewer Is Better OpenVKL 0.9 CPU Power Consumption Monitor 100 200 300 400 500 1. P-State powersave: Approximate power consumption of 2999 Joules per run. 2. P-State performance: Approximate power consumption of 4024 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 3446 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 3992 Joules per run. CPU Temp
Min Avg Max P-State powersave 43.0 47.6 55.0 P-State performance 46.0 50.5 62.0 intel_cpufreq schedutil 42.0 48.1 56.0 intel_cpufreq performance 45.0 46.9 58.0 OpenBenchmarking.org Celsius, Fewer Is Better OpenVKL 0.9 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Items / Sec, More Is Better OpenVKL 0.9 Benchmark: vklBenchmarkVdbVolume P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 5M 10M 15M 20M 25M Min: 21646886 / Avg: 21722325.67 / Max: 21810279 Min: 28927640 / Avg: 29996538.8 / Max: 32168515 Min: 23523744 / Avg: 23599249 / Max: 23653216 Min: 28845208 / Avg: 30016318.4 / Max: 33452620
Zstd Compression OpenBenchmarking.org MB/s Per Watt, More Is Better Zstd Compression 1.5.0 Compression Level: 19 - Decompression Speed P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 10.980 9.740 10.931 10.026
Result
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 8, Long Mode - Compression Speed intel_cpufreq performance P-State powersave P-State performance intel_cpufreq schedutil 200 400 600 800 1000 SE +/- 2.28, N = 3 SE +/- 0.62, N = 3 SE +/- 4.09, N = 3 SE +/- 2.96, N = 3 1015.8 296.0 1011.9 311.7 1. (CC) gcc options: -O3 -pthread -lz -llzma MB/s Per Watt
OpenBenchmarking.org MB/s Per Watt, More Is Better Zstd Compression 1.5.0 Compression Level: 8, Long Mode - Compression Speed intel_cpufreq performance 0.9358 1.8716 2.8074 3.7432 4.679 4.159 Result Confidence
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 8, Long Mode - Compression Speed intel_cpufreq performance P-State powersave P-State performance intel_cpufreq schedutil 200 400 600 800 1000 Min: 1011.2 / Avg: 1015.77 / Max: 1018.1 Min: 295.3 / Avg: 295.97 / Max: 297.2 Min: 1005.7 / Avg: 1011.87 / Max: 1019.6 Min: 308.1 / Avg: 311.73 / Max: 317.6 1. (CC) gcc options: -O3 -pthread -lz -llzma
Result
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 19 - Decompression Speed P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 600 1200 1800 2400 3000 SE +/- 16.96, N = 3 SE +/- 5.29, N = 15 SE +/- 4.60, N = 15 SE +/- 4.37, N = 3 2563.2 2561.8 2594.7 2609.7 1. (CC) gcc options: -O3 -pthread -lz -llzma MB/s Per Watt
OpenBenchmarking.org MB/s Per Watt, More Is Better Zstd Compression 1.5.0 Compression Level: 19 - Decompression Speed P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 10.980 9.740 10.931 10.026 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 3297 3408 P-State performance 3400 3404 3411 intel_cpufreq schedutil 1007 3311 3413 intel_cpufreq performance 3403 3404 3407 OpenBenchmarking.org Megahertz, More Is Better Zstd Compression 1.5.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 67.7 233.5 355.4 P-State performance 150.3 263.0 417.7 intel_cpufreq schedutil 67.8 237.4 377.1 intel_cpufreq performance 149.9 260.3 397.2 OpenBenchmarking.org Watts, Fewer Is Better Zstd Compression 1.5.0 CPU Power Consumption Monitor 110 220 330 440 550 1. P-State powersave: Approximate power consumption of 2801 Joules per run. 2. P-State performance: Approximate power consumption of 3051 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2943 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 3037 Joules per run. CPU Temp
Min Avg Max P-State powersave 33.0 42.5 55.0 P-State performance 46.0 48.1 59.0 intel_cpufreq schedutil 37.0 53.0 65.0 intel_cpufreq performance 46.0 47.3 53.0 OpenBenchmarking.org Celsius, Fewer Is Better Zstd Compression 1.5.0 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 19 - Decompression Speed P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 500 1000 1500 2000 2500 Min: 2537.4 / Avg: 2563.23 / Max: 2595.2 Min: 2511.3 / Avg: 2561.79 / Max: 2615.4 Min: 2542.8 / Avg: 2594.74 / Max: 2612.2 Min: 2601.6 / Avg: 2609.7 / Max: 2616.6 1. (CC) gcc options: -O3 -pthread -lz -llzma
Result
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 19, Long Mode - Decompression Speed P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 600 1200 1800 2400 3000 SE +/- 4.94, N = 15 SE +/- 9.19, N = 3 SE +/- 8.19, N = 6 SE +/- 2.06, N = 3 2711.6 2677.9 2693.4 2713.1 1. (CC) gcc options: -O3 -pthread -lz -llzma MB/s Per Watt
OpenBenchmarking.org MB/s Per Watt, More Is Better Zstd Compression 1.5.0 Compression Level: 19, Long Mode - Decompression Speed P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 12.48 11.10 12.20 11.31 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 3288 3409 P-State performance 3403 3404 3409 intel_cpufreq schedutil 1416 3332 3409 intel_cpufreq performance 3403 3404 3410 OpenBenchmarking.org Megahertz, More Is Better Zstd Compression 1.5.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 69.3 217.3 300.8 P-State performance 150.3 241.4 314.1 intel_cpufreq schedutil 56.0 220.9 303.6 intel_cpufreq performance 179.5 239.9 319.4 OpenBenchmarking.org Watts, Fewer Is Better Zstd Compression 1.5.0 CPU Power Consumption Monitor 80 160 240 320 400 1. P-State powersave: Approximate power consumption of 2477 Joules per run. 2. P-State performance: Approximate power consumption of 2655 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2503 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2639 Joules per run. CPU Temp
Min Avg Max P-State powersave 39.0 47.4 51.0 P-State performance 46.0 47.3 50.0 intel_cpufreq schedutil 44.0 49.9 53.0 intel_cpufreq performance 46.0 48.0 56.0 OpenBenchmarking.org Celsius, Fewer Is Better Zstd Compression 1.5.0 CPU Temperature Monitor 16 32 48 64 80 Result Confidence
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 19, Long Mode - Decompression Speed P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 500 1000 1500 2000 2500 Min: 2674.9 / Avg: 2711.6 / Max: 2729.2 Min: 2663.2 / Avg: 2677.9 / Max: 2694.8 Min: 2663.8 / Avg: 2693.35 / Max: 2715.6 Min: 2709.7 / Avg: 2713.07 / Max: 2716.8 1. (CC) gcc options: -O3 -pthread -lz -llzma
NAS Parallel Benchmarks Min Avg Max P-State powersave 802 3013 3300 P-State performance 3107 3268 3404 intel_cpufreq schedutil 1534 3159 3405 intel_cpufreq performance 3107 3283 3406 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 1876 3000 P-State performance 3110 3381 3410 intel_cpufreq schedutil 1971 2876 3404 intel_cpufreq performance 3400 3403 3409 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2591 3011 P-State performance 3001 3144 3403 intel_cpufreq schedutil 802 2945 3403 intel_cpufreq performance 3005 3158 3404 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2499 3300 P-State performance 3202 3359 3403 intel_cpufreq schedutil 2176 3223 3405 intel_cpufreq performance 2294 3298 3405 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 802 2884 3300 P-State performance 3105 3267 3403 intel_cpufreq schedutil 1872 3146 3404 intel_cpufreq performance 3104 3288 3403 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 802 2584 3301 P-State performance 3101 3386 3404 intel_cpufreq schedutil 2299 3186 3403 intel_cpufreq performance 3401 3403 3403 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 802 2989 3302 P-State performance 3102 3270 3404 intel_cpufreq schedutil 1801 3114 3403 intel_cpufreq performance 3103 3241 3403 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2761 3307 P-State performance 3202 3282 3403 intel_cpufreq schedutil 3102 3231 3405 intel_cpufreq performance 3105 3279 3403 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 1941 3057 P-State performance 3400 3402 3404 intel_cpufreq schedutil 2300 3085 3404 intel_cpufreq performance 3348 3399 3405 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2440 3304 P-State performance 3200 3371 3404 intel_cpufreq schedutil 2301 3185 3404 intel_cpufreq performance 3202 3342 3403 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
OpenFOAM Min Avg Max P-State powersave 802 2836 3400 P-State performance 3003 3227 3409 intel_cpufreq schedutil 1966 3139 3404 intel_cpufreq performance 3003 3231 3408 OpenBenchmarking.org Megahertz, More Is Better OpenFOAM 8 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 3053 3405 P-State performance 3002 3171 3413 intel_cpufreq schedutil 2209 3157 3405 intel_cpufreq performance 3002 3170 3409 OpenBenchmarking.org Megahertz, More Is Better OpenFOAM 8 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
WRF OpenBenchmarking.org Megahertz, More Is Better WRF 4.2.2 CPU Peak Freq (Highest CPU Core Frequency) Monitor P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 600 1200 1800 2400 3000 Min: 938 / Avg: 3009.34 / Max: 3537 Min: 2953 / Avg: 3010.99 / Max: 3403 Min: 2904 / Avg: 3009.83 / Max: 3400 Min: 2950 / Avg: 3010.05 / Max: 3403
Xcompact3d Incompact3d Min Avg Max P-State powersave 802 2697 3075 P-State performance 3004 3177 3403 intel_cpufreq schedutil 3003 3152 3405 intel_cpufreq performance 3002 3205 3405 OpenBenchmarking.org Megahertz, More Is Better Xcompact3d Incompact3d 2021-03-11 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2969 3313 P-State performance 3002 3019 3403 intel_cpufreq schedutil 2213 3014 3403 intel_cpufreq performance 3002 3019 3403 OpenBenchmarking.org Megahertz, More Is Better Xcompact3d Incompact3d 2021-03-11 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
ArrayFire Min Avg Max P-State powersave 801 2570 3405 P-State performance 3403 3404 3408 intel_cpufreq schedutil 2303 3284 3405 intel_cpufreq performance 3403 3404 3410 OpenBenchmarking.org Megahertz, More Is Better ArrayFire 3.7 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
LAMMPS Molecular Dynamics Simulator Min Avg Max P-State powersave 802 2981 3204 P-State performance 3000 3018 3403 intel_cpufreq schedutil 3002 3015 3403 intel_cpufreq performance 3000 3018 3403 OpenBenchmarking.org Megahertz, More Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
NAMD Min Avg Max P-State powersave 801 2883 3010 P-State performance 2946 3124 3403 intel_cpufreq schedutil 2942 3094 3404 intel_cpufreq performance 2922 3115 3403 OpenBenchmarking.org Megahertz, More Is Better NAMD 2.14 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
GROMACS Min Avg Max P-State powersave 802 2836 3400 P-State performance 2900 3141 3403 intel_cpufreq schedutil 2231 3064 3403 intel_cpufreq performance 2903 3164 3404 OpenBenchmarking.org Megahertz, More Is Better GROMACS 2021.2 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Quantum ESPRESSO OpenBenchmarking.org Megahertz, More Is Better Quantum ESPRESSO 6.7 CPU Peak Freq (Highest CPU Core Frequency) Monitor P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 600 1200 1800 2400 3000 Min: 801 / Avg: 3232.42 / Max: 3439 Min: 3103 / Avg: 3296.48 / Max: 3518 Min: 1730 / Avg: 3291.57 / Max: 3413 Min: 3104 / Avg: 3274.2 / Max: 3471
Darmstadt Automotive Parallel Heterogeneous Suite Min Avg Max P-State powersave 801 2875 3406 P-State performance 3403 3403 3406 intel_cpufreq schedutil 1391 2972 3405 intel_cpufreq performance 3403 3403 3404 OpenBenchmarking.org Megahertz, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2820 3406 P-State performance 3403 3404 3407 intel_cpufreq schedutil 801 2838 3406 intel_cpufreq performance 3403 3403 3405 OpenBenchmarking.org Megahertz, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 802 2620 3390 P-State performance 3400 3403 3410 intel_cpufreq schedutil 871 2612 3399 intel_cpufreq performance 3401 3403 3407 OpenBenchmarking.org Megahertz, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
RELION Min Avg Max P-State powersave 801 2965 3200 P-State performance 2941 3007 3403 intel_cpufreq schedutil 1501 2989 3403 intel_cpufreq performance 2908 3002 3403 OpenBenchmarking.org Megahertz, More Is Better RELION 3.1.1 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
TensorFlow Lite Min Avg Max P-State powersave 801 2956 3103 P-State performance 3000 3031 3403 intel_cpufreq schedutil 3000 3033 3403 intel_cpufreq performance 3000 3050 3404 OpenBenchmarking.org Megahertz, More Is Better TensorFlow Lite 2020-08-23 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2902 3050 P-State performance 3003 3059 3403 intel_cpufreq schedutil 2402 3038 3404 intel_cpufreq performance 3005 3068 3403 OpenBenchmarking.org Megahertz, More Is Better TensorFlow Lite 2020-08-23 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 802 3002 3210 P-State performance 3033 3134 3404 intel_cpufreq schedutil 3031 3138 3407 intel_cpufreq performance 3029 3143 3403 OpenBenchmarking.org Megahertz, More Is Better TensorFlow Lite 2020-08-23 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Rodinia Min Avg Max P-State powersave 801 2410 3406 P-State performance 3403 3403 3404 intel_cpufreq schedutil 2226 3333 3403 intel_cpufreq performance 3403 3403 3404 OpenBenchmarking.org Megahertz, More Is Better Rodinia 3.1 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2819 3406 P-State performance 2809 3011 3408 intel_cpufreq schedutil 1128 2880 3405 intel_cpufreq performance 2808 2997 3407 OpenBenchmarking.org Megahertz, More Is Better Rodinia 3.1 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2358 3401 P-State performance 3233 3403 3414 intel_cpufreq schedutil 1444 2685 3413 intel_cpufreq performance 3400 3404 3417 OpenBenchmarking.org Megahertz, More Is Better Rodinia 3.1 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2548 3405 P-State performance 3000 3286 3411 intel_cpufreq schedutil 1420 3152 3403 intel_cpufreq performance 3000 3261 3404 OpenBenchmarking.org Megahertz, More Is Better Rodinia 3.1 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 802 3277 3406 P-State performance 3000 3334 3408 intel_cpufreq schedutil 2303 3305 3406 intel_cpufreq performance 3000 3324 3410 OpenBenchmarking.org Megahertz, More Is Better Rodinia 3.1 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
libavif avifenc Min Avg Max P-State powersave 801 3210 3409 P-State performance 3403 3406 3442 intel_cpufreq schedutil 1421 3285 3409 intel_cpufreq performance 3403 3405 3410 OpenBenchmarking.org Megahertz, More Is Better libavif avifenc 0.9.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2498 3404 P-State performance 3403 3404 3410 intel_cpufreq schedutil 802 2852 3404 intel_cpufreq performance 3403 3405 3410 OpenBenchmarking.org Megahertz, More Is Better libavif avifenc 0.9.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2545 3404 P-State performance 3403 3403 3404 intel_cpufreq schedutil 802 3113 3405 intel_cpufreq performance 3403 3404 3408 OpenBenchmarking.org Megahertz, More Is Better libavif avifenc 0.9.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
PJSIP Min Avg Max P-State powersave 801 858 2300 P-State performance 3403 3405 3428 intel_cpufreq schedutil 802 1510 2300 intel_cpufreq performance 3400 3407 3483 OpenBenchmarking.org Megahertz, More Is Better PJSIP 2.11 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 3165 3408 P-State performance 3403 3407 3413 intel_cpufreq schedutil 908 3357 3409 intel_cpufreq performance 3403 3408 3435 OpenBenchmarking.org Megahertz, More Is Better PJSIP 2.11 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 963 3400 P-State performance 3400 3404 3409 intel_cpufreq schedutil 801 1738 3403 intel_cpufreq performance 3403 3405 3414 OpenBenchmarking.org Megahertz, More Is Better PJSIP 2.11 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Timed Godot Game Engine Compilation Min Avg Max P-State powersave 802 2935 3406 P-State performance 3100 3368 3410 intel_cpufreq schedutil 987 3246 3407 intel_cpufreq performance 3003 3374 3409 OpenBenchmarking.org Megahertz, More Is Better Timed Godot Game Engine Compilation 3.2.3 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Timed Linux Kernel Compilation Min Avg Max P-State powersave 801 2455 3405 P-State performance 2771 3290 3409 intel_cpufreq schedutil 2301 3260 3408 intel_cpufreq performance 2870 3269 3408 OpenBenchmarking.org Megahertz, More Is Better Timed Linux Kernel Compilation 5.10.20 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Timed Mesa Compilation Min Avg Max P-State powersave 801 2918 3404 P-State performance 2948 3346 3407 intel_cpufreq schedutil 802 3134 3404 intel_cpufreq performance 3041 3364 3406 OpenBenchmarking.org Megahertz, More Is Better Timed Mesa Compilation 21.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Timed LLVM Compilation Min Avg Max P-State powersave 801 2870 3404 P-State performance 2776 3108 3408 intel_cpufreq schedutil 1802 3071 3407 intel_cpufreq performance 2775 3097 3409 OpenBenchmarking.org Megahertz, More Is Better Timed LLVM Compilation 12.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 802 3095 3407 P-State performance 2843 3319 3409 intel_cpufreq schedutil 2302 3299 3410 intel_cpufreq performance 2196 3311 3413 OpenBenchmarking.org Megahertz, More Is Better Timed LLVM Compilation 12.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Timed Node.js Compilation Min Avg Max P-State powersave 801 2947 3408 P-State performance 2806 3150 3408 intel_cpufreq schedutil 2200 3131 3408 intel_cpufreq performance 2780 3147 3406 OpenBenchmarking.org Megahertz, More Is Better Timed Node.js Compilation 15.11 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Liquid-DSP Min Avg Max P-State powersave 801 2843 3300 P-State performance 3102 3214 3404 intel_cpufreq schedutil 1823 3216 3407 intel_cpufreq performance 3100 3242 3403 OpenBenchmarking.org Megahertz, More Is Better Liquid-DSP 2021.01.31 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 802 2873 3008 P-State performance 2990 3096 3405 intel_cpufreq schedutil 2880 3066 3403 intel_cpufreq performance 2183 3000 3408 OpenBenchmarking.org Megahertz, More Is Better Liquid-DSP 2021.01.31 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2836 3000 P-State performance 2958 3071 3403 intel_cpufreq schedutil 2959 3071 3404 intel_cpufreq performance 2956 3069 3403 OpenBenchmarking.org Megahertz, More Is Better Liquid-DSP 2021.01.31 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Blender Min Avg Max P-State powersave 801 2704 3408 P-State performance 2793 3140 3408 intel_cpufreq schedutil 2792 3107 3406 intel_cpufreq performance 2786 3127 3405 OpenBenchmarking.org Megahertz, More Is Better Blender 2.92 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2801 3400 P-State performance 2775 2930 3405 intel_cpufreq schedutil 2200 2856 3403 intel_cpufreq performance 2774 2918 3405 OpenBenchmarking.org Megahertz, More Is Better Blender 2.92 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 802 2936 3405 P-State performance 2904 3010 3405 intel_cpufreq schedutil 2301 2976 3406 intel_cpufreq performance 2892 3014 3408 OpenBenchmarking.org Megahertz, More Is Better Blender 2.92 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
PostgreSQL pgbench Min Avg Max P-State powersave 801 2759 3405 P-State performance 3000 3361 3413 intel_cpufreq schedutil 1412 3064 3405 intel_cpufreq performance 3000 3367 3431 OpenBenchmarking.org Megahertz, More Is Better PostgreSQL pgbench 13.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2821 3436 P-State performance 3000 3355 3413 intel_cpufreq schedutil 1652 3302 3405 intel_cpufreq performance 3000 3362 3441 OpenBenchmarking.org Megahertz, More Is Better PostgreSQL pgbench 13.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 2792 3405 P-State performance 3000 3360 3410 intel_cpufreq schedutil 1827 3302 3407 intel_cpufreq performance 2185 3350 3409 OpenBenchmarking.org Megahertz, More Is Better PostgreSQL pgbench 13.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Min Avg Max P-State powersave 801 3022 3437 P-State performance 3000 3366 3407 intel_cpufreq schedutil 958 3317 3442 intel_cpufreq performance 2190 3356 3436 OpenBenchmarking.org Megahertz, More Is Better PostgreSQL pgbench 13.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Stockfish Min Avg Max P-State powersave 801 2652 3002 P-State performance 2987 3106 3406 intel_cpufreq schedutil 2173 3060 3404 intel_cpufreq performance 2974 3105 3405 OpenBenchmarking.org Megahertz, More Is Better Stockfish 13 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
NWChem Min Avg Max P-State powersave 801 3007 3409 P-State performance 3003 3017 3409 intel_cpufreq schedutil 2300 3015 3409 intel_cpufreq performance 3003 3018 3409 OpenBenchmarking.org Megahertz, More Is Better NWChem 7.0.2 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000
Stockfish OpenBenchmarking.org Nodes Per Second Per Watt, More Is Better Stockfish 13 Total Time P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 80K 160K 240K 320K 400K 392264.24 378389.79 368173.69 380770.10
Result
OpenBenchmarking.org Nodes Per Second, More Is Better Stockfish 13 Total Time P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 40M 80M 120M 160M 200M SE +/- 1103002.46, N = 3 SE +/- 1592768.45, N = 3 SE +/- 1451422.61, N = 15 SE +/- 2150017.23, N = 3 177031122 177131428 175446182 178629474 1. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fprofile-use -fno-peel-loops -fno-tracer -pedantic -O3 -msse -msse3 -mpopcnt -mavx2 -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto=jobserver Nodes Per Second Per Watt
OpenBenchmarking.org Nodes Per Second Per Watt, More Is Better Stockfish 13 Total Time P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 80K 160K 240K 320K 400K 392264.24 378389.79 368173.69 380770.10 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2652 3002 P-State performance 2987 3106 3406 intel_cpufreq schedutil 2173 3060 3404 intel_cpufreq performance 2974 3105 3405 OpenBenchmarking.org Megahertz, More Is Better Stockfish 13 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 90.8 451.3 621.5 P-State performance 201.4 468.1 625.2 intel_cpufreq schedutil 88.3 476.5 620.1 intel_cpufreq performance 194.9 469.1 622.5 OpenBenchmarking.org Watts, Fewer Is Better Stockfish 13 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 2858 Joules per run. 2. P-State performance: Approximate power consumption of 2965 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 3145 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2971 Joules per run. CPU Temp
Min Avg Max P-State powersave 47.0 65.8 74.0 P-State performance 49.0 66.1 73.0 intel_cpufreq schedutil 47.0 70.5 76.0 intel_cpufreq performance 53.0 69.3 76.0 OpenBenchmarking.org Celsius, Fewer Is Better Stockfish 13 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Nodes Per Second, More Is Better Stockfish 13 Total Time P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 30M 60M 90M 120M 150M Min: 175295614 / Avg: 177031122 / Max: 179078220 Min: 174778062 / Avg: 177131428 / Max: 180167408 Min: 166205419 / Avg: 175446182.33 / Max: 189004933 Min: 174374932 / Avg: 178629474 / Max: 181297001 1. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fprofile-use -fno-peel-loops -fno-tracer -pedantic -O3 -msse -msse3 -mpopcnt -mavx2 -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto=jobserver
GROMACS OpenBenchmarking.org Ns Per Day Per Watt, More Is Better GROMACS 2021.2 Implementation: MPI CPU - Input: water_GMX50_bare P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0045 0.009 0.0135 0.018 0.0225 0.02 0.02 0.02 0.02
Result
OpenBenchmarking.org Ns Per Day, More Is Better GROMACS 2021.2 Implementation: MPI CPU - Input: water_GMX50_bare P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 SE +/- 0.038, N = 3 SE +/- 0.048, N = 3 SE +/- 0.035, N = 3 SE +/- 0.011, N = 3 8.999 9.052 9.042 9.102 1. (CXX) g++ options: -O3 -pthread Ns Per Day Per Watt
OpenBenchmarking.org Ns Per Day Per Watt, More Is Better GROMACS 2021.2 Implementation: MPI CPU - Input: water_GMX50_bare P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0045 0.009 0.0135 0.018 0.0225 0.02 0.02 0.02 0.02 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 802 2836 3400 P-State performance 2900 3141 3403 intel_cpufreq schedutil 2231 3064 3403 intel_cpufreq performance 2903 3164 3404 OpenBenchmarking.org Megahertz, More Is Better GROMACS 2021.2 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 63 445 722 P-State performance 71 461 575 intel_cpufreq schedutil 178 461 571 intel_cpufreq performance 155 465 570 OpenBenchmarking.org Watts, Fewer Is Better GROMACS 2021.2 CPU Power Consumption Monitor 200 400 600 800 1000 1. P-State powersave: Approximate power consumption of 4893 Joules per run. 2. P-State performance: Approximate power consumption of 5076 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 4914 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 5119 Joules per run. CPU Temp
Min Avg Max P-State powersave 48.0 63.1 74.0 P-State performance 52.0 66.6 76.0 intel_cpufreq schedutil 53.0 67.4 76.0 intel_cpufreq performance 52.0 62.8 71.0 OpenBenchmarking.org Celsius, Fewer Is Better GROMACS 2021.2 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Ns Per Day, More Is Better GROMACS 2021.2 Implementation: MPI CPU - Input: water_GMX50_bare P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 Min: 8.95 / Avg: 9 / Max: 9.08 Min: 8.97 / Avg: 9.05 / Max: 9.14 Min: 8.98 / Avg: 9.04 / Max: 9.1 Min: 9.09 / Avg: 9.1 / Max: 9.12 1. (CXX) g++ options: -O3 -pthread
LAMMPS Molecular Dynamics Simulator OpenBenchmarking.org ns/day Per Watt, More Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 Model: 20k Atoms P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0151 0.0302 0.0453 0.0604 0.0755 0.067 0.067 0.067 0.067
Result
OpenBenchmarking.org ns/day, More Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 Model: 20k Atoms P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 8 16 24 32 40 SE +/- 0.06, N = 3 SE +/- 0.05, N = 3 SE +/- 0.04, N = 3 SE +/- 0.05, N = 3 35.50 35.83 35.79 35.85 1. (CXX) g++ options: -O3 -pthread -lm ns/day Per Watt
OpenBenchmarking.org ns/day Per Watt, More Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 Model: 20k Atoms P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0151 0.0302 0.0453 0.0604 0.0755 0.067 0.067 0.067 0.067 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 802 2981 3204 P-State performance 3000 3018 3403 intel_cpufreq schedutil 3002 3015 3403 intel_cpufreq performance 3000 3018 3403 OpenBenchmarking.org Megahertz, More Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 124.1 531.5 568.5 P-State performance 191.2 535.5 569.6 intel_cpufreq schedutil 149.4 532.5 567.5 intel_cpufreq performance 189.0 534.9 577.7 OpenBenchmarking.org Watts, Fewer Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 31360 Joules per run. 2. P-State performance: Approximate power consumption of 31419 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 31065 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 31023 Joules per run. CPU Temp
Min Avg Max P-State powersave 48.0 71.2 75.0 P-State performance 49.0 71.0 76.0 intel_cpufreq schedutil 51.0 69.8 72.0 intel_cpufreq performance 50.0 72.0 74.0 OpenBenchmarking.org Celsius, Fewer Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org ns/day, More Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 Model: 20k Atoms P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 8 16 24 32 40 Min: 35.37 / Avg: 35.5 / Max: 35.58 Min: 35.74 / Avg: 35.82 / Max: 35.89 Min: 35.71 / Avg: 35.79 / Max: 35.84 Min: 35.8 / Avg: 35.85 / Max: 35.95 1. (CXX) g++ options: -O3 -pthread -lm
PJSIP OpenBenchmarking.org Responses Per Second Per Watt, More Is Better PJSIP 2.11 Method: OPTIONS, Stateful P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 10 20 30 40 50 38.12 45.87 38.23 45.12
Result
OpenBenchmarking.org Responses Per Second, More Is Better PJSIP 2.11 Method: OPTIONS, Stateful P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 2K 4K 6K 8K 10K SE +/- 31.56, N = 3 SE +/- 0.58, N = 3 SE +/- 13.93, N = 3 SE +/- 6.00, N = 3 3769 10362 3797 10409 1. (CC) gcc options: -lstdc++ -lssl -lcrypto -luuid -lm -lrt -lpthread -lasound -O2 Responses Per Second Per Watt
OpenBenchmarking.org Responses Per Second Per Watt, More Is Better PJSIP 2.11 Method: OPTIONS, Stateful P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 10 20 30 40 50 38.12 45.87 38.23 45.12 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 858 2300 P-State performance 3403 3405 3428 intel_cpufreq schedutil 802 1510 2300 intel_cpufreq performance 3400 3407 3483 OpenBenchmarking.org Megahertz, More Is Better PJSIP 2.11 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 62.5 98.9 107.7 P-State performance 185.9 225.9 244.3 intel_cpufreq schedutil 63.5 99.3 109.9 intel_cpufreq performance 167.0 230.7 260.3 OpenBenchmarking.org Watts, Fewer Is Better PJSIP 2.11 CPU Power Consumption Monitor 70 140 210 280 350 1. P-State powersave: Approximate power consumption of 1648 Joules per run. 2. P-State performance: Approximate power consumption of 3615 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1622 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 3614 Joules per run. CPU Temp
Min Avg Max P-State powersave 33.0 34.5 36.0 P-State performance 44.0 44.3 45.0 intel_cpufreq schedutil 35.0 36.5 38.0 intel_cpufreq performance 47.0 51.3 53.0 OpenBenchmarking.org Celsius, Fewer Is Better PJSIP 2.11 CPU Temperature Monitor 15 30 45 60 75 Result Confidence
OpenBenchmarking.org Responses Per Second, More Is Better PJSIP 2.11 Method: OPTIONS, Stateful P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 2K 4K 6K 8K 10K Min: 3712 / Avg: 3769 / Max: 3821 Min: 10361 / Avg: 10362 / Max: 10363 Min: 3777 / Avg: 3797.33 / Max: 3824 Min: 10397 / Avg: 10409 / Max: 10415 1. (CC) gcc options: -lstdc++ -lssl -lcrypto -luuid -lm -lrt -lpthread -lasound -O2
OpenBenchmarking.org Responses Per Second Per Watt, More Is Better PJSIP 2.11 Method: OPTIONS, Stateless P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 40 80 120 160 200 174.15 163.09 166.04 160.57
OpenBenchmarking.org Responses Per Second Per Watt, More Is Better PJSIP 2.11 Method: INVITE P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 6 12 18 24 30 24.58 23.27 23.62 23.05
Result
OpenBenchmarking.org Responses Per Second, More Is Better PJSIP 2.11 Method: OPTIONS, Stateless P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 9K 18K 27K 36K 45K SE +/- 341.07, N = 9 SE +/- 274.68, N = 3 SE +/- 380.67, N = 7 SE +/- 83.53, N = 3 41421 42389 41743 42338 1. (CC) gcc options: -lstdc++ -lssl -lcrypto -luuid -lm -lrt -lpthread -lasound -O2 Responses Per Second Per Watt
OpenBenchmarking.org Responses Per Second Per Watt, More Is Better PJSIP 2.11 Method: OPTIONS, Stateless P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 40 80 120 160 200 174.15 163.09 166.04 160.57 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 3165 3408 P-State performance 3403 3407 3413 intel_cpufreq schedutil 908 3357 3409 intel_cpufreq performance 3403 3408 3435 OpenBenchmarking.org Megahertz, More Is Better PJSIP 2.11 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 68.2 237.8 275.4 P-State performance 181.0 259.9 289.3 intel_cpufreq schedutil 67.5 251.4 286.7 intel_cpufreq performance 183.3 263.7 291.9 OpenBenchmarking.org Watts, Fewer Is Better PJSIP 2.11 CPU Power Consumption Monitor 70 140 210 280 350 1. P-State powersave: Approximate power consumption of 3066 Joules per run. 2. P-State performance: Approximate power consumption of 3292 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 3232 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 3340 Joules per run. CPU Temp
Min Avg Max P-State powersave 32.0 41.2 44.0 P-State performance 43.0 44.5 49.0 intel_cpufreq schedutil 34.0 54.7 64.0 intel_cpufreq performance 48.0 52.8 61.0 OpenBenchmarking.org Celsius, Fewer Is Better PJSIP 2.11 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Responses Per Second, More Is Better PJSIP 2.11 Method: OPTIONS, Stateless P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 7K 14K 21K 28K 35K Min: 39845 / Avg: 41421 / Max: 42798 Min: 42109 / Avg: 42388.67 / Max: 42938 Min: 39971 / Avg: 41743 / Max: 42953 Min: 42245 / Avg: 42338.33 / Max: 42505 1. (CC) gcc options: -lstdc++ -lssl -lcrypto -luuid -lm -lrt -lpthread -lasound -O2
Result
OpenBenchmarking.org Responses Per Second, More Is Better PJSIP 2.11 Method: INVITE P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1200 2400 3600 4800 6000 SE +/- 29.86, N = 15 SE +/- 7.17, N = 3 SE +/- 50.72, N = 15 SE +/- 15.21, N = 3 2548 5393 2540 5416 1. (CC) gcc options: -lstdc++ -lssl -lcrypto -luuid -lm -lrt -lpthread -lasound -O2 Responses Per Second Per Watt
OpenBenchmarking.org Responses Per Second Per Watt, More Is Better PJSIP 2.11 Method: INVITE P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 6 12 18 24 30 24.58 23.27 23.62 23.05 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 963 3400 P-State performance 3400 3404 3409 intel_cpufreq schedutil 801 1738 3403 intel_cpufreq performance 3403 3405 3414 OpenBenchmarking.org Megahertz, More Is Better PJSIP 2.11 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 59.1 103.7 116.6 P-State performance 166.8 231.8 281.1 intel_cpufreq schedutil 60.9 107.6 231.7 intel_cpufreq performance 184.2 235.0 256.4 OpenBenchmarking.org Watts, Fewer Is Better PJSIP 2.11 CPU Power Consumption Monitor 70 140 210 280 350 1. P-State powersave: Approximate power consumption of 1707 Joules per run. 2. P-State performance: Approximate power consumption of 3863 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1800 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 3917 Joules per run. CPU Temp
Min Avg Max P-State powersave 32.0 33.7 39.0 P-State performance 44.0 46.4 49.0 intel_cpufreq schedutil 32.0 36.6 55.0 intel_cpufreq performance 51.0 53.5 57.0 OpenBenchmarking.org Celsius, Fewer Is Better PJSIP 2.11 CPU Temperature Monitor 16 32 48 64 80 Result Confidence
OpenBenchmarking.org Responses Per Second, More Is Better PJSIP 2.11 Method: INVITE P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 900 1800 2700 3600 4500 Min: 2318 / Avg: 2547.53 / Max: 2633 Min: 5379 / Avg: 5393.33 / Max: 5401 Min: 2337 / Avg: 2539.6 / Max: 3151 Min: 5387 / Avg: 5416.33 / Max: 5438 1. (CC) gcc options: -lstdc++ -lssl -lcrypto -luuid -lm -lrt -lpthread -lasound -O2
Liquid-DSP OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 2021.01.31 Threads: 64 - Buffer Length: 256 - Filter Length: 57 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1.4M 2.8M 4.2M 5.6M 7M 6754065.39 6392760.37 6436493.63 6437714.60
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 2021.01.31 Threads: 128 - Buffer Length: 256 - Filter Length: 57 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1.5M 3M 4.5M 6M 7.5M 6911581.20 7053355.32 7058626.64 6985743.48
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 2021.01.31 Threads: 64 - Buffer Length: 256 - Filter Length: 57 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 700M 1400M 2100M 2800M 3500M SE +/- 19182283.49, N = 3 SE +/- 5500404.03, N = 3 SE +/- 69564540.69, N = 12 SE +/- 8774394.57, N = 3 3029900000 3066133333 3003125000 3072600000 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 2021.01.31 Threads: 64 - Buffer Length: 256 - Filter Length: 57 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1.4M 2.8M 4.2M 5.6M 7M 6754065.39 6392760.37 6436493.63 6437714.60 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2843 3300 P-State performance 3102 3214 3404 intel_cpufreq schedutil 1823 3216 3407 intel_cpufreq performance 3100 3242 3403 OpenBenchmarking.org Megahertz, More Is Better Liquid-DSP 2021.01.31 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 57.2 448.6 567.9 P-State performance 178.1 479.6 584.3 intel_cpufreq schedutil 75.6 466.6 591.2 intel_cpufreq performance 212.4 477.3 581.1 OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 2021.01.31 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 2692 Joules per run. 2. P-State performance: Approximate power consumption of 2718 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2644 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2705 Joules per run. CPU Temp
Min Avg Max P-State powersave 44.0 65.9 74.0 P-State performance 53.0 68.2 74.0 intel_cpufreq schedutil 45.0 65.9 76.0 intel_cpufreq performance 53.0 64.5 75.0 OpenBenchmarking.org Celsius, Fewer Is Better Liquid-DSP 2021.01.31 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 2021.01.31 Threads: 64 - Buffer Length: 256 - Filter Length: 57 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 500M 1000M 1500M 2000M 2500M Min: 2992300000 / Avg: 3029900000 / Max: 3055300000 Min: 3059900000 / Avg: 3066133333.33 / Max: 3077100000 Min: 2240000000 / Avg: 3003125000 / Max: 3106100000 Min: 3055300000 / Avg: 3072600000 / Max: 3083800000 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 2021.01.31 Threads: 128 - Buffer Length: 256 - Filter Length: 57 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 700M 1400M 2100M 2800M 3500M SE +/- 10910901.79, N = 3 SE +/- 6806124.69, N = 3 SE +/- 4817445.15, N = 3 SE +/- 5329477.77, N = 3 3270133333 3293000000 3291466667 3293400000 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 2021.01.31 Threads: 128 - Buffer Length: 256 - Filter Length: 57 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1.5M 3M 4.5M 6M 7.5M 6911581.20 7053355.32 7058626.64 6985743.48 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 802 2873 3008 P-State performance 2990 3096 3405 intel_cpufreq schedutil 2880 3066 3403 intel_cpufreq performance 2183 3000 3408 OpenBenchmarking.org Megahertz, More Is Better Liquid-DSP 2021.01.31 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 101.4 473.1 582.5 P-State performance 148.2 466.9 588.2 intel_cpufreq schedutil 157.2 466.3 582.9 intel_cpufreq performance 168.0 471.4 579.0 OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 2021.01.31 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 2681 Joules per run. 2. P-State performance: Approximate power consumption of 2490 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2642 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2672 Joules per run. CPU Temp
Min Avg Max P-State powersave 47.0 67.4 73.0 P-State performance 51.0 68.0 74.0 intel_cpufreq schedutil 49.0 67.9 74.0 intel_cpufreq performance 53.0 68.2 73.0 OpenBenchmarking.org Celsius, Fewer Is Better Liquid-DSP 2021.01.31 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 2021.01.31 Threads: 128 - Buffer Length: 256 - Filter Length: 57 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 600M 1200M 1800M 2400M 3000M Min: 3248700000 / Avg: 3270133333.33 / Max: 3284400000 Min: 3285700000 / Avg: 3293000000 / Max: 3306600000 Min: 3286500000 / Avg: 3291466666.67 / Max: 3301100000 Min: 3283300000 / Avg: 3293400000 / Max: 3301400000 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 2021.01.31 Threads: 160 - Buffer Length: 256 - Filter Length: 57 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 700M 1400M 2100M 2800M 3500M SE +/- 12922117.13, N = 3 SE +/- 2605762.84, N = 3 SE +/- 1071343.91, N = 3 SE +/- 5955109.85, N = 3 3073933333 3085100000 3082533333 3096700000 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 2021.01.31 Threads: 160 - Buffer Length: 256 - Filter Length: 57 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1.4M 2.8M 4.2M 5.6M 7M 6650836.15 6607333.41 6576582.79 6658045.41 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2836 3000 P-State performance 2958 3071 3403 intel_cpufreq schedutil 2959 3071 3404 intel_cpufreq performance 2956 3069 3403 OpenBenchmarking.org Megahertz, More Is Better Liquid-DSP 2021.01.31 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 157.2 462.2 625.0 P-State performance 196.5 466.9 607.3 intel_cpufreq schedutil 169.3 468.7 615.2 intel_cpufreq performance 196.4 465.1 607.4 OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 2021.01.31 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 2465 Joules per run. 2. P-State performance: Approximate power consumption of 2646 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2656 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2636 Joules per run. CPU Temp
Min Avg Max P-State powersave 52.0 69.7 75.0 P-State performance 55.0 70.5 74.0 intel_cpufreq schedutil 55.0 70.5 75.0 intel_cpufreq performance 55.0 69.9 75.0 OpenBenchmarking.org Celsius, Fewer Is Better Liquid-DSP 2021.01.31 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 2021.01.31 Threads: 160 - Buffer Length: 256 - Filter Length: 57 P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 500M 1000M 1500M 2000M 2500M Min: 3048100000 / Avg: 3073933333.33 / Max: 3087500000 Min: 3080400000 / Avg: 3085100000 / Max: 3089400000 Min: 3080600000 / Avg: 3082533333.33 / Max: 3084300000 Min: 3089400000 / Avg: 3096700000 / Max: 3108500000 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
Darmstadt Automotive Parallel Heterogeneous Suite DAPHNE is the Darmstadt Automotive Parallel HeterogeNEous Benchmark Suite with OpenCL / CUDA / OpenMP test cases for these automotive benchmarks for evaluating programming models in context to vehicle autonomous driving capabilities. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Test Cases Per Minute, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: Euclidean Cluster P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 200 400 600 800 1000 SE +/- 7.84, N = 15 SE +/- 2.39, N = 3 SE +/- 9.84, N = 15 SE +/- 1.25, N = 3 831.33 1020.77 884.75 1021.62 1. (CXX) g++ options: -O3 -std=c++11 -fopenmp Test Cases Per Minute Per Watt
OpenBenchmarking.org Test Cases Per Minute Per Watt, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: Euclidean Cluster P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1.0235 2.047 3.0705 4.094 5.1175 4.549 3.778 4.491 3.816 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2875 3406 P-State performance 3403 3403 3406 intel_cpufreq schedutil 1391 2972 3405 intel_cpufreq performance 3403 3403 3404 OpenBenchmarking.org Megahertz, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 62.2 182.8 261.7 P-State performance 211.6 270.2 317.9 intel_cpufreq schedutil 75.6 197.0 300.8 intel_cpufreq performance 209.8 267.7 311.3 OpenBenchmarking.org Watts, Fewer Is Better Darmstadt Automotive Parallel Heterogeneous Suite CPU Power Consumption Monitor 80 160 240 320 400 1. P-State powersave: Approximate power consumption of 1304 Joules per run. 2. P-State performance: Approximate power consumption of 1621 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1340 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1517 Joules per run. CPU Temp
Min Avg Max P-State powersave 37.0 41.4 47.0 P-State performance 47.0 49.3 52.0 intel_cpufreq schedutil 41.0 46.4 50.0 intel_cpufreq performance 47.0 49.9 52.0 OpenBenchmarking.org Celsius, Fewer Is Better Darmstadt Automotive Parallel Heterogeneous Suite CPU Temperature Monitor 15 30 45 60 75 Result Confidence
OpenBenchmarking.org Test Cases Per Minute, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: Euclidean Cluster P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 200 400 600 800 1000 Min: 797.28 / Avg: 831.33 / Max: 884.98 Min: 1017.19 / Avg: 1020.77 / Max: 1025.31 Min: 828.25 / Avg: 884.75 / Max: 952.64 Min: 1019.65 / Avg: 1021.62 / Max: 1023.94 1. (CXX) g++ options: -O3 -std=c++11 -fopenmp
OpenBenchmarking.org Test Cases Per Minute Per Watt, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: Points2Image P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 12 24 36 48 60 42.62 51.36 45.67 50.29
Result
OpenBenchmarking.org Test Cases Per Minute, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: NDT Mapping P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 200 400 600 800 1000 SE +/- 18.61, N = 15 SE +/- 4.79, N = 3 SE +/- 19.99, N = 14 SE +/- 4.97, N = 3 742.30 924.84 743.01 914.70 1. (CXX) g++ options: -O3 -std=c++11 -fopenmp Test Cases Per Minute Per Watt
OpenBenchmarking.org Test Cases Per Minute Per Watt, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: NDT Mapping P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.9999 1.9998 2.9997 3.9996 4.9995 4.444 3.948 4.322 3.913 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2820 3406 P-State performance 3403 3404 3407 intel_cpufreq schedutil 801 2838 3406 intel_cpufreq performance 3403 3403 3405 OpenBenchmarking.org Megahertz, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 54.6 167.0 233.4 P-State performance 149.0 234.2 277.2 intel_cpufreq schedutil 73.0 171.9 241.8 intel_cpufreq performance 148.3 233.7 287.4 OpenBenchmarking.org Watts, Fewer Is Better Darmstadt Automotive Parallel Heterogeneous Suite CPU Power Consumption Monitor 70 140 210 280 350 1. P-State powersave: Approximate power consumption of 1314 Joules per run. 2. P-State performance: Approximate power consumption of 1484 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1363 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1558 Joules per run. CPU Temp
Min Avg Max P-State powersave 35.0 38.6 41.0 P-State performance 46.0 48.2 49.0 intel_cpufreq schedutil 40.0 44.3 48.0 intel_cpufreq performance 47.0 49.6 50.0 OpenBenchmarking.org Celsius, Fewer Is Better Darmstadt Automotive Parallel Heterogeneous Suite CPU Temperature Monitor 14 28 42 56 70 Result Confidence
OpenBenchmarking.org Test Cases Per Minute, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: NDT Mapping P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 160 320 480 640 800 Min: 595.43 / Avg: 742.3 / Max: 826.32 Min: 915.28 / Avg: 924.84 / Max: 930.07 Min: 599.99 / Avg: 743.01 / Max: 867.36 Min: 907.75 / Avg: 914.7 / Max: 924.32 1. (CXX) g++ options: -O3 -std=c++11 -fopenmp
Result
OpenBenchmarking.org Test Cases Per Minute, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: Points2Image P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3K 6K 9K 12K 15K SE +/- 86.30, N = 3 SE +/- 114.27, N = 15 SE +/- 115.67, N = 12 SE +/- 100.42, N = 15 6634.58 13527.24 7077.76 13249.68 1. (CXX) g++ options: -O3 -std=c++11 -fopenmp Test Cases Per Minute Per Watt
OpenBenchmarking.org Test Cases Per Minute Per Watt, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: Points2Image P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 12 24 36 48 60 42.62 51.36 45.67 50.29 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 802 2620 3390 P-State performance 3400 3403 3410 intel_cpufreq schedutil 871 2612 3399 intel_cpufreq performance 3401 3403 3407 OpenBenchmarking.org Megahertz, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 80.0 155.7 242.9 P-State performance 160.0 263.4 291.5 intel_cpufreq schedutil 60.9 155.0 248.1 intel_cpufreq performance 154.1 263.4 299.3 OpenBenchmarking.org Watts, Fewer Is Better Darmstadt Automotive Parallel Heterogeneous Suite CPU Power Consumption Monitor 80 160 240 320 400 1. P-State powersave: Approximate power consumption of 4670 Joules per run. 2. P-State performance: Approximate power consumption of 5549 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 4559 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 5585 Joules per run. CPU Temp
Min Avg Max P-State powersave 35.0 37.5 42.0 P-State performance 46.0 51.9 54.0 intel_cpufreq schedutil 37.0 41.7 47.0 intel_cpufreq performance 48.0 54.1 63.0 OpenBenchmarking.org Celsius, Fewer Is Better Darmstadt Automotive Parallel Heterogeneous Suite CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Test Cases Per Minute, More Is Better Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: Points2Image P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 2K 4K 6K 8K 10K Min: 6484.21 / Avg: 6634.58 / Max: 6783.15 Min: 12950.1 / Avg: 13527.24 / Max: 14298.38 Min: 6389.05 / Avg: 7077.76 / Max: 7664.41 Min: 12570.34 / Avg: 13249.68 / Max: 13852.01 1. (CXX) g++ options: -O3 -std=c++11 -fopenmp
NAS Parallel Benchmarks OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 10 20 30 40 50 42.62 27.70 31.04 28.91
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.D P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 5 10 15 20 25 20.74 19.49 20.08 19.33
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: FT.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 60 120 180 240 300 296.44 285.10 280.10 293.75
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 100 200 300 400 500 478.36 440.34 467.72 445.54
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.B P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 100 200 300 400 500 458.43 369.71 441.37 372.37
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 50 100 150 200 250 220.93 203.56 223.98 201.24
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: IS.D P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 2 4 6 8 10 8.745 7.745 8.220 7.784
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: MG.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 150 300 450 600 750 680.15 412.87 540.22 427.24
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: CG.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 30 60 90 120 150 137.13 116.28 128.90 117.69
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: BT.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 40K 80K 120K 160K 200K SE +/- 416.58, N = 4 SE +/- 313.16, N = 4 SE +/- 322.38, N = 4 SE +/- 230.36, N = 4 191166.41 197815.31 191705.02 197821.88 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: BT.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 100 200 300 400 500 473.91 454.64 475.18 445.67 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 802 3013 3300 P-State performance 3107 3268 3404 intel_cpufreq schedutil 1534 3159 3405 intel_cpufreq performance 3107 3283 3406 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 62.1 403.4 562.2 P-State performance 142.1 435.1 592.6 intel_cpufreq schedutil 62.8 403.4 562.6 intel_cpufreq performance 137.8 443.9 612.4 OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 2017 Joules per run. 2. P-State performance: Approximate power consumption of 1958 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2017 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1997 Joules per run. CPU Temp
Min Avg Max P-State powersave 28.0 50.7 65.0 P-State performance 36.0 56.0 67.0 intel_cpufreq schedutil 28.0 52.3 66.0 intel_cpufreq performance 36.0 55.6 68.0 OpenBenchmarking.org Celsius, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: BT.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 30K 60K 90K 120K 150K Min: 190433.54 / Avg: 191166.41 / Max: 192039.54 Min: 196947.42 / Avg: 197815.31 / Max: 198332.24 Min: 191071.55 / Avg: 191705.02 / Max: 192503.07 Min: 197423.37 / Avg: 197821.88 / Max: 198486.89 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1700 3400 5100 6800 8500 SE +/- 109.82, N = 12 SE +/- 165.19, N = 15 SE +/- 121.94, N = 15 SE +/- 128.21, N = 15 6051.90 7610.18 6157.58 7705.44 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 10 20 30 40 50 42.62 27.70 31.04 28.91 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 1876 3000 P-State performance 3110 3381 3410 intel_cpufreq schedutil 1971 2876 3404 intel_cpufreq performance 3400 3403 3409 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 68.7 142.0 325.6 P-State performance 166.2 274.7 393.6 intel_cpufreq schedutil 103.0 198.4 382.4 intel_cpufreq performance 175.0 266.5 346.9 OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor 110 220 330 440 550 1. P-State powersave: Approximate power consumption of 284 Joules per run. 2. P-State performance: Approximate power consumption of 440 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 383 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 409 Joules per run. CPU Temp
Min Avg Max P-State powersave 39.0 45.0 53.0 P-State performance 53.0 56.0 67.0 intel_cpufreq schedutil 41.0 46.6 55.0 intel_cpufreq performance 52.0 55.0 67.0 OpenBenchmarking.org Celsius, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1300 2600 3900 5200 6500 Min: 5120.33 / Avg: 6051.9 / Max: 6487.12 Min: 6085.07 / Avg: 7610.18 / Max: 8332.98 Min: 5002.25 / Avg: 6157.58 / Max: 6669.01 Min: 6545.11 / Avg: 7705.44 / Max: 8291.96 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.D P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 2K 4K 6K 8K 10K SE +/- 65.72, N = 3 SE +/- 60.08, N = 13 SE +/- 22.19, N = 3 SE +/- 38.79, N = 4 8620.97 8920.10 8634.30 8938.09 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.D P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 5 10 15 20 25 20.74 19.49 20.08 19.33 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2591 3011 P-State performance 3001 3144 3403 intel_cpufreq schedutil 802 2945 3403 intel_cpufreq performance 3005 3158 3404 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 68.1 415.6 569.8 P-State performance 165.5 457.8 585.0 intel_cpufreq schedutil 72.0 429.9 565.8 intel_cpufreq performance 184.4 462.4 576.1 OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 2217 Joules per run. 2. P-State performance: Approximate power consumption of 2289 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2150 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2312 Joules per run. CPU Temp
Min Avg Max P-State powersave 37.0 55.3 67.0 P-State performance 51.0 69.2 76.0 intel_cpufreq schedutil 38.0 57.9 68.0 intel_cpufreq performance 50.0 66.5 74.0 OpenBenchmarking.org Celsius, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.D P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1600 3200 4800 6400 8000 Min: 8490.11 / Avg: 8620.97 / Max: 8696.99 Min: 8360.09 / Avg: 8920.1 / Max: 9191.16 Min: 8595.29 / Avg: 8634.3 / Max: 8672.13 Min: 8880.61 / Avg: 8938.09 / Max: 9049.68 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: FT.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 20K 40K 60K 80K 100K SE +/- 842.11, N = 7 SE +/- 83.85, N = 7 SE +/- 807.72, N = 7 SE +/- 63.65, N = 7 90989.73 100543.72 91970.99 100763.11 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: FT.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 60 120 180 240 300 296.44 285.10 280.10 293.75 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2499 3300 P-State performance 3202 3359 3403 intel_cpufreq schedutil 2176 3223 3405 intel_cpufreq performance 2294 3298 3405 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 76.0 306.9 553.3 P-State performance 172.6 352.7 574.1 intel_cpufreq schedutil 116.2 328.4 543.4 intel_cpufreq performance 177.4 343.0 548.2 OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 789 Joules per run. 2. P-State performance: Approximate power consumption of 806 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 844 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 735 Joules per run. CPU Temp
Min Avg Max P-State powersave 48.0 55.3 63.0 P-State performance 57.0 62.9 70.0 intel_cpufreq schedutil 49.0 56.5 67.0 intel_cpufreq performance 55.0 61.2 68.0 OpenBenchmarking.org Celsius, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: FT.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 20K 40K 60K 80K 100K Min: 88152.28 / Avg: 90989.73 / Max: 95161.66 Min: 100220.81 / Avg: 100543.72 / Max: 100841.36 Min: 89431.45 / Avg: 91970.99 / Max: 94436.99 Min: 100515.52 / Avg: 100763.11 / Max: 100957.17 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 40K 80K 120K 160K 200K SE +/- 1423.68, N = 4 SE +/- 364.13, N = 4 SE +/- 196.76, N = 4 SE +/- 495.98, N = 4 174911.02 187300.47 177878.68 186973.06 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 100 200 300 400 500 478.36 440.34 467.72 445.54 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 802 2884 3300 P-State performance 3105 3267 3403 intel_cpufreq schedutil 1872 3146 3404 intel_cpufreq performance 3104 3288 3403 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 84.6 365.6 554.0 P-State performance 190.2 425.4 593.0 intel_cpufreq schedutil 134.8 380.3 554.5 intel_cpufreq performance 195.4 419.7 579.6 OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 1463 Joules per run. 2. P-State performance: Approximate power consumption of 1595 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1521 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1574 Joules per run. CPU Temp
Min Avg Max P-State powersave 47.0 58.8 69.0 P-State performance 55.0 66.6 73.0 intel_cpufreq schedutil 47.0 60.7 68.0 intel_cpufreq performance 54.0 65.2 72.0 OpenBenchmarking.org Celsius, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 30K 60K 90K 120K 150K Min: 171769.36 / Avg: 174911.02 / Max: 178187.71 Min: 186706.87 / Avg: 187300.47 / Max: 188292.05 Min: 177474.51 / Avg: 177878.68 / Max: 178398.74 Min: 185984.7 / Avg: 186973.06 / Max: 188227.21 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.B P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 30K 60K 90K 120K 150K SE +/- 480.22, N = 8 SE +/- 134.33, N = 9 SE +/- 203.66, N = 8 SE +/- 133.35, N = 9 114946.46 123544.09 115501.41 123176.72 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.B P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 100 200 300 400 500 458.43 369.71 441.37 372.37 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 802 2584 3301 P-State performance 3101 3386 3404 intel_cpufreq schedutil 2299 3186 3403 intel_cpufreq performance 3401 3403 3403 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 104.7 250.7 444.1 P-State performance 151.7 334.2 435.1 intel_cpufreq schedutil 103.8 261.7 454.3 intel_cpufreq performance 196.0 330.8 404.0 OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor 120 240 360 480 600 1. P-State powersave: Approximate power consumption of 501 Joules per run. 2. P-State performance: Approximate power consumption of 668 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 523 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 662 Joules per run. CPU Temp
Min Avg Max P-State powersave 49.0 52.2 59.0 P-State performance 55.0 62.2 70.0 intel_cpufreq schedutil 50.0 52.7 55.0 intel_cpufreq performance 55.0 61.8 69.0 OpenBenchmarking.org Celsius, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.B P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 20K 40K 60K 80K 100K Min: 112909.96 / Avg: 114946.46 / Max: 117270.82 Min: 123048.82 / Avg: 123544.09 / Max: 124187.87 Min: 114547.2 / Avg: 115501.41 / Max: 116322.05 Min: 122615.53 / Avg: 123176.72 / Max: 123909.89 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 20K 40K 60K 80K 100K SE +/- 139.83, N = 3 SE +/- 159.72, N = 4 SE +/- 142.93, N = 3 SE +/- 55.17, N = 4 91313.60 91817.76 91642.33 91830.56 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 50 100 150 200 250 220.93 203.56 223.98 201.24 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 802 2989 3302 P-State performance 3102 3270 3404 intel_cpufreq schedutil 1801 3114 3403 intel_cpufreq performance 3103 3241 3403 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 95.4 413.3 571.9 P-State performance 193.4 451.1 596.8 intel_cpufreq schedutil 94.7 409.2 564.8 intel_cpufreq performance 191.4 456.3 597.0 OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor 160 320 480 640 800 1. P-State powersave: Approximate power consumption of 2067 Joules per run. 2. P-State performance: Approximate power consumption of 2255 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2182 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2282 Joules per run. CPU Temp
Min Avg Max P-State powersave 46.0 60.0 68.0 P-State performance 55.0 66.7 73.0 intel_cpufreq schedutil 45.0 60.6 69.0 intel_cpufreq performance 54.0 67.3 72.0 OpenBenchmarking.org Celsius, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 16K 32K 48K 64K 80K Min: 91068.82 / Avg: 91313.6 / Max: 91553.12 Min: 91403.58 / Avg: 91817.76 / Max: 92172.43 Min: 91416.18 / Avg: 91642.33 / Max: 91906.82 Min: 91677.56 / Avg: 91830.56 / Max: 91920.79 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: IS.D P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 600 1200 1800 2400 3000 SE +/- 29.00, N = 4 SE +/- 15.07, N = 4 SE +/- 32.51, N = 4 SE +/- 5.86, N = 4 2999.72 3014.77 2926.26 2978.52 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: IS.D P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 2 4 6 8 10 8.745 7.745 8.220 7.784 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2761 3307 P-State performance 3202 3282 3403 intel_cpufreq schedutil 3102 3231 3405 intel_cpufreq performance 3105 3279 3403 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 89.1 343.0 549.9 P-State performance 189.4 389.2 549.9 intel_cpufreq schedutil 121.9 356.0 540.0 intel_cpufreq performance 191.3 382.7 557.9 OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 1715 Joules per run. 2. P-State performance: Approximate power consumption of 1557 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 1691 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 1531 Joules per run. CPU Temp
Min Avg Max P-State powersave 50.0 62.3 73.0 P-State performance 57.0 61.6 68.0 intel_cpufreq schedutil 50.0 61.6 69.0 intel_cpufreq performance 59.0 66.5 74.0 OpenBenchmarking.org Celsius, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: IS.D P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 500 1000 1500 2000 2500 Min: 2943.44 / Avg: 2999.72 / Max: 3073.84 Min: 2969.56 / Avg: 3014.77 / Max: 3030.39 Min: 2854.37 / Avg: 2926.26 / Max: 3004.56 Min: 2967.29 / Avg: 2978.52 / Max: 2988.91 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: MG.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 30K 60K 90K 120K 150K SE +/- 333.81, N = 10 SE +/- 89.79, N = 10 SE +/- 233.06, N = 10 SE +/- 209.39, N = 10 116754.16 120084.27 117869.49 120124.05 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: MG.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 150 300 450 600 750 680.15 412.87 540.22 427.24 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 1941 3057 P-State performance 3400 3402 3404 intel_cpufreq schedutil 2300 3085 3404 intel_cpufreq performance 3348 3399 3405 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 95.5 171.7 333.1 P-State performance 151.1 290.9 431.5 intel_cpufreq schedutil 78.8 218.2 471.0 intel_cpufreq performance 188.3 281.2 385.2 OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor 120 240 360 480 600 1. P-State powersave: Approximate power consumption of 343 Joules per run. 2. P-State performance: Approximate power consumption of 465 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 393 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 450 Joules per run. CPU Temp
Min Avg Max P-State powersave 44.0 50.1 59.0 P-State performance 50.0 52.9 60.0 intel_cpufreq schedutil 46.0 51.7 61.0 intel_cpufreq performance 54.0 57.4 64.0 OpenBenchmarking.org Celsius, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: MG.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 20K 40K 60K 80K 100K Min: 114463.06 / Avg: 116754.16 / Max: 117996.91 Min: 119574.58 / Avg: 120084.27 / Max: 120655.3 Min: 116191.23 / Avg: 117869.49 / Max: 118738.5 Min: 119055.9 / Avg: 120124.05 / Max: 121232.86 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: CG.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 9K 18K 27K 36K 45K SE +/- 83.38, N = 7 SE +/- 113.89, N = 8 SE +/- 42.28, N = 8 SE +/- 69.22, N = 8 39485.72 40261.26 39430.66 40226.02 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: CG.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 30 60 90 120 150 137.13 116.28 128.90 117.69 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2440 3304 P-State performance 3200 3371 3404 intel_cpufreq schedutil 2301 3185 3404 intel_cpufreq performance 3202 3342 3403 OpenBenchmarking.org Megahertz, More Is Better NAS Parallel Benchmarks 3.4 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 56.5 287.9 551.8 P-State performance 178.9 346.2 549.4 intel_cpufreq schedutil 70.7 305.9 510.9 intel_cpufreq performance 182.4 341.8 554.4 OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor 140 280 420 560 700 1. P-State powersave: Approximate power consumption of 699 Joules per run. 2. P-State performance: Approximate power consumption of 736 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 688 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 684 Joules per run. CPU Temp
Min Avg Max P-State powersave 41.0 49.6 61.0 P-State performance 48.0 57.8 67.0 intel_cpufreq schedutil 43.0 51.9 62.0 intel_cpufreq performance 52.0 57.1 64.0 OpenBenchmarking.org Celsius, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: CG.C P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 7K 14K 21K 28K 35K Min: 39158.68 / Avg: 39485.72 / Max: 39729.32 Min: 39894.3 / Avg: 40261.26 / Max: 40800.55 Min: 39293.73 / Avg: 39430.66 / Max: 39639.4 Min: 39928.96 / Avg: 40226.02 / Max: 40474.55 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent_core -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.1.0
OpenFOAM Min Avg Max P-State powersave 43.0 53.2 67.0 P-State performance 54.0 64.3 74.0 intel_cpufreq schedutil 46.0 59.5 71.0 intel_cpufreq performance 52.0 61.8 71.0 OpenBenchmarking.org Celsius, Fewer Is Better OpenFOAM 8 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 44.0 60.8 75.0 P-State performance 46.0 62.4 74.0 intel_cpufreq schedutil 45.0 62.5 75.0 intel_cpufreq performance 46.0 62.9 76.0 OpenBenchmarking.org Celsius, Fewer Is Better OpenFOAM 8 CPU Temperature Monitor 20 40 60 80 100
WRF OpenBenchmarking.org Celsius, Fewer Is Better WRF 4.2.2 CPU Temperature Monitor P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 14 28 42 56 70 Min: 52 / Avg: 68.72 / Max: 74 Min: 55 / Avg: 67.24 / Max: 74 Min: 55 / Avg: 70.13 / Max: 74 Min: 55 / Avg: 69.28 / Max: 74
Xcompact3d Incompact3d Min Avg Max P-State powersave 46.0 56.9 66.0 P-State performance 51.0 60.4 66.0 intel_cpufreq schedutil 51.0 58.4 65.0 intel_cpufreq performance 52.0 60.9 65.0 OpenBenchmarking.org Celsius, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 49.0 65.8 74.0 P-State performance 50.0 68.2 73.0 intel_cpufreq schedutil 48.0 67.1 74.0 intel_cpufreq performance 50.0 68.2 73.0 OpenBenchmarking.org Celsius, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 CPU Temperature Monitor 20 40 60 80 100
ArrayFire Min Avg Max P-State powersave 38.0 41.6 44.0 P-State performance 46.0 47.2 54.0 intel_cpufreq schedutil 42.0 44.8 48.0 intel_cpufreq performance 46.0 46.9 49.0 OpenBenchmarking.org Celsius, Fewer Is Better ArrayFire 3.7 CPU Temperature Monitor 15 30 45 60 75
NAMD Min Avg Max P-State powersave 51.0 66.9 70.0 P-State performance 54.0 65.3 69.0 intel_cpufreq schedutil 51.0 63.4 69.0 intel_cpufreq performance 54.0 65.4 70.0 OpenBenchmarking.org Celsius, Fewer Is Better NAMD 2.14 CPU Temperature Monitor 20 40 60 80 100
Quantum ESPRESSO OpenBenchmarking.org Celsius, Fewer Is Better Quantum ESPRESSO 6.7 CPU Temperature Monitor P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 15 30 45 60 75 Min: 43 / Avg: 64.58 / Max: 79 Min: 48 / Avg: 64.61 / Max: 75 Min: 46 / Avg: 65.52 / Max: 77 Min: 47 / Avg: 66.29 / Max: 78
RELION Min Avg Max P-State powersave 32.0 67.9 77.0 P-State performance 49.0 68.0 76.0 intel_cpufreq schedutil 37.0 67.3 76.0 intel_cpufreq performance 50.0 67.6 76.0 OpenBenchmarking.org Celsius, Fewer Is Better RELION 3.1.1 CPU Temperature Monitor 20 40 60 80 100
TensorFlow Lite Min Avg Max P-State powersave 50.0 68.3 72.0 P-State performance 55.0 69.1 73.0 intel_cpufreq schedutil 51.0 67.4 71.0 intel_cpufreq performance 50.0 68.8 73.0 OpenBenchmarking.org Celsius, Fewer Is Better TensorFlow Lite 2020-08-23 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 50.0 72.0 77.0 P-State performance 54.0 71.7 76.0 intel_cpufreq schedutil 49.0 69.5 73.0 intel_cpufreq performance 52.0 71.8 76.0 OpenBenchmarking.org Celsius, Fewer Is Better TensorFlow Lite 2020-08-23 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 55.0 71.1 74.0 P-State performance 56.0 72.1 75.0 intel_cpufreq schedutil 53.0 70.2 74.0 intel_cpufreq performance 56.0 73.2 76.0 OpenBenchmarking.org Celsius, Fewer Is Better TensorFlow Lite 2020-08-23 CPU Temperature Monitor 20 40 60 80 100
Rodinia Min Avg Max P-State powersave 48.0 53.0 63.0 P-State performance 54.0 59.6 68.0 intel_cpufreq schedutil 49.0 53.5 63.0 intel_cpufreq performance 52.0 55.9 65.0 OpenBenchmarking.org Celsius, Fewer Is Better Rodinia 3.1 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 46.0 66.6 75.0 P-State performance 52.0 69.1 76.0 intel_cpufreq schedutil 46.0 68.1 77.0 intel_cpufreq performance 50.0 69.0 77.0 OpenBenchmarking.org Celsius, Fewer Is Better Rodinia 3.1 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 43.0 50.5 59.0 P-State performance 57.0 71.6 78.0 intel_cpufreq schedutil 46.0 55.9 66.0 intel_cpufreq performance 58.0 70.7 77.0 OpenBenchmarking.org Celsius, Fewer Is Better Rodinia 3.1 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 39.0 54.8 63.0 P-State performance 52.0 57.6 64.0 intel_cpufreq schedutil 46.0 58.0 65.0 intel_cpufreq performance 54.0 59.2 64.0 OpenBenchmarking.org Celsius, Fewer Is Better Rodinia 3.1 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 47.0 54.6 65.0 P-State performance 47.0 52.4 62.0 intel_cpufreq schedutil 47.0 56.0 67.0 intel_cpufreq performance 45.0 51.1 62.0 OpenBenchmarking.org Celsius, Fewer Is Better Rodinia 3.1 CPU Temperature Monitor 20 40 60 80 100
libavif avifenc Min Avg Max P-State powersave 35.0 54.1 71.0 P-State performance 51.0 53.8 58.0 intel_cpufreq schedutil 35.0 55.1 69.0 intel_cpufreq performance 55.0 61.9 73.0 OpenBenchmarking.org Celsius, Fewer Is Better libavif avifenc 0.9.0 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 38.0 41.7 46.0 P-State performance 46.0 47.7 49.0 intel_cpufreq schedutil 40.0 45.4 51.0 intel_cpufreq performance 52.0 54.1 56.0 OpenBenchmarking.org Celsius, Fewer Is Better libavif avifenc 0.9.0 CPU Temperature Monitor 16 32 48 64 80
Min Avg Max P-State powersave 36.0 39.1 50.0 P-State performance 44.0 45.3 46.0 intel_cpufreq schedutil 38.0 40.7 43.0 intel_cpufreq performance 47.0 49.5 52.0 OpenBenchmarking.org Celsius, Fewer Is Better libavif avifenc 0.9.0 CPU Temperature Monitor 15 30 45 60 75
Timed Godot Game Engine Compilation Min Avg Max P-State powersave 33.0 44.3 63.0 P-State performance 44.0 54.9 67.0 intel_cpufreq schedutil 32.0 53.6 68.0 intel_cpufreq performance 49.0 57.0 70.0 OpenBenchmarking.org Celsius, Fewer Is Better Timed Godot Game Engine Compilation 3.2.3 CPU Temperature Monitor 20 40 60 80 100
Timed Linux Kernel Compilation Min Avg Max P-State powersave 40.0 51.9 66.0 P-State performance 50.0 57.6 68.0 intel_cpufreq schedutil 45.0 57.9 70.0 intel_cpufreq performance 51.0 59.2 69.0 OpenBenchmarking.org Celsius, Fewer Is Better Timed Linux Kernel Compilation 5.10.20 CPU Temperature Monitor 20 40 60 80 100
Timed Mesa Compilation Min Avg Max P-State powersave 43.0 52.0 63.0 P-State performance 52.0 56.2 68.0 intel_cpufreq schedutil 46.0 54.3 65.0 intel_cpufreq performance 52.0 57.3 67.0 OpenBenchmarking.org Celsius, Fewer Is Better Timed Mesa Compilation 21.0 CPU Temperature Monitor 20 40 60 80 100
Timed LLVM Compilation Min Avg Max P-State powersave 42.0 62.0 77.0 P-State performance 50.0 66.0 80.0 intel_cpufreq schedutil 44.0 63.6 76.0 intel_cpufreq performance 51.0 66.9 78.0 OpenBenchmarking.org Celsius, Fewer Is Better Timed LLVM Compilation 12.0 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 40.0 60.3 76.0 P-State performance 49.0 64.6 76.0 intel_cpufreq schedutil 42.0 62.9 75.0 intel_cpufreq performance 50.0 64.4 76.0 OpenBenchmarking.org Celsius, Fewer Is Better Timed LLVM Compilation 12.0 CPU Temperature Monitor 20 40 60 80 100
Timed Node.js Compilation Min Avg Max P-State powersave 44.0 61.0 75.0 P-State performance 52.0 66.2 76.0 intel_cpufreq schedutil 49.0 64.7 75.0 intel_cpufreq performance 54.0 65.5 75.0 OpenBenchmarking.org Celsius, Fewer Is Better Timed Node.js Compilation 15.11 CPU Temperature Monitor 20 40 60 80 100
Blender Min Avg Max P-State powersave 53.0 67.6 75.0 P-State performance 56.0 68.6 74.0 intel_cpufreq schedutil 56.0 68.4 74.0 intel_cpufreq performance 56.0 69.4 75.0 OpenBenchmarking.org Celsius, Fewer Is Better Blender 2.92 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 50.0 71.8 76.0 P-State performance 57.0 73.0 77.0 intel_cpufreq schedutil 52.0 72.5 78.0 intel_cpufreq performance 57.0 73.4 78.0 OpenBenchmarking.org Celsius, Fewer Is Better Blender 2.92 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 49.0 69.8 75.0 P-State performance 54.0 71.9 76.0 intel_cpufreq schedutil 52.0 70.2 76.0 intel_cpufreq performance 55.0 71.0 75.0 OpenBenchmarking.org Celsius, Fewer Is Better Blender 2.92 CPU Temperature Monitor 20 40 60 80 100
PostgreSQL pgbench Min Avg Max P-State powersave 41.0 46.6 52.0 P-State performance 44.0 49.0 57.0 intel_cpufreq schedutil 39.0 44.7 50.0 intel_cpufreq performance 48.0 51.3 58.0 OpenBenchmarking.org Celsius, Fewer Is Better PostgreSQL pgbench 13.0 CPU Temperature Monitor 16 32 48 64 80
Min Avg Max P-State powersave 39.0 45.3 55.0 P-State performance 45.0 48.6 63.0 intel_cpufreq schedutil 37.0 47.1 62.0 intel_cpufreq performance 50.0 54.0 67.0 OpenBenchmarking.org Celsius, Fewer Is Better PostgreSQL pgbench 13.0 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 39.0 43.2 47.0 P-State performance 45.0 47.3 56.0 intel_cpufreq schedutil 42.0 47.3 57.0 intel_cpufreq performance 50.0 52.8 62.0 OpenBenchmarking.org Celsius, Fewer Is Better PostgreSQL pgbench 13.0 CPU Temperature Monitor 20 40 60 80 100
Min Avg Max P-State powersave 39.0 50.8 66.0 P-State performance 45.0 48.5 62.0 intel_cpufreq schedutil 41.0 51.3 65.0 intel_cpufreq performance 49.0 55.2 67.0 OpenBenchmarking.org Celsius, Fewer Is Better PostgreSQL pgbench 13.0 CPU Temperature Monitor 20 40 60 80 100
NWChem Min Avg Max P-State powersave 52.0 72.4 78.0 P-State performance 55.0 70.5 79.0 intel_cpufreq schedutil 53.0 70.5 76.0 intel_cpufreq performance 58.0 70.7 77.0 OpenBenchmarking.org Celsius, Fewer Is Better NWChem 7.0.2 CPU Temperature Monitor 20 40 60 80 100
NAMD NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org days/ns, Fewer Is Better NAMD 2.14 ATPase Simulation - 327,506 Atoms P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 0.0613 0.1226 0.1839 0.2452 0.3065 SE +/- 0.00040, N = 3 SE +/- 0.00055, N = 3 SE +/- 0.00043, N = 3 SE +/- 0.00021, N = 3 0.27093 0.27168 0.27227 0.27051 CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2883 3010 P-State performance 2946 3124 3403 intel_cpufreq schedutil 2942 3094 3404 intel_cpufreq performance 2922 3115 3403 OpenBenchmarking.org Megahertz, More Is Better NAMD 2.14 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 129 477 647 P-State performance 189 456 636 intel_cpufreq schedutil 145 429 609 intel_cpufreq performance 162 459 635 OpenBenchmarking.org Watts, Fewer Is Better NAMD 2.14 CPU Power Consumption Monitor 200 400 600 800 1000 1. P-State powersave: Approximate power consumption of 3813 Joules per run. 2. P-State performance: Approximate power consumption of 2280 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2288 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2449 Joules per run. CPU Temp
Min Avg Max P-State powersave 51.0 66.9 70.0 P-State performance 54.0 65.3 69.0 intel_cpufreq schedutil 51.0 63.4 69.0 intel_cpufreq performance 54.0 65.4 70.0 OpenBenchmarking.org Celsius, Fewer Is Better NAMD 2.14 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org days/ns, Fewer Is Better NAMD 2.14 ATPase Simulation - 327,506 Atoms P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 1 2 3 4 5 Min: 0.27 / Avg: 0.27 / Max: 0.27 Min: 0.27 / Avg: 0.27 / Max: 0.27 Min: 0.27 / Avg: 0.27 / Max: 0.27 Min: 0.27 / Avg: 0.27 / Max: 0.27
WRF WRF, the Weather Research and Forecasting Model, is a "next-generation mesoscale numerical weather prediction system designed for both atmospheric research and operational forecasting applications. It features two dynamical cores, a data assimilation system, and a software architecture supporting parallel computation and system extensibility." Learn more via the OpenBenchmarking.org test page.
Xcompact3d Incompact3d Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.
Quantum ESPRESSO Quantum ESPRESSO is an integrated suite of Open-Source computer codes for electronic-structure calculations and materials modeling at the nanoscale. It is based on density-functional theory, plane waves, and pseudopotentials. Learn more via the OpenBenchmarking.org test page.
RELION RELION - REgularised LIkelihood OptimisatioN - is a stand-alone computer program for Maximum A Posteriori refinement of (multiple) 3D reconstructions or 2D class averages in cryo-electron microscopy (cryo-EM). It is developed in the research group of Sjors Scheres at the MRC Laboratory of Molecular Biology. Learn more via the OpenBenchmarking.org test page.
Rodinia Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes select OpenCL, NVIDIA CUDA and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.
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.
Result
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 0.9.0 Encoder Speed: 6, Lossless P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 9 18 27 36 45 SE +/- 0.28, N = 15 SE +/- 0.22, N = 3 SE +/- 0.41, N = 3 SE +/- 0.20, N = 3 37.47 30.35 31.99 30.01 1. (CXX) g++ options: -O3 -fPIC -lm CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 3210 3409 P-State performance 3403 3406 3442 intel_cpufreq schedutil 1421 3285 3409 intel_cpufreq performance 3403 3405 3410 OpenBenchmarking.org Megahertz, More Is Better libavif avifenc 0.9.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 68.2 268.1 414.9 P-State performance 151.5 312.4 462.5 intel_cpufreq schedutil 68.2 281.5 459.7 intel_cpufreq performance 166.8 314.9 476.7 OpenBenchmarking.org Watts, Fewer Is Better libavif avifenc 0.9.0 CPU Power Consumption Monitor 120 240 360 480 600 1. P-State powersave: Approximate power consumption of 2628 Joules per run. 2. P-State performance: Approximate power consumption of 2500 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 2440 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 2519 Joules per run. CPU Temp
Min Avg Max P-State powersave 35.0 54.1 71.0 P-State performance 51.0 53.8 58.0 intel_cpufreq schedutil 35.0 55.1 69.0 intel_cpufreq performance 55.0 61.9 73.0 OpenBenchmarking.org Celsius, Fewer Is Better libavif avifenc 0.9.0 CPU Temperature Monitor 20 40 60 80 100 Result Confidence
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 0.9.0 Encoder Speed: 6, Lossless P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 8 16 24 32 40 Min: 35.88 / Avg: 37.47 / Max: 38.97 Min: 30.07 / Avg: 30.35 / Max: 30.77 Min: 31.51 / Avg: 31.99 / Max: 32.81 Min: 29.62 / Avg: 30.01 / Max: 30.26 1. (CXX) g++ options: -O3 -fPIC -lm
Result
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 0.9.0 Encoder Speed: 10, Lossless P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 2 4 6 8 10 SE +/- 0.141, N = 15 SE +/- 0.031, N = 6 SE +/- 0.124, N = 15 SE +/- 0.040, N = 6 8.833 7.248 8.486 7.276 1. (CXX) g++ options: -O3 -fPIC -lm CPU Peak Freq (Highest CPU Core Frequency
Min Avg Max P-State powersave 801 2545 3404 P-State performance 3403 3403 3404 intel_cpufreq schedutil 802 3113 3405 intel_cpufreq performance 3403 3404 3408 OpenBenchmarking.org Megahertz, More Is Better libavif avifenc 0.9.0 CPU Peak Freq (Highest CPU Core Frequency) Monitor 800 1600 2400 3200 4000 CPU Power Consumption
Min Avg Max P-State powersave 61.8 132.4 209.6 P-State performance 174.7 208.8 243.2 intel_cpufreq schedutil 71.0 148.0 226.2 intel_cpufreq performance 176.2 208.8 247.3 OpenBenchmarking.org Watts, Fewer Is Better libavif avifenc 0.9.0 CPU Power Consumption Monitor 60 120 180 240 300 1. P-State powersave: Approximate power consumption of 406 Joules per run. 2. P-State performance: Approximate power consumption of 557 Joules per run. 3. intel_cpufreq schedutil: Approximate power consumption of 434 Joules per run. 4. intel_cpufreq performance: Approximate power consumption of 592 Joules per run. CPU Temp
Min Avg Max P-State powersave 36.0 39.1 50.0 P-State performance 44.0 45.3 46.0 intel_cpufreq schedutil 38.0 40.7 43.0 intel_cpufreq performance 47.0 49.5 52.0 OpenBenchmarking.org Celsius, Fewer Is Better libavif avifenc 0.9.0 CPU Temperature Monitor 15 30 45 60 75 Result Confidence
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 0.9.0 Encoder Speed: 10, Lossless P-State powersave P-State performance intel_cpufreq schedutil intel_cpufreq performance 3 6 9 12 15 Min: 7.79 / Avg: 8.83 / Max: 9.65 Min: 7.17 / Avg: 7.25 / Max: 7.37 Min: 7.9 / Avg: 8.49 / Max: 9.6 Min: 7.09 / Avg: 7.28 / Max: 7.35 1. (CXX) g++ options: -O3 -fPIC -lm
Timed Godot Game Engine Compilation This test times how long it takes to compile the Godot Game Engine. Godot is a popular, open-source, cross-platform 2D/3D game engine and is built using the SCons build system and targeting the X11 platform. Learn more via the OpenBenchmarking.org test page.
Timed Mesa Compilation This test profile times how long it takes to compile Mesa with Meson/Ninja. For minimizing build dependencies and avoid versioning conflicts, test this is just the core Mesa build without LLVM or the extra Gallium3D/Mesa drivers enabled. Learn more via the OpenBenchmarking.org test page.
Timed Node.js Compilation This test profile times how long it takes to build/compile Node.js itself from source. Node.js is a JavaScript run-time built from the Chrome V8 JavaScript engine while itself is written in C/C++. Learn more via the OpenBenchmarking.org test page.
Blender Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL, NVIDIA OptiX, and NVIDIA CUDA is supported. Learn more via the OpenBenchmarking.org test page.
NWChem NWChem is an open-source high performance computational chemistry package. Per NWChem's documentation, "NWChem aims to provide its users with computational chemistry tools that are scalable both in their ability to treat large scientific computational chemistry problems efficiently, and in their use of available parallel computing resources from high-performance parallel supercomputers to conventional workstation clusters." Learn more via the OpenBenchmarking.org test page.
P-State powersave Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --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: intel_pstate powersave - CPU Microcode: 0xd000270Python Notes: Python 3.9.4Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 23 May 2021 12:11 by user root.
P-State performance Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --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: intel_pstate performance - CPU Microcode: 0xd000270Python Notes: Python 3.9.4Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 24 May 2021 05:21 by user root.
intel_cpufreq schedutil Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --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: intel_cpufreq schedutil - CPU Microcode: 0xd000270Python Notes: Python 3.9.4Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 24 May 2021 17:18 by user root.
intel_cpufreq performance Processor: 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads), Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS), Chipset: Intel Device 0998, Memory: 16 x 32 GB DDR4-3200MT/s Hynix HMA84GR7CJR4N-XN, Disk: 800GB INTEL SSDPF21Q800GB, Graphics: ASPEED, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP
OS: Ubuntu 21.04, Kernel: 5.11.0-17-generic (x86_64), Desktop: GNOME Shell 3.38.4, Display Server: X Server, Vulkan: 1.0.2, Compiler: GCC 10.3.0, File-System: ext4, Screen Resolution: 1024x768
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --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: intel_cpufreq performance - CPU Microcode: 0xd000270Python Notes: Python 3.9.4Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 25 May 2021 09:25 by user root.