AMD EPYC 7F72 2P Linux 5.11 Perf Governor 2 x AMD EPYC 7F72 24-Core testing looking at CPU freq invariance on 5.11 with patch. CPU power consumption monitoring via AMD_Energy interface at 1 second polling. Additional data with CPUFreq performance governor included.
HTML result view exported from: https://openbenchmarking.org/result/2102048-HA-AMDEPYC7F37 .
AMD EPYC 7F72 2P Linux 5.11 Perf Governor Processor Motherboard Chipset Memory Disk Graphics Network Monitor OS Kernel Desktop Display Server Display Driver Compiler File-System Screen Resolution CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 2 x AMD EPYC 7F72 24-Core @ 3.20GHz (48 Cores / 96 Threads) Supermicro H11DSi-NT v2.00 (2.1 BIOS) AMD Starship/Matisse 16 x 8192 MB DDR4-3200MT/s HMA81GR7CJR8N-XN 1000GB Western Digital WD_BLACK SN850 1TB ASPEED 2 x Intel 10G X550T Ubuntu 20.10 5.11.0-rc4-max-boost-inv-patch (x86_64) 20210121 GNOME Shell 3.38.1 X Server 1.20.9 modesetting 1.20.9 GCC 10.2.0 ext4 1920x1080 5.11.0-051100rc4daily20210122-generic (x86_64) 20210121 VE228 5.11.0-rc4-max-boost-inv-patch (x86_64) 20210121 2 x AMD EPYC 7F72 24-Core @ 3.71GHz (48 Cores / 96 Threads) 5.11.0-rc6-phx (x86_64) 20210203 aspeed OpenBenchmarking.org Kernel Details - Transparent Huge Pages: madvise Compiler Details - --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --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-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/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-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 Disk Details - CPUFreq Performance, Linux 5.11 Git, Linux 5.11 Giovanni Patch: NONE / errors=remount-ro,relatime,rw / Block Size: 4096 Processor Details - CPUFreq Performance: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0x8301034 - Linux 5.11 Git: Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0x8301034 - Linux 5.11 Giovanni Patch: Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0x8301034 - Linux 5.11 Rafael Patch: Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0x8301034 Java Details - OpenJDK Runtime Environment (build 11.0.9.1+1-Ubuntu-0ubuntu1.20.10) Python Details - Python 3.8.6 Security Details - itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected
AMD EPYC 7F72 2P Linux 5.11 Perf Governor build-gdb: Time To Compile cpuminer-opt: Skeincoin cpuminer-opt: LBC, LBRY Credits dav1d: Summer Nature 4K dav1d: Chimera 1080p 10-bit lammps: Rhodopsin Protein npb: LU.C onnx: yolov4 - OpenMP CPU openfoam: Motorbike 30M openfoam: Motorbike 60M qe: AUSURF112 qmcpack: simple-H2O rav1e: 10 rav1e: 6 rav1e: 5 redis: SET ai-benchmark: Device Inference Score ai-benchmark: Device Training Score ai-benchmark: Device AI Score dacapobench: Jython dacapobench: Tradebeans dacapobench: Tradesoap ospray: Magnetic Reconnection - Path Tracer ospray: San Miguel - SciVis clomp: Static OMP Speedup influxdb: 4 - 10000 - 2,5000,1 - 10000 influxdb: 64 - 10000 - 2,5000,1 - 10000 lczero: BLAS lczero: Eigen onednn: Convolution Batch Shapes Auto - f32 - CPU onednn: Deconvolution Batch shapes_1d - f32 - CPU onednn: IP Shapes 3D - f32 - CPU onednn: Matrix Multiply Batch Shapes Transformer - f32 - CPU ttsiod-renderer: Phong Rendering With Soft-Shadow Mapping rodinia: OpenMP CFD Solver rodinia: OpenMP Streamcluster tensorflow-lite: Mobilenet Float tensorflow-lite: Mobilenet Quant tensorflow-lite: NASNet Mobile tensorflow-lite: SqueezeNet tensorflow-lite: Inception ResNet V2 tensorflow-lite: Inception V4 x265: Bosphorus 1080p x265: Bosphorus 4K CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 85.277 522948 194087 363.33 181.95 24.639 153770.57 185 18.29 128.81 1249.06 28.971 3.177 1.446 1.095 1454741.42 1775 1133 2908 4740 4671 4621 333.33 55.56 47.4 956189.4 1360163.0 4147 4450 0.867545 2.30467 0.813628 0.514604 665.448 8.510 10.432 39981.9 41180.1 132844 61347.2 737993 818887 62.29 20.75 97.641 363784 132477 308.29 130.61 21.129 147443.86 175 18.71 129.67 1217.49 31.177 2.902 1.370 1.045 1380890.22 1697 1059 2756 4897 5954 5170 250 52.63 43.9 807463.1 1231991.2 4106 4284 0.914198 2.40549 0.881348 0.547674 627.208 9.255 11.209 46659.2 45083.9 189771 65193.0 765726 894640 47.66 18.63 92.916 364017 139037 317.45 133.37 23.787 154376.76 181 18.30 128.28 1171.03 29.281 3.054 1.408 1.068 1427348.10 1720 1067 2787 4778 5591 5148 250 54.97 47.8 812193.6 1256112.1 4061 4433 0.863782 2.33290 0.849248 0.521968 655.225 8.882 10.338 39523.5 41034.0 134044 62195.4 736285 810750 49.45 19.74 93.098 366648 137557 315.99 133.69 24.795 154111.43 181 18.32 129.05 1222.60 29.040 3.045 1.398 1.064 1401394.30 1713 1061 2774 4843 5600 5167 250 54.97 46.9 812464.3 1264917.7 3522 3749 0.870351 2.31171 0.823825 0.511809 642.251 8.952 10.567 40060.0 40544.1 135467 62054.2 724454 815478 50.43 19.93 OpenBenchmarking.org
Timed GDB GNU Debugger Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed GDB GNU Debugger Compilation 9.1 Time To Compile CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 20 40 60 80 100 SE +/- 0.14, N = 3 SE +/- 0.40, N = 3 SE +/- 0.43, N = 3 SE +/- 0.23, N = 3 85.28 97.64 92.92 93.10
Timed GDB GNU Debugger Compilation CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.7 167.0 492.1 Linux 5.11 Git 121.0 157.1 492.3 Linux 5.11 Giovanni Patch 120.6 159.2 492.3 OpenBenchmarking.org Watts, Fewer Is Better Timed GDB GNU Debugger Compilation 9.1 CPU Power Consumption Monitor 130 260 390 520 650
Cpuminer-Opt Algorithm: Skeincoin OpenBenchmarking.org kH/s, More Is Better Cpuminer-Opt 3.15.5 Algorithm: Skeincoin CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 110K 220K 330K 440K 550K SE +/- 9614.65, N = 12 SE +/- 3597.13, N = 15 SE +/- 5604.68, N = 12 SE +/- 2614.77, N = 13 522948 363784 364017 366648 1. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
Cpuminer-Opt Algorithm: Skeincoin OpenBenchmarking.org kH/s Per Watt, More Is Better Cpuminer-Opt 3.15.5 Algorithm: Skeincoin CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 800 1600 2400 3200 4000 3625.65 2899.04 2897.78
Cpuminer-Opt CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 72.0 144.2 193.5 Linux 5.11 Git 120.4 125.5 148.6 Linux 5.11 Giovanni Patch 120.0 125.6 144.4 OpenBenchmarking.org Watts, Fewer Is Better Cpuminer-Opt 3.15.5 CPU Power Consumption Monitor 50 100 150 200 250
Cpuminer-Opt Algorithm: LBC, LBRY Credits OpenBenchmarking.org kH/s, More Is Better Cpuminer-Opt 3.15.5 Algorithm: LBC, LBRY Credits CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 40K 80K 120K 160K 200K SE +/- 861.90, N = 3 SE +/- 1036.73, N = 3 SE +/- 1380.06, N = 3 SE +/- 1545.00, N = 3 194087 132477 139037 137557 1. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
Cpuminer-Opt Algorithm: LBC, LBRY Credits OpenBenchmarking.org kH/s Per Watt, More Is Better Cpuminer-Opt 3.15.5 Algorithm: LBC, LBRY Credits CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 300 600 900 1200 1500 1315.46 1052.71 1091.62
Cpuminer-Opt CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.6 147.5 245.0 Linux 5.11 Git 120.3 125.8 163.2 Linux 5.11 Giovanni Patch 119.9 127.4 169.1 OpenBenchmarking.org Watts, Fewer Is Better Cpuminer-Opt 3.15.5 CPU Power Consumption Monitor 60 120 180 240 300
dav1d Video Input: Summer Nature 4K OpenBenchmarking.org FPS, More Is Better dav1d 0.8.1 Video Input: Summer Nature 4K CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 80 160 240 320 400 SE +/- 3.54, N = 15 SE +/- 1.86, N = 3 SE +/- 0.53, N = 3 SE +/- 2.68, N = 3 363.33 308.29 317.45 315.99 MIN: 186.32 / MAX: 403.05 MIN: 163.13 / MAX: 334.13 MIN: 173.69 / MAX: 340.43 MIN: 157.61 / MAX: 340.72 1. (CC) gcc options: -pthread
dav1d Video Input: Summer Nature 4K OpenBenchmarking.org FPS Per Watt, More Is Better dav1d 0.8.1 Video Input: Summer Nature 4K CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.414 0.828 1.242 1.656 2.07 1.84 1.77 1.72
dav1d CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 119.9 197.5 386.9 Linux 5.11 Git 120.4 174.0 312.0 Linux 5.11 Giovanni Patch 119.8 184.4 344.4 OpenBenchmarking.org Watts, Fewer Is Better dav1d 0.8.1 CPU Power Consumption Monitor 100 200 300 400 500
dav1d Video Input: Chimera 1080p 10-bit OpenBenchmarking.org FPS, More Is Better dav1d 0.8.1 Video Input: Chimera 1080p 10-bit CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 40 80 120 160 200 SE +/- 0.23, N = 3 SE +/- 0.23, N = 3 SE +/- 0.14, N = 3 SE +/- 0.19, N = 3 181.95 130.61 133.37 133.69 MIN: 125.32 / MAX: 275.36 MIN: 90.23 / MAX: 199.74 MIN: 92.59 / MAX: 205.11 MIN: 92.24 / MAX: 206.31 1. (CC) gcc options: -pthread
dav1d Video Input: Chimera 1080p 10-bit OpenBenchmarking.org FPS Per Watt, More Is Better dav1d 0.8.1 Video Input: Chimera 1080p 10-bit CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.2723 0.5446 0.8169 1.0892 1.3615 1.21 0.91 0.92
dav1d CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 119.6 150.2 248.8 Linux 5.11 Git 67.4 143.3 181.4 Linux 5.11 Giovanni Patch 119.7 144.8 190.2 OpenBenchmarking.org Watts, Fewer Is Better dav1d 0.8.1 CPU Power Consumption Monitor 60 120 180 240 300
LAMMPS Molecular Dynamics Simulator Model: Rhodopsin Protein OpenBenchmarking.org ns/day, More Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 Model: Rhodopsin Protein CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 6 12 18 24 30 SE +/- 0.19, N = 15 SE +/- 0.23, N = 15 SE +/- 0.17, N = 12 SE +/- 0.16, N = 3 24.64 21.13 23.79 24.80 1. (CXX) g++ options: -O3 -pthread -lm
LAMMPS Molecular Dynamics Simulator Model: Rhodopsin Protein OpenBenchmarking.org ns/day Per Watt, More Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 Model: Rhodopsin Protein CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.0315 0.063 0.0945 0.126 0.1575 0.14 0.12 0.13
LAMMPS Molecular Dynamics Simulator CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.5 176.0 401.4 Linux 5.11 Git 120.7 178.3 411.4 Linux 5.11 Giovanni Patch 120.4 176.6 409.0 OpenBenchmarking.org Watts, Fewer Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 CPU Power Consumption Monitor 110 220 330 440 550
NAS Parallel Benchmarks Test / Class: LU.C OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 30K 60K 90K 120K 150K SE +/- 121.33, N = 4 SE +/- 1780.52, N = 15 SE +/- 509.59, N = 4 SE +/- 237.90, N = 3 153770.57 147443.86 154376.76 154111.43 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent -levent_pthreads -lutil -lm -lrt -lz 2. Open MPI 4.0.3
NAS Parallel Benchmarks Test / Class: LU.C OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 100 200 300 400 500 466.44 440.23 469.62
NAS Parallel Benchmarks CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.1 329.7 405.5 Linux 5.11 Git 122.6 334.9 407.9 Linux 5.11 Giovanni Patch 121.0 328.7 407.9 OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor 110 220 330 440 550
ONNX Runtime Model: yolov4 - Device: OpenMP CPU OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.6 Model: yolov4 - Device: OpenMP CPU CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 40 80 120 160 200 SE +/- 2.62, N = 3 SE +/- 1.60, N = 12 SE +/- 1.86, N = 3 SE +/- 1.86, N = 4 185 175 181 181 1. (CXX) g++ options: -fopenmp -ffunction-sections -fdata-sections -O3 -ldl -lrt
ONNX Runtime Model: yolov4 - Device: OpenMP CPU OpenBenchmarking.org Inferences Per Minute Per Watt, More Is Better ONNX Runtime 1.6 Model: yolov4 - Device: OpenMP CPU CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.1575 0.315 0.4725 0.63 0.7875 0.70 0.67 0.69
ONNX Runtime CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.3 263.5 278.7 Linux 5.11 Git 122.5 262.9 279.9 Linux 5.11 Giovanni Patch 121.1 263.9 279.9 OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.6 CPU Power Consumption Monitor 70 140 210 280 350
OpenFOAM Input: Motorbike 30M OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 8 Input: Motorbike 30M CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 5 10 15 20 25 SE +/- 0.10, N = 3 SE +/- 0.14, N = 3 SE +/- 0.08, N = 3 SE +/- 0.16, N = 3 18.29 18.71 18.30 18.32 1. (CXX) g++ options: -std=c++11 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm
OpenFOAM CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.9 309.6 472.6 Linux 5.11 Git 122.6 314.5 470.8 Linux 5.11 Giovanni Patch 121.2 309.5 472.2 OpenBenchmarking.org Watts, Fewer Is Better OpenFOAM 8 CPU Power Consumption Monitor 120 240 360 480 600
OpenFOAM Input: Motorbike 60M OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 8 Input: Motorbike 60M CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 30 60 90 120 150 SE +/- 0.09, N = 3 SE +/- 0.09, N = 3 SE +/- 0.07, N = 3 SE +/- 0.05, N = 3 128.81 129.67 128.28 129.05 1. (CXX) g++ options: -std=c++11 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm
OpenFOAM CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.1 290.0 399.7 Linux 5.11 Git 122.1 288.9 399.6 Linux 5.11 Giovanni Patch 121.0 290.6 400.9 OpenBenchmarking.org Watts, Fewer Is Better OpenFOAM 8 CPU Power Consumption Monitor 110 220 330 440 550
Quantum ESPRESSO Input: AUSURF112 OpenBenchmarking.org Seconds, Fewer Is Better Quantum ESPRESSO 6.7 Input: AUSURF112 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 300 600 900 1200 1500 SE +/- 19.05, N = 9 SE +/- 11.28, N = 3 SE +/- 12.21, N = 4 SE +/- 23.97, N = 6 1249.06 1217.49 1171.03 1222.60 1. (F9X) gfortran options: -lopenblas -lFoX_dom -lFoX_sax -lFoX_wxml -lFoX_common -lFoX_utils -lFoX_fsys -lfftw3 -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -ldl -levent -levent_pthreads -lutil -lm -lrt -lz
Quantum ESPRESSO CPU Power Consumption Monitor OpenBenchmarking.org Watts, Fewer Is Better Quantum ESPRESSO 6.7 CPU Power Consumption Monitor CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 90 180 270 360 450 Min: 70.27 / Avg: 389.94 / Max: 492.33 Min: 73.52 / Avg: 387.82 / Max: 492.31 Min: 120.56 / Avg: 385.69 / Max: 492.5
QMCPACK Input: simple-H2O OpenBenchmarking.org Total Execution Time - Seconds, Fewer Is Better QMCPACK 3.10 Input: simple-H2O CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 7 14 21 28 35 SE +/- 0.06, N = 3 SE +/- 0.53, N = 15 SE +/- 0.08, N = 3 SE +/- 0.04, N = 3 28.97 31.18 29.28 29.04 1. (CXX) g++ options: -fopenmp -finline-limit=1000 -fstrict-aliasing -funroll-all-loops -march=native -O3 -fomit-frame-pointer -ffast-math -pthread -lm
QMCPACK CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.6 438.1 491.9 Linux 5.11 Git 121.2 433.2 491.7 Linux 5.11 Giovanni Patch 120.3 437.6 491.3 OpenBenchmarking.org Watts, Fewer Is Better QMCPACK 3.10 CPU Power Consumption Monitor 130 260 390 520 650
rav1e Speed: 10 OpenBenchmarking.org Frames Per Second, More Is Better rav1e 0.4 Speed: 10 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 0.7148 1.4296 2.1444 2.8592 3.574 SE +/- 0.018, N = 3 SE +/- 0.016, N = 3 SE +/- 0.008, N = 3 SE +/- 0.011, N = 3 3.177 2.902 3.054 3.045
rav1e Speed: 10 OpenBenchmarking.org Frames Per Second Per Watt, More Is Better rav1e 0.4 Speed: 10 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.0045 0.009 0.0135 0.018 0.0225 0.02 0.02 0.02
rav1e CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 119.9 138.6 146.5 Linux 5.11 Git 120.8 137.3 145.9 Linux 5.11 Giovanni Patch 120.2 136.5 145.8 OpenBenchmarking.org Watts, Fewer Is Better rav1e 0.4 CPU Power Consumption Monitor 40 80 120 160 200
rav1e Speed: 6 OpenBenchmarking.org Frames Per Second, More Is Better rav1e 0.4 Speed: 6 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 0.3254 0.6508 0.9762 1.3016 1.627 SE +/- 0.001, N = 3 SE +/- 0.002, N = 3 SE +/- 0.003, N = 3 SE +/- 0.002, N = 3 1.446 1.370 1.408 1.398
rav1e Speed: 6 OpenBenchmarking.org Frames Per Second Per Watt, More Is Better rav1e 0.4 Speed: 6 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.0023 0.0046 0.0069 0.0092 0.0115 0.01 0.01 0.01
rav1e CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 74.7 143.0 152.2 Linux 5.11 Git 120.7 142.0 151.6 Linux 5.11 Giovanni Patch 74.9 142.0 151.6 OpenBenchmarking.org Watts, Fewer Is Better rav1e 0.4 CPU Power Consumption Monitor 40 80 120 160 200
rav1e Speed: 5 OpenBenchmarking.org Frames Per Second, More Is Better rav1e 0.4 Speed: 5 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 0.2464 0.4928 0.7392 0.9856 1.232 SE +/- 0.003, N = 3 SE +/- 0.001, N = 3 SE +/- 0.001, N = 3 SE +/- 0.002, N = 3 1.095 1.045 1.068 1.064
rav1e Speed: 5 OpenBenchmarking.org Frames Per Second Per Watt, More Is Better rav1e 0.4 Speed: 5 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.0023 0.0046 0.0069 0.0092 0.0115 0.01 0.01 0.01
rav1e CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 119.8 143.6 155.6 Linux 5.11 Git 120.4 143.1 157.6 Linux 5.11 Giovanni Patch 120.0 143.3 155.2 OpenBenchmarking.org Watts, Fewer Is Better rav1e 0.4 CPU Power Consumption Monitor 40 80 120 160 200
Redis Test: SET OpenBenchmarking.org Requests Per Second, More Is Better Redis 6.0.9 Test: SET CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 300K 600K 900K 1200K 1500K SE +/- 10017.79, N = 13 SE +/- 10410.66, N = 15 SE +/- 13176.39, N = 15 SE +/- 15335.21, N = 5 1454741.42 1380890.22 1427348.10 1401394.30 1. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3
Redis Test: SET OpenBenchmarking.org Requests Per Second Per Watt, More Is Better Redis 6.0.9 Test: SET CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 3K 6K 9K 12K 15K 11683.85 11015.92 11500.61
Redis CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 119.7 124.5 135.4 Linux 5.11 Git 120.5 125.4 135.4 Linux 5.11 Giovanni Patch 119.7 124.1 134.6 OpenBenchmarking.org Watts, Fewer Is Better Redis 6.0.9 CPU Power Consumption Monitor 40 80 120 160 200
AI Benchmark Alpha Device Inference Score OpenBenchmarking.org Score, More Is Better AI Benchmark Alpha 0.1.2 Device Inference Score CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 400 800 1200 1600 2000 1775 1697 1720 1713
AI Benchmark Alpha Device Training Score OpenBenchmarking.org Score, More Is Better AI Benchmark Alpha 0.1.2 Device Training Score CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 200 400 600 800 1000 1133 1059 1067 1061
AI Benchmark Alpha Device AI Score OpenBenchmarking.org Score, More Is Better AI Benchmark Alpha 0.1.2 Device AI Score CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 600 1200 1800 2400 3000 2908 2756 2787 2774
AI Benchmark Alpha Device AI Score OpenBenchmarking.org Score Per Watt, More Is Better AI Benchmark Alpha 0.1.2 Device AI Score CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 3 6 9 12 15 11.59 12.95 12.98
AI Benchmark Alpha CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 119.8 250.8 388.9 Linux 5.11 Git 120.4 212.9 368.5 Linux 5.11 Giovanni Patch 119.9 214.7 378.2 OpenBenchmarking.org Watts, Fewer Is Better AI Benchmark Alpha 0.1.2 CPU Power Consumption Monitor 100 200 300 400 500
DaCapo Benchmark Java Test: Jython OpenBenchmarking.org msec, Fewer Is Better DaCapo Benchmark 9.12-MR1 Java Test: Jython CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 1000 2000 3000 4000 5000 SE +/- 20.83, N = 6 SE +/- 28.66, N = 18 SE +/- 43.93, N = 6 SE +/- 24.16, N = 4 4740 4897 4778 4843
DaCapo Benchmark CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 119.9 138.2 194.4 Linux 5.11 Git 120.6 138.7 198.5 Linux 5.11 Giovanni Patch 120.0 137.8 195.6 OpenBenchmarking.org Watts, Fewer Is Better DaCapo Benchmark 9.12-MR1 CPU Power Consumption Monitor 50 100 150 200 250
DaCapo Benchmark Java Test: Tradebeans OpenBenchmarking.org msec, Fewer Is Better DaCapo Benchmark 9.12-MR1 Java Test: Tradebeans CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 1300 2600 3900 5200 6500 SE +/- 52.34, N = 20 SE +/- 50.83, N = 20 SE +/- 66.39, N = 20 SE +/- 53.96, N = 20 4671 5954 5591 5600
DaCapo Benchmark CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 119.8 149.7 216.0 Linux 5.11 Git 120.6 140.2 198.0 Linux 5.11 Giovanni Patch 119.9 140.5 200.9 OpenBenchmarking.org Watts, Fewer Is Better DaCapo Benchmark 9.12-MR1 CPU Power Consumption Monitor 60 120 180 240 300
DaCapo Benchmark Java Test: Tradesoap OpenBenchmarking.org msec, Fewer Is Better DaCapo Benchmark 9.12-MR1 Java Test: Tradesoap CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 1100 2200 3300 4400 5500 SE +/- 42.72, N = 5 SE +/- 44.82, N = 4 SE +/- 61.21, N = 4 SE +/- 47.07, N = 4 4621 5170 5148 5167
DaCapo Benchmark CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.2 171.7 314.7 Linux 5.11 Git 120.8 163.6 307.8 Linux 5.11 Giovanni Patch 120.1 165.5 309.0 OpenBenchmarking.org Watts, Fewer Is Better DaCapo Benchmark 9.12-MR1 CPU Power Consumption Monitor 80 160 240 320 400
OSPray Demo: Magnetic Reconnection - Renderer: Path Tracer OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: Magnetic Reconnection - Renderer: Path Tracer CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 70 140 210 280 350 SE +/- 0.00, N = 11 333.33 250.00 250.00 250.00 MIN: 100 / MAX: 500 MIN: 90.91 / MAX: 500 MIN: 90.91 / MAX: 333.33 MIN: 111.11 / MAX: 333.33
OSPray Demo: Magnetic Reconnection - Renderer: Path Tracer OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: Magnetic Reconnection - Renderer: Path Tracer CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.4478 0.8956 1.3434 1.7912 2.239 1.99 1.61 1.65
OSPray CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.4 167.3 303.3 Linux 5.11 Git 120.9 154.8 256.7 Linux 5.11 Giovanni Patch 120.2 151.8 245.9 OpenBenchmarking.org Watts, Fewer Is Better OSPray 1.8.5 CPU Power Consumption Monitor 80 160 240 320 400
OSPray Demo: San Miguel - Renderer: SciVis OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: San Miguel - Renderer: SciVis CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 12 24 36 48 60 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.58, N = 5 SE +/- 0.58, N = 5 55.56 52.63 54.97 54.97 MIN: 33.33 / MAX: 58.82 MIN: 27.03 / MAX: 58.82 MIN: 31.25 / MAX: 58.82 MIN: 25.64 / MAX: 58.82
OSPray Demo: San Miguel - Renderer: SciVis OpenBenchmarking.org FPS Per Watt, More Is Better OSPray 1.8.5 Demo: San Miguel - Renderer: SciVis CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.0743 0.1486 0.2229 0.2972 0.3715 0.33 0.30 0.32
OSPray CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.6 168.2 482.0 Linux 5.11 Git 121.4 173.5 481.5 Linux 5.11 Giovanni Patch 120.7 171.8 482.2 OpenBenchmarking.org Watts, Fewer Is Better OSPray 1.8.5 CPU Power Consumption Monitor 120 240 360 480 600
CLOMP Static OMP Speedup OpenBenchmarking.org Speedup, More Is Better CLOMP 1.2 Static OMP Speedup CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 11 22 33 44 55 SE +/- 0.55, N = 3 SE +/- 0.60, N = 3 SE +/- 0.47, N = 3 SE +/- 0.98, N = 15 47.4 43.9 47.8 46.9 1. (CC) gcc options: -fopenmp -O3 -lm
CLOMP Static OMP Speedup OpenBenchmarking.org Speedup Per Watt, More Is Better CLOMP 1.2 Static OMP Speedup CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.0698 0.1396 0.2094 0.2792 0.349 0.31 0.29 0.31
CLOMP CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 119.9 151.3 312.5 Linux 5.11 Git 120.4 153.4 298.4 Linux 5.11 Giovanni Patch 119.9 152.6 321.5 OpenBenchmarking.org Watts, Fewer Is Better CLOMP 1.2 CPU Power Consumption Monitor 80 160 240 320 400
InfluxDB Concurrent Streams: 4 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 OpenBenchmarking.org val/sec, More Is Better InfluxDB 1.8.2 Concurrent Streams: 4 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 200K 400K 600K 800K 1000K SE +/- 2401.68, N = 3 SE +/- 2183.04, N = 3 SE +/- 1525.09, N = 3 SE +/- 757.15, N = 3 956189.4 807463.1 812193.6 812464.3
InfluxDB Concurrent Streams: 4 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 OpenBenchmarking.org val/sec Per Watt, More Is Better InfluxDB 1.8.2 Concurrent Streams: 4 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 1100 2200 3300 4400 5500 4963.60 5060.30 4857.31
InfluxDB CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.8 192.6 218.3 Linux 5.11 Git 120.9 159.6 179.4 Linux 5.11 Giovanni Patch 120.5 167.2 187.5 OpenBenchmarking.org Watts, Fewer Is Better InfluxDB 1.8.2 CPU Power Consumption Monitor 60 120 180 240 300
InfluxDB Concurrent Streams: 64 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 OpenBenchmarking.org val/sec, More Is Better InfluxDB 1.8.2 Concurrent Streams: 64 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 300K 600K 900K 1200K 1500K SE +/- 9433.94, N = 3 SE +/- 6204.63, N = 3 SE +/- 2545.78, N = 3 SE +/- 5312.25, N = 3 1360163.0 1231991.2 1256112.1 1264917.7
InfluxDB Concurrent Streams: 64 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 OpenBenchmarking.org val/sec Per Watt, More Is Better InfluxDB 1.8.2 Concurrent Streams: 64 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 1500 3000 4500 6000 7500 6626.07 6781.37 6672.14
InfluxDB CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.9 205.3 232.0 Linux 5.11 Git 120.9 181.7 213.6 Linux 5.11 Giovanni Patch 120.3 188.3 217.2 OpenBenchmarking.org Watts, Fewer Is Better InfluxDB 1.8.2 CPU Power Consumption Monitor 60 120 180 240 300
LeelaChessZero Backend: BLAS OpenBenchmarking.org Nodes Per Second, More Is Better LeelaChessZero 0.26 Backend: BLAS CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 900 1800 2700 3600 4500 SE +/- 17.79, N = 3 SE +/- 50.84, N = 3 SE +/- 49.90, N = 9 SE +/- 36.49, N = 5 4147 4106 4061 3522 1. (CXX) g++ options: -flto -pthread
LeelaChessZero Backend: BLAS OpenBenchmarking.org Nodes Per Second Per Watt, More Is Better LeelaChessZero 0.26 Backend: BLAS CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 3 6 9 12 15 9.72 9.69 9.72
LeelaChessZero CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.0 426.6 492.4 Linux 5.11 Git 120.8 423.8 492.3 Linux 5.11 Giovanni Patch 120.4 417.7 492.3 OpenBenchmarking.org Watts, Fewer Is Better LeelaChessZero 0.26 CPU Power Consumption Monitor 130 260 390 520 650
LeelaChessZero Backend: Eigen OpenBenchmarking.org Nodes Per Second, More Is Better LeelaChessZero 0.26 Backend: Eigen CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 1000 2000 3000 4000 5000 SE +/- 26.71, N = 3 SE +/- 49.20, N = 4 SE +/- 36.23, N = 3 SE +/- 54.19, N = 9 4450 4284 4433 3749 1. (CXX) g++ options: -flto -pthread
LeelaChessZero Backend: Eigen OpenBenchmarking.org Nodes Per Second Per Watt, More Is Better LeelaChessZero 0.26 Backend: Eigen CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 3 6 9 12 15 10.39 10.08 10.45
LeelaChessZero CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.9 428.4 492.3 Linux 5.11 Git 122.0 425.0 492.3 Linux 5.11 Giovanni Patch 121.4 424.2 492.4 OpenBenchmarking.org Watts, Fewer Is Better LeelaChessZero 0.26 CPU Power Consumption Monitor 130 260 390 520 650
oneDNN Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.0 Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 0.2057 0.4114 0.6171 0.8228 1.0285 SE +/- 0.001697, N = 7 SE +/- 0.006064, N = 7 SE +/- 0.001510, N = 7 SE +/- 0.007039, N = 3 0.867545 0.914198 0.863782 0.870351 MIN: 0.78 MIN: 0.78 MIN: 0.79 MIN: 0.8 1. (CXX) g++ options: -O3 -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread
oneDNN CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.8 332.1 492.4 Linux 5.11 Git 122.1 326.0 492.0 Linux 5.11 Giovanni Patch 121.6 331.1 492.0 OpenBenchmarking.org Watts, Fewer Is Better oneDNN 2.0 CPU Power Consumption Monitor 130 260 390 520 650
oneDNN Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.0 Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 0.5412 1.0824 1.6236 2.1648 2.706 SE +/- 0.02560, N = 15 SE +/- 0.03372, N = 3 SE +/- 0.01587, N = 3 SE +/- 0.02377, N = 15 2.30467 2.40549 2.33290 2.31171 MIN: 1.88 MIN: 1.92 MIN: 2 MIN: 1.89 1. (CXX) g++ options: -O3 -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread
oneDNN CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.0 284.2 491.8 Linux 5.11 Git 123.1 282.9 453.8 Linux 5.11 Giovanni Patch 120.8 285.6 468.7 OpenBenchmarking.org Watts, Fewer Is Better oneDNN 2.0 CPU Power Consumption Monitor 130 260 390 520 650
oneDNN Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.0 Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 0.1983 0.3966 0.5949 0.7932 0.9915 SE +/- 0.005456, N = 5 SE +/- 0.005127, N = 5 SE +/- 0.004000, N = 5 SE +/- 0.006790, N = 15 0.813628 0.881348 0.849248 0.823825 MIN: 0.69 MIN: 0.71 MIN: 0.73 MIN: 0.68 1. (CXX) g++ options: -O3 -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread
oneDNN CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.1 239.4 430.2 Linux 5.11 Git 121.6 235.5 411.7 Linux 5.11 Giovanni Patch 120.7 235.0 415.0 OpenBenchmarking.org Watts, Fewer Is Better oneDNN 2.0 CPU Power Consumption Monitor 110 220 330 440 550
oneDNN Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.0 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 0.1232 0.2464 0.3696 0.4928 0.616 SE +/- 0.005906, N = 4 SE +/- 0.005010, N = 4 SE +/- 0.004601, N = 4 SE +/- 0.003992, N = 10 0.514604 0.547674 0.521968 0.511809 MIN: 0.43 MIN: 0.43 MIN: 0.43 MIN: 0.43 1. (CXX) g++ options: -O3 -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread
oneDNN CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.7 278.8 480.5 Linux 5.11 Git 122.2 271.2 438.7 Linux 5.11 Giovanni Patch 120.6 274.8 455.7 OpenBenchmarking.org Watts, Fewer Is Better oneDNN 2.0 CPU Power Consumption Monitor 120 240 360 480 600
TTSIOD 3D Renderer Phong Rendering With Soft-Shadow Mapping OpenBenchmarking.org FPS, More Is Better TTSIOD 3D Renderer 2.3b Phong Rendering With Soft-Shadow Mapping CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 140 280 420 560 700 SE +/- 5.92, N = 15 SE +/- 9.04, N = 15 SE +/- 3.22, N = 3 SE +/- 24.62, N = 12 665.45 627.21 655.23 642.25 1. (CXX) g++ options: -O3 -fomit-frame-pointer -ffast-math -mtune=native -flto -msse -mrecip -mfpmath=sse -msse2 -mssse3 -lSDL -fopenmp -fwhole-program -lstdc++
TTSIOD 3D Renderer Phong Rendering With Soft-Shadow Mapping OpenBenchmarking.org FPS Per Watt, More Is Better TTSIOD 3D Renderer 2.3b Phong Rendering With Soft-Shadow Mapping CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.6165 1.233 1.8495 2.466 3.0825 2.74 2.59 2.70
TTSIOD 3D Renderer CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.0 243.0 271.3 Linux 5.11 Git 122.3 242.6 267.9 Linux 5.11 Giovanni Patch 120.7 242.4 266.3 OpenBenchmarking.org Watts, Fewer Is Better TTSIOD 3D Renderer 2.3b CPU Power Consumption Monitor 70 140 210 280 350
Rodinia Test: OpenMP CFD Solver OpenBenchmarking.org Seconds, Fewer Is Better Rodinia 3.1 Test: OpenMP CFD Solver CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 3 6 9 12 15 SE +/- 0.053, N = 5 SE +/- 0.148, N = 15 SE +/- 0.141, N = 15 SE +/- 0.168, N = 15 8.510 9.255 8.882 8.952 1. (CXX) g++ options: -O2 -lOpenCL
Rodinia CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.9 253.1 370.6 Linux 5.11 Git 121.8 253.0 370.4 Linux 5.11 Giovanni Patch 121.0 253.4 373.0 OpenBenchmarking.org Watts, Fewer Is Better Rodinia 3.1 CPU Power Consumption Monitor 100 200 300 400 500
Rodinia Test: OpenMP Streamcluster OpenBenchmarking.org Seconds, Fewer Is Better Rodinia 3.1 Test: OpenMP Streamcluster CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 3 6 9 12 15 SE +/- 0.04, N = 5 SE +/- 0.22, N = 15 SE +/- 0.03, N = 5 SE +/- 0.13, N = 4 10.43 11.21 10.34 10.57 1. (CXX) g++ options: -O2 -lOpenCL
Rodinia CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.7 216.9 281.4 Linux 5.11 Git 121.2 220.5 291.1 Linux 5.11 Giovanni Patch 120.3 215.0 284.6 OpenBenchmarking.org Watts, Fewer Is Better Rodinia 3.1 CPU Power Consumption Monitor 70 140 210 280 350
TensorFlow Lite Model: Mobilenet Float OpenBenchmarking.org Microseconds, Fewer Is Better TensorFlow Lite 2020-08-23 Model: Mobilenet Float CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 10K 20K 30K 40K 50K SE +/- 473.02, N = 3 SE +/- 1144.94, N = 15 SE +/- 395.37, N = 3 SE +/- 126.68, N = 3 39981.9 46659.2 39523.5 40060.0
TensorFlow Lite CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.8 451.8 481.0 Linux 5.11 Git 122.2 441.5 477.4 Linux 5.11 Giovanni Patch 120.6 451.2 480.3 OpenBenchmarking.org Watts, Fewer Is Better TensorFlow Lite 2020-08-23 CPU Power Consumption Monitor 120 240 360 480 600
TensorFlow Lite Model: Mobilenet Quant OpenBenchmarking.org Microseconds, Fewer Is Better TensorFlow Lite 2020-08-23 Model: Mobilenet Quant CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 10K 20K 30K 40K 50K SE +/- 211.03, N = 3 SE +/- 759.19, N = 15 SE +/- 400.94, N = 6 SE +/- 60.30, N = 3 41180.1 45083.9 41034.0 40544.1
TensorFlow Lite CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.6 453.0 481.2 Linux 5.11 Git 123.7 447.4 480.2 Linux 5.11 Giovanni Patch 121.2 453.6 483.1 OpenBenchmarking.org Watts, Fewer Is Better TensorFlow Lite 2020-08-23 CPU Power Consumption Monitor 120 240 360 480 600
TensorFlow Lite Model: NASNet Mobile OpenBenchmarking.org Microseconds, Fewer Is Better TensorFlow Lite 2020-08-23 Model: NASNet Mobile CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 40K 80K 120K 160K 200K SE +/- 2146.20, N = 15 SE +/- 7366.43, N = 15 SE +/- 2393.85, N = 15 SE +/- 1630.66, N = 3 132844 189771 134044 135467
TensorFlow Lite CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.8 401.7 439.7 Linux 5.11 Git 123.2 385.7 435.9 Linux 5.11 Giovanni Patch 121.5 403.7 446.4 OpenBenchmarking.org Watts, Fewer Is Better TensorFlow Lite 2020-08-23 CPU Power Consumption Monitor 120 240 360 480 600
TensorFlow Lite Model: SqueezeNet OpenBenchmarking.org Microseconds, Fewer Is Better TensorFlow Lite 2020-08-23 Model: SqueezeNet CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 14K 28K 42K 56K 70K SE +/- 715.70, N = 4 SE +/- 690.93, N = 3 SE +/- 412.91, N = 15 SE +/- 479.59, N = 15 61347.2 65193.0 62195.4 62054.2
TensorFlow Lite CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.9 451.3 481.7 Linux 5.11 Git 123.4 441.7 476.9 Linux 5.11 Giovanni Patch 121.4 449.5 481.2 OpenBenchmarking.org Watts, Fewer Is Better TensorFlow Lite 2020-08-23 CPU Power Consumption Monitor 120 240 360 480 600
TensorFlow Lite Model: Inception ResNet V2 OpenBenchmarking.org Microseconds, Fewer Is Better TensorFlow Lite 2020-08-23 Model: Inception ResNet V2 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 160K 320K 480K 640K 800K SE +/- 2132.19, N = 3 SE +/- 4257.59, N = 3 SE +/- 5824.36, N = 9 SE +/- 1613.84, N = 3 737993 765726 736285 724454
TensorFlow Lite CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.8 439.4 470.5 Linux 5.11 Git 123.3 436.2 470.2 Linux 5.11 Giovanni Patch 121.6 442.7 471.4 OpenBenchmarking.org Watts, Fewer Is Better TensorFlow Lite 2020-08-23 CPU Power Consumption Monitor 120 240 360 480 600
TensorFlow Lite Model: Inception V4 OpenBenchmarking.org Microseconds, Fewer Is Better TensorFlow Lite 2020-08-23 Model: Inception V4 CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 200K 400K 600K 800K 1000K SE +/- 4685.69, N = 3 SE +/- 2435.29, N = 3 SE +/- 1163.43, N = 3 SE +/- 2213.60, N = 3 818887 894640 810750 815478
TensorFlow Lite CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 121.7 429.0 458.6 Linux 5.11 Git 122.2 414.2 453.9 Linux 5.11 Giovanni Patch 121.5 431.6 460.9 OpenBenchmarking.org Watts, Fewer Is Better TensorFlow Lite 2020-08-23 CPU Power Consumption Monitor 120 240 360 480 600
x265 Video Input: Bosphorus 1080p OpenBenchmarking.org Frames Per Second, More Is Better x265 3.4 Video Input: Bosphorus 1080p CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 14 28 42 56 70 SE +/- 0.73, N = 15 SE +/- 0.42, N = 7 SE +/- 0.52, N = 4 SE +/- 0.14, N = 3 62.29 47.66 49.45 50.43 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
x265 Video Input: Bosphorus 1080p OpenBenchmarking.org Frames Per Second Per Watt, More Is Better x265 3.4 Video Input: Bosphorus 1080p CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.0698 0.1396 0.2094 0.2792 0.349 0.31 0.30 0.29
x265 CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.6 200.8 253.8 Linux 5.11 Git 121.3 158.7 187.2 Linux 5.11 Giovanni Patch 121.1 169.9 202.1 OpenBenchmarking.org Watts, Fewer Is Better x265 3.4 CPU Power Consumption Monitor 70 140 210 280 350
x265 Video Input: Bosphorus 4K OpenBenchmarking.org Frames Per Second, More Is Better x265 3.4 Video Input: Bosphorus 4K CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch Linux 5.11 Rafael Patch 5 10 15 20 25 SE +/- 0.09, N = 3 SE +/- 0.10, N = 3 SE +/- 0.14, N = 3 SE +/- 0.07, N = 3 20.75 18.63 19.74 19.93 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
x265 Video Input: Bosphorus 4K OpenBenchmarking.org Frames Per Second Per Watt, More Is Better x265 3.4 Video Input: Bosphorus 4K CPUFreq Performance Linux 5.11 Git Linux 5.11 Giovanni Patch 0.018 0.036 0.054 0.072 0.09 0.08 0.08 0.08
x265 CPU Power Consumption Monitor Min Avg Max CPUFreq Performance 120.3 250.4 301.8 Linux 5.11 Git 121.0 221.5 281.7 Linux 5.11 Giovanni Patch 120.6 238.0 293.2 OpenBenchmarking.org Watts, Fewer Is Better x265 3.4 CPU Power Consumption Monitor 80 160 240 320 400
Phoronix Test Suite v10.8.4