AMD EPYC 9754 Bergamo SMT On/Off Comparison Benchmarks by Michael Larabel for a future article (post 19th) looking at SMT on/off comparison toggled via BIOS. SMT comparison testing of AMD EPYC 9754 128-Core CPUs on Titanite with Ubuntu 22.04 LTS.
HTML result view exported from: https://openbenchmarking.org/result/2307190-NE-BERGAMOSM27&grs&sro .
Processor Motherboard Chipset Memory Disk Graphics Network OS Kernel Desktop Display Server Vulkan Compiler File-System Screen Resolution EPYC 9754 1P EPYC 9754 2P SMT On SMT Off SMT On SMT Off AMD EPYC 9754 128-Core @ 2.25GHz (128 Cores / 256 Threads) AMD Titanite_4G (RTI1007B BIOS) AMD Device 14a4 768GB 2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007 ASPEED Broadcom NetXtreme BCM5720 PCIe Ubuntu 22.04 5.19.0-41-generic (x86_64) GNOME Shell 42.5 X Server 1.21.1.4 1.3.224 GCC 11.3.0 ext4 1024x768 AMD EPYC 9754 128-Core @ 2.25GHz (128 Cores) 2 x AMD EPYC 9754 128-Core @ 2.25GHz (256 Cores / 512 Threads) 1520GB 2 x AMD EPYC 9754 128-Core @ 2.25GHz (256 Cores) OpenBenchmarking.org Kernel Details - Transparent Huge Pages: madvise Compiler Details - --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --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-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-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 Processor Details - Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa0010b Python Details - Python 3.10.6 Security Details - EPYC 9754 1P: SMT On: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9754 1P: SMT Off: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: disabled RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9754 2P: SMT On: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9754 2P: SMT Off: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: disabled RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
openvino: Vehicle Detection FP16 - CPU openssl: SHA256 toybrot: TBB specfem3d: Water-layered Halfspace compress-7zip: Decompression Rating deepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Stream john-the-ripper: Blowfish john-the-ripper: bcrypt deepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream deepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream deepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream deepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream embree: Pathtracer ISPC - Crown deepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Stream graph500: 26 openssl: ChaCha20 embree: Pathtracer ISPC - Asian Dragon specfem3d: Mount St. Helens graph500: 26 blender: Classroom - CPU-Only npb: LU.C liquid-dsp: 512 - 256 - 512 ospray-studio: 3 - 4K - 32 - Path Tracer ospray-studio: 3 - 4K - 16 - Path Tracer ospray-studio: 1 - 4K - 16 - Path Tracer cp2k: H2O-DFT-LS ospray-studio: 1 - 4K - 1 - Path Tracer ospray-studio: 2 - 4K - 16 - Path Tracer cloverleaf: Lagrangian-Eulerian Hydrodynamics ospray-studio: 3 - 4K - 1 - Path Tracer ospray-studio: 2 - 4K - 32 - Path Tracer openssl: ChaCha20-Poly1305 ospray-studio: 2 - 4K - 1 - Path Tracer graph500: 26 ospray-studio: 1 - 4K - 32 - Path Tracer blender: Pabellon Barcelona - CPU-Only john-the-ripper: WPA PSK xmrig: Wownero - 1M astcenc: Exhaustive openvino: Person Detection FP32 - CPU openvino: Machine Translation EN To DE FP16 - CPU openvino: Person Detection FP16 - CPU specfem3d: Layered Halfspace blender: Barbershop - CPU-Only blender: BMW27 - CPU-Only helsing: 14 digit graph500: 26 openssl: RSA4096 npb: MG.C astcenc: Fast ospray: gravity_spheres_volume/dim_512/ao/real_time openssl: RSA4096 john-the-ripper: MD5 ospray: particle_volume/scivis/real_time blender: Fishy Cat - CPU-Only ospray: particle_volume/ao/real_time ospray: gravity_spheres_volume/dim_512/scivis/real_time openssl: SHA512 openvino: Face Detection FP16 - CPU openssl: AES-128-GCM openssl: AES-256-GCM openvino: Machine Translation EN To DE FP16 - CPU openvino: Face Detection FP16-INT8 - CPU openvino: Face Detection FP16-INT8 - CPU openvino: Face Detection FP16 - CPU deepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Stream astcenc: Thorough openvino: Weld Porosity Detection FP16-INT8 - CPU openvino: Weld Porosity Detection FP16 - CPU namd: ATPase Simulation - 327,506 Atoms liquid-dsp: 256 - 256 - 512 openvino: Person Vehicle Bike Detection FP16 - CPU build-linux-kernel: allmodconfig openvino: Weld Porosity Detection FP16 - CPU primesieve: 1e13 libxsmm: 256 toybrot: OpenMP openvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPU minibude: OpenMP - BM2 minibude: OpenMP - BM2 npb: IS.D npb: BT.C heffte: r2c - FFTW - float - 512 npb: SP.C luxcorerender: LuxCore Benchmark - CPU heffte: c2c - FFTW - float - 512 primesieve: 1e12 openvino: Person Vehicle Bike Detection FP16 - CPU tensorflow: CPU - 256 - GoogLeNet appleseed: Material Tester npb: FT.C compress-7zip: Compression Rating openvkl: vklBenchmark ISPC tensorflow: CPU - 512 - ResNet-50 openvino: Person Detection FP16 - CPU openvino: Person Detection FP32 - CPU tensorflow: CPU - 512 - GoogLeNet openvino: Age Gender Recognition Retail 0013 FP16 - CPU npb: CG.C build-linux-kernel: defconfig luxcorerender: DLSC - CPU deepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream deepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream deepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream deepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream deepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Stream oidn: RTLightmap.hdr.4096x4096 - CPU-Only deepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Stream deepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Stream mysqlslap: 2048 appleseed: Emily oidn: RT.ldr_alb_nrm.3840x2160 - CPU-Only openvino: Age Gender Recognition Retail 0013 FP16 - CPU aircrack-ng: oidn: RT.hdr_alb_nrm.3840x2160 - CPU-Only tensorflow: CPU - 256 - AlexNet mysqlslap: 4096 build-llvm: Ninja tensorflow: CPU - 512 - AlexNet build-nodejs: Time To Compile tensorflow: CPU - 256 - ResNet-50 minife: Small openvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPU appleseed: Disney Material build-gem5: Time To Compile build-llvm: Unix Makefiles build-godot: Time To Compile openvino: Weld Porosity Detection FP16-INT8 - CPU nekrs: Kershaw openvino: Vehicle Detection FP16-INT8 - CPU openvino: Vehicle Detection FP16-INT8 - CPU openvino: Vehicle Detection FP16 - CPU deepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Stream deepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Stream deepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Stream deepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Stream pgbench: 1000 - 800 - Read Only - Average Latency pgbench: 1000 - 800 - Read Only stockfish: Total Time luxcorerender: Rainbow Colors and Prism - CPU luxcorerender: Orange Juice - CPU srsran: PUSCH Processor Benchmark, Throughput Total xmrig: Monero - 1M nekrs: TurboPipe Periodic specfem3d: Homogeneous Halfspace specfem3d: Tomographic Model heffte: r2c - FFTW - double - 512 heffte: c2c - FFTW - double - 512 libxsmm: 128 minibude: OpenMP - BM1 minibude: OpenMP - BM1 npb: SP.B npb: EP.D EPYC 9754 1P EPYC 9754 2P SMT On SMT Off SMT On SMT Off 43.73 163633625553 3591 17.194492779 791787 417.1186 216115 216064 73.2077 968.6267 73.3807 624.4789 125.5813 316.0466 445912000 659346857987 157.6504 6.216379849 880249000 31.12 279662.55 1783700000 32972 16484 13759 5012.26 861 13963 12.00 1032 27885 462415837320 873 333445000 27538 39.27 810375 74803.6 8.1743 2378.07 582.08 2377.79 15.964931688 116.54 12.77 50.473 857890000 1890935.3 128129.56 1190.7549 32.6750 54195.1 20312667 30.8165 16.49 30.8507 31.9245 53005879330 1049.31 1169673735557 1012537766210 110.00 541.29 117.88 60.77 1376.2142 75.0831 11794.32 6067.78 0.20702 1696766667 6148.89 227.517 10.53 21.286 3331.7 4081 85400.88 5972.643 238.905 5300.29 292243.61 245.542 131909.91 12.18 128.124 1.944 10.40 504.09 166.676131 140791.52 726271 1396 122.77 26.61 26.57 416.03 120515.11 45686.88 26.225 16.34 859.8653 859.877 65.9747 102.2265 46.4318 1.74 201.6258 152.9895 780 122.587709 3.62 0.83 171120.354 3.62 1422.08 655 125.364 1628.80 113.704 118.45 51784.1 1.21 44.30492 161.648 211.079 105.797 10.84 5808636667 12.07 5311.48 1464.29 499.5866 126.9310 267.7088 240.8667 0.952 855569 365034349 20.88 24.47 8389.0 24409.4 2538406923 9.404348500 7.480674706 66.3418 34.8451 2713.4 237.763 5944.062 149355.54 13264.79 11.49 111414557280 5590 13.419454501 510470 504.3578 163263 163220 97.0913 1275.7707 96.9697 812.1023 85.2893 404.2338 493535000 550700307433 107.3608 5.030031580 928320000 38.44 289518.14 1414700000 43371 21666 18004 4957.686 1127 18269 9.27 1358 36539 392782689987 1146 363750000 35926 50.01 676218 63182.2 7.3147 1102.82 545.13 1105.96 15.028121929 147.22 15.15 57.956 893624000 1799293.0 136942.13 1278.6466 25.7178 56647.3 16751667 23.7210 20.47 23.7333 25.5121 51804333637 513.23 1152838928063 997034621883 58.71 268.82 119.01 62.18 1780.4421 68.3072 11710.59 5837.62 0.20595 1313933333 5118.29 177.758 5.47 21.132 3813.4 6242 71673.85 5903.205 236.128 5315.15 298801.44 248.557 133415.42 8.88 128.597 1.796 6.24 525.39 167.900597 147448.72 592741 1107 123.88 28.74 28.83 429.16 113162.29 48672.24 22.984 13.27 644.9768 645.0019 49.5111 77.8148 35.5549 1.72 156.0563 125.4467 783 123.298396 3.57 0.70 149056.953 3.57 1375.33 695 119.576 1526.81 116.122 121.47 51741.0 1.30 38.492263 148.484 213.907 102.588 10.91 5734266667 4.74 6744.66 2784.67 468.7809 134.5717 259.2779 246.3732 0.917 877722 272722940 14.37 20.98 20430.9 51218.9 2586083333 6.092750299 4.937265506 67.9907 35.4049 2696.5 234.684 5867.108 161475.24 14274.53 11.03 327926038513 2014 9.778381437 1353957 797.6248 409850 407860 139.8821 1868.3310 139.6590 1190.7168 210.0714 602.2602 960471000 1317549954027 255.9864 3.906675095 1724030000 16.28 591505.17 3330166667 18455 9253 7698 2143.513 482 7817 21.65 582 15731 909817360110 495 672541000 15619 21.75 1524533 142082.7 15.9305 1552.15 1159.99 1559.95 9.997447570 69.03 7.12 27.280 1571100000 3782091.8 249109.09 610.1137 53.8295 108490.5 34879333 49.2311 9.87 49.1328 52.7318 106049655203 526.98 2339890700107 2019317070327 55.12 270.93 235.65 121.03 2627.3142 134.8853 22954.44 11299.32 0.10646 2544400000 9889.61 145.929 5.64 11.120 6112.6 3321 133931.53 7888.085 315.524 9849.01 491231.83 433.601 224243.28 9.86 221.765 1.508 6.46 329.22 211432.89 925820 1720 172.36 40.67 40.90 538.52 168372.89 67554.74 20.344 18.61 902.8593 902.7792 68.3638 107.1835 48.6206 2.35 211.4848 159.8861 580 164.567164 4.78 0.62 4.76 1225.41 545 107.721 1770.91 93.271 124.98 62774.2 1.08 152.586 199.365 100.790 11.00 4.83 13214.20 5810.41 521.3108 242.9589 229.1624 556.9411 0.974 827878 582386924 19.02 34.45 17891.4 86533.7 4.727830148 3.782741798 207.197 109.645 4976.7 253.123 6328.069 236490.76 23705.40 10.26 222269428867 2976 6.218069257 913091 1058.6776 320885 317046 182.5974 2409.8035 182.4453 1541.9512 146.2051 770.1392 1075900000 1100453394570 178.3046 2.630122599 2078880000 20.24 658754.21 2610933333 24238 12127 10046 2363.579 631 10232 15.15 763 20545 784489570523 645 770180000 20133 27.47 1301500 100754.3 14.4099 1579.14 1174.99 1545.23 7.470772930 85.54 8.45 50.054 1815160000 3598946.3 268721.05 693.3961 44.7104 113251.4 30221667 41.5531 11.72 41.6567 44.2889 102232233590 526.46 2307524535457 1998234148863 54.42 271.24 235.37 121.08 2730.4980 127.2088 22955.64 11373.95 0.13969 2542633333 9878.88 118.099 5.61 11.152 6373.0 3671 118225.55 10989.938 439.597 8635.30 536518.74 430.265 231041.57 6.97 223.584 1.160 6.47 452.99 265.885808 224178.10 771462 1530 189.40 41.05 40.20 634.13 142135.58 66822.97 18.474 14.45 676.8128 676.4342 51.8103 81.1994 45.8765 2.09 162.0191 118.1285 591 159.906614 4.31 0.67 128050.219 4.35 1581.66 579 99.215 1908.76 93.952 146.74 53798.6 1.13 40.570691 148.376 198.750 100.240 11.10 4.86 13141.51 6242.68 426.7362 290.7562 212.6821 590.8789 1.018 785968 447023143 13.51 25.11 36573.8 85946.0 3.451830169 2.709205428 210.933 112.413 4505.5 439.067 10976.682 246272.79 25983.74 OpenBenchmarking.org
OpenVINO Model: Vehicle Detection FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Vehicle Detection FP16 - Device: CPU SMT Off SMT On 10 20 30 40 50 SE +/- 0.10, N = 13 SE +/- 0.55, N = 14 SE +/- 0.12, N = 13 SE +/- 0.15, N = 15 11.49 43.73 10.26 11.03 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenSSL Algorithm: SHA256 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.1 Algorithm: SHA256 SMT Off SMT On 70000M 140000M 210000M 280000M 350000M SE +/- 15184699.42, N = 3 SE +/- 161352968.33, N = 3 SE +/- 224959781.68, N = 3 SE +/- 207719324.78, N = 3 111414557280 163633625553 222269428867 327926038513 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
toyBrot Fractal Generator Implementation: TBB EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better toyBrot Fractal Generator 2020-11-18 Implementation: TBB SMT Off SMT On 1200 2400 3600 4800 6000 SE +/- 43.30, N = 15 SE +/- 21.75, N = 9 SE +/- 31.80, N = 15 SE +/- 23.04, N = 15 5590 3591 2976 2014 1. (CXX) g++ options: -O3 -lpthread -lm -lgcc -lgcc_s -lc
SPECFEM3D Model: Water-layered Halfspace EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.0 Model: Water-layered Halfspace SMT Off SMT On 4 8 12 16 20 SE +/- 0.021567327, N = 3 SE +/- 0.142353904, N = 3 SE +/- 0.062799204, N = 15 SE +/- 0.063515594, N = 4 13.419454501 17.194492779 6.218069257 9.778381437 1. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
7-Zip Compression Test: Decompression Rating EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 22.01 Test: Decompression Rating SMT Off SMT On 300K 600K 900K 1200K 1500K SE +/- 429.17, N = 3 SE +/- 1608.29, N = 3 SE +/- 1897.81, N = 3 SE +/- 5499.45, N = 3 510470 791787 913091 1353957 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
Neural Magic DeepSparse Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.5 Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Stream SMT Off SMT On 200 400 600 800 1000 SE +/- 5.07, N = 15 SE +/- 0.46, N = 3 SE +/- 0.90, N = 3 SE +/- 0.32, N = 3 504.36 417.12 1058.68 797.62
John The Ripper Test: Blowfish EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: Blowfish SMT Off SMT On 90K 180K 270K 360K 450K SE +/- 12.67, N = 3 SE +/- 117.40, N = 3 SE +/- 1247.49, N = 3 SE +/- 3726.56, N = 3 163263 216115 320885 409850 1. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
John The Ripper Test: bcrypt EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: bcrypt SMT Off SMT On 90K 180K 270K 360K 450K SE +/- 33.22, N = 3 SE +/- 92.54, N = 3 SE +/- 1964.45, N = 3 SE +/- 2298.36, N = 3 163220 216064 317046 407860 1. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
Neural Magic DeepSparse Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.5 Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream SMT Off SMT On 40 80 120 160 200 SE +/- 0.06, N = 3 SE +/- 0.25, N = 3 SE +/- 0.18, N = 3 SE +/- 0.08, N = 3 97.09 73.21 182.60 139.88
Neural Magic DeepSparse Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.5 Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream SMT Off SMT On 500 1000 1500 2000 2500 SE +/- 5.02, N = 3 SE +/- 0.74, N = 3 SE +/- 1.63, N = 3 SE +/- 1.60, N = 3 1275.77 968.63 2409.80 1868.33
Neural Magic DeepSparse Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.5 Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream SMT Off SMT On 40 80 120 160 200 SE +/- 0.09, N = 3 SE +/- 0.25, N = 3 SE +/- 0.07, N = 3 SE +/- 0.15, N = 3 96.97 73.38 182.45 139.66
Neural Magic DeepSparse Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.5 Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream SMT Off SMT On 300 600 900 1200 1500 SE +/- 3.93, N = 3 SE +/- 0.29, N = 3 SE +/- 1.38, N = 3 SE +/- 1.02, N = 3 812.10 624.48 1541.95 1190.72
Embree Binary: Pathtracer ISPC - Model: Crown EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.1 Binary: Pathtracer ISPC - Model: Crown SMT Off SMT On 50 100 150 200 250 SE +/- 0.05, N = 6 SE +/- 0.13, N = 7 SE +/- 0.13, N = 7 SE +/- 0.16, N = 9 85.29 125.58 146.21 210.07
Neural Magic DeepSparse Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.5 Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream SMT Off SMT On 170 340 510 680 850 SE +/- 2.15, N = 3 SE +/- 1.02, N = 3 SE +/- 0.92, N = 3 SE +/- 0.72, N = 3 404.23 316.05 770.14 602.26
Graph500 Scale: 26 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org sssp max_TEPS, More Is Better Graph500 3.0 Scale: 26 SMT Off SMT On 200M 400M 600M 800M 1000M 493535000 445912000 1075900000 960471000 1. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi
OpenSSL Algorithm: ChaCha20 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.1 Algorithm: ChaCha20 SMT Off SMT On 300000M 600000M 900000M 1200000M 1500000M SE +/- 70893114.79, N = 3 SE +/- 23916253.09, N = 3 SE +/- 107897909.37, N = 3 SE +/- 46014499.85, N = 3 550700307433 659346857987 1100453394570 1317549954027 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
Embree Binary: Pathtracer ISPC - Model: Asian Dragon EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.1 Binary: Pathtracer ISPC - Model: Asian Dragon SMT Off SMT On 60 120 180 240 300 SE +/- 0.09, N = 6 SE +/- 0.09, N = 8 SE +/- 0.26, N = 8 SE +/- 0.48, N = 9 107.36 157.65 178.30 255.99
SPECFEM3D Model: Mount St. Helens EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.0 Model: Mount St. Helens SMT Off SMT On 2 4 6 8 10 SE +/- 0.069356320, N = 12 SE +/- 0.034771470, N = 5 SE +/- 0.013893024, N = 5 SE +/- 0.014293172, N = 5 5.030031580 6.216379849 2.630122599 3.906675095 1. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
Graph500 Scale: 26 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org bfs max_TEPS, More Is Better Graph500 3.0 Scale: 26 SMT Off SMT On 400M 800M 1200M 1600M 2000M 928320000 880249000 2078880000 1724030000 1. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi
Blender Blend File: Classroom - Compute: CPU-Only EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: Classroom - Compute: CPU-Only SMT Off SMT On 9 18 27 36 45 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 38.44 31.12 20.24 16.28
NAS Parallel Benchmarks Test / Class: LU.C EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C SMT Off SMT On 140K 280K 420K 560K 700K SE +/- 1485.96, N = 6 SE +/- 2132.06, N = 6 SE +/- 4916.03, N = 15 SE +/- 7199.59, N = 15 289518.14 279662.55 658754.21 591505.17 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
Liquid-DSP Threads: 512 - Buffer Length: 256 - Filter Length: 512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 512 - Buffer Length: 256 - Filter Length: 512 SMT Off SMT On 700M 1400M 2100M 2800M 3500M SE +/- 2946183.97, N = 3 SE +/- 1814754.35, N = 3 SE +/- 3555434.03, N = 3 SE +/- 569600.25, N = 3 1414700000 1783700000 2610933333 3330166667 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OSPRay Studio Camera: 3 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 0.11 Camera: 3 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer SMT Off SMT On 9K 18K 27K 36K 45K SE +/- 76.61, N = 3 SE +/- 17.58, N = 3 SE +/- 17.91, N = 3 SE +/- 34.44, N = 3 43371 32972 24238 18455 1. (CXX) g++ options: -O3 -lm -ldl
OSPRay Studio Camera: 3 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 0.11 Camera: 3 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer SMT Off SMT On 5K 10K 15K 20K 25K SE +/- 15.65, N = 3 SE +/- 11.89, N = 3 SE +/- 8.65, N = 3 SE +/- 27.17, N = 3 21666 16484 12127 9253 1. (CXX) g++ options: -O3 -lm -ldl
OSPRay Studio Camera: 1 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 0.11 Camera: 1 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer SMT Off SMT On 4K 8K 12K 16K 20K SE +/- 18.26, N = 3 SE +/- 5.78, N = 3 SE +/- 14.52, N = 3 SE +/- 7.80, N = 3 18004 13759 10046 7698 1. (CXX) g++ options: -O3 -lm -ldl
CP2K Molecular Dynamics Input: H2O-DFT-LS EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better CP2K Molecular Dynamics 2023.1 Input: H2O-DFT-LS SMT Off SMT On 1100 2200 3300 4400 5500 4957.69 5012.26 2363.58 2143.51 1. (F9X) gfortran options: -fopenmp -mtune=native -O3 -funroll-loops -fbacktrace -ffree-form -fimplicit-none -std=f2008 -lcp2kstart -lcp2kmc -lcp2kswarm -lcp2kmotion -lcp2kthermostat -lcp2kemd -lcp2ktmc -lcp2kmain -lcp2kdbt -lcp2ktas -lcp2kdbm -lcp2kgrid -lcp2kgridcpu -lcp2kgridref -lcp2kgridcommon -ldbcsrarnoldi -ldbcsrx -lcp2kshg_int -lcp2keri_mme -lcp2kminimax -lcp2khfxbase -lcp2ksubsys -lcp2kxc -lcp2kao -lcp2kpw_env -lcp2kinput -lcp2kpw -lcp2kgpu -lcp2kfft -lcp2kfpga -lcp2kfm -lcp2kcommon -lcp2koffload -lcp2kmpiwrap -lcp2kbase -ldbcsr -lsirius -lspla -lspfft -lsymspg -lvdwxc -lhdf5 -lhdf5_hl -lz -lgsl -lelpa_openmp -lcosma -lcosta -lscalapack -lxsmmf -lxsmm -ldl -lpthread -lxcf03 -lxc -lint2 -lfftw3_mpi -lfftw3 -lfftw3_omp -lmpi_cxx -lmpi -lopenblas -lvori -lstdc++ -lmpi_usempif08 -lmpi_mpifh -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm
OSPRay Studio Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 0.11 Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer SMT Off SMT On 200 400 600 800 1000 SE +/- 1.20, N = 3 SE +/- 0.88, N = 3 SE +/- 0.58, N = 3 SE +/- 2.85, N = 3 1127 861 631 482 1. (CXX) g++ options: -O3 -lm -ldl
OSPRay Studio Camera: 2 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 0.11 Camera: 2 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer SMT Off SMT On 4K 8K 12K 16K 20K SE +/- 11.93, N = 3 SE +/- 6.12, N = 3 SE +/- 5.29, N = 3 SE +/- 19.63, N = 3 18269 13963 10232 7817 1. (CXX) g++ options: -O3 -lm -ldl
CloverLeaf Lagrangian-Eulerian Hydrodynamics EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better CloverLeaf Lagrangian-Eulerian Hydrodynamics SMT Off SMT On 5 10 15 20 25 SE +/- 0.09, N = 5 SE +/- 0.11, N = 4 SE +/- 0.04, N = 4 SE +/- 0.27, N = 4 9.27 12.00 15.15 21.65 1. (F9X) gfortran options: -O3 -march=native -funroll-loops -fopenmp
OSPRay Studio Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 0.11 Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer SMT Off SMT On 300 600 900 1200 1500 SE +/- 0.33, N = 3 SE +/- 0.58, N = 3 SE +/- 0.88, N = 3 SE +/- 0.33, N = 3 1358 1032 763 582 1. (CXX) g++ options: -O3 -lm -ldl
OSPRay Studio Camera: 2 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 0.11 Camera: 2 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer SMT Off SMT On 8K 16K 24K 32K 40K SE +/- 53.62, N = 3 SE +/- 21.53, N = 3 SE +/- 96.56, N = 3 SE +/- 41.40, N = 3 36539 27885 20545 15731 1. (CXX) g++ options: -O3 -lm -ldl
OpenSSL Algorithm: ChaCha20-Poly1305 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.1 Algorithm: ChaCha20-Poly1305 SMT Off SMT On 200000M 400000M 600000M 800000M 1000000M SE +/- 86138373.27, N = 3 SE +/- 15599558.91, N = 3 SE +/- 39882779.79, N = 3 SE +/- 267574958.87, N = 3 392782689987 462415837320 784489570523 909817360110 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OSPRay Studio Camera: 2 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 0.11 Camera: 2 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer SMT Off SMT On 200 400 600 800 1000 SE +/- 1.53, N = 3 SE +/- 0.88, N = 3 SE +/- 0.33, N = 3 SE +/- 1.45, N = 3 1146 873 645 495 1. (CXX) g++ options: -O3 -lm -ldl
Graph500 Scale: 26 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org sssp median_TEPS, More Is Better Graph500 3.0 Scale: 26 SMT Off SMT On 160M 320M 480M 640M 800M 363750000 333445000 770180000 672541000 1. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi
OSPRay Studio Camera: 1 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 0.11 Camera: 1 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer SMT Off SMT On 8K 16K 24K 32K 40K SE +/- 29.21, N = 3 SE +/- 10.48, N = 3 SE +/- 38.84, N = 3 SE +/- 64.83, N = 3 35926 27538 20133 15619 1. (CXX) g++ options: -O3 -lm -ldl
Blender Blend File: Pabellon Barcelona - Compute: CPU-Only EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: Pabellon Barcelona - Compute: CPU-Only SMT Off SMT On 11 22 33 44 55 SE +/- 0.14, N = 3 SE +/- 0.08, N = 3 SE +/- 0.09, N = 3 SE +/- 0.08, N = 3 50.01 39.27 27.47 21.75
John The Ripper Test: WPA PSK EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: WPA PSK SMT Off SMT On 300K 600K 900K 1200K 1500K SE +/- 6933.79, N = 3 SE +/- 505.98, N = 3 SE +/- 15887.63, N = 4 SE +/- 18163.51, N = 15 676218 810375 1301500 1524533 1. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
Xmrig Variant: Wownero - Hash Count: 1M EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org H/s, More Is Better Xmrig 6.18.1 Variant: Wownero - Hash Count: 1M SMT Off SMT On 30K 60K 90K 120K 150K SE +/- 13.77, N = 4 SE +/- 513.11, N = 15 SE +/- 741.13, N = 4 SE +/- 677.82, N = 5 63182.2 74803.6 100754.3 142082.7 1. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
ASTC Encoder Preset: Exhaustive EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 4.0 Preset: Exhaustive SMT Off SMT On 4 8 12 16 20 SE +/- 0.0028, N = 5 SE +/- 0.0006, N = 5 SE +/- 0.0257, N = 6 SE +/- 0.0081, N = 6 7.3147 8.1743 14.4099 15.9305 1. (CXX) g++ options: -O3 -flto -pthread
OpenVINO Model: Person Detection FP32 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Person Detection FP32 - Device: CPU SMT Off SMT On 500 1000 1500 2000 2500 SE +/- 10.77, N = 5 SE +/- 12.47, N = 15 SE +/- 4.61, N = 3 SE +/- 8.82, N = 3 1102.82 2378.07 1579.14 1552.15 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenVINO Model: Machine Translation EN To DE FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Machine Translation EN To DE FP16 - Device: CPU SMT Off SMT On 300 600 900 1200 1500 SE +/- 4.46, N = 15 SE +/- 6.51, N = 15 SE +/- 4.51, N = 3 SE +/- 7.44, N = 3 545.13 582.08 1174.99 1159.99 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenVINO Model: Person Detection FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Person Detection FP16 - Device: CPU SMT Off SMT On 500 1000 1500 2000 2500 SE +/- 8.36, N = 9 SE +/- 16.41, N = 12 SE +/- 10.43, N = 3 SE +/- 4.07, N = 3 1105.96 2377.79 1545.23 1559.95 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
SPECFEM3D Model: Layered Halfspace EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.0 Model: Layered Halfspace SMT Off SMT On 4 8 12 16 20 SE +/- 0.159428812, N = 3 SE +/- 0.034780206, N = 3 SE +/- 0.056354216, N = 4 SE +/- 0.066481426, N = 15 15.028121929 15.964931688 7.470772930 9.997447570 1. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
Blender Blend File: Barbershop - Compute: CPU-Only EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: Barbershop - Compute: CPU-Only SMT Off SMT On 30 60 90 120 150 SE +/- 0.17, N = 3 SE +/- 0.11, N = 3 SE +/- 0.08, N = 3 SE +/- 0.19, N = 3 147.22 116.54 85.54 69.03
Blender Blend File: BMW27 - Compute: CPU-Only EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: BMW27 - Compute: CPU-Only SMT Off SMT On 4 8 12 16 20 SE +/- 0.05, N = 4 SE +/- 0.02, N = 4 SE +/- 0.05, N = 5 SE +/- 0.02, N = 6 15.15 12.77 8.45 7.12
Helsing Digit Range: 14 digit EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Helsing 1.0-beta Digit Range: 14 digit SMT Off SMT On 13 26 39 52 65 SE +/- 0.03, N = 3 SE +/- 0.09, N = 3 SE +/- 0.31, N = 3 SE +/- 0.37, N = 3 57.96 50.47 50.05 27.28 1. (CC) gcc options: -O2 -pthread
Graph500 Scale: 26 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org bfs median_TEPS, More Is Better Graph500 3.0 Scale: 26 SMT Off SMT On 400M 800M 1200M 1600M 2000M 893624000 857890000 1815160000 1571100000 1. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi
OpenSSL Algorithm: RSA4096 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org verify/s, More Is Better OpenSSL 3.1 Algorithm: RSA4096 SMT Off SMT On 800K 1600K 2400K 3200K 4000K SE +/- 1031.37, N = 3 SE +/- 405.88, N = 3 SE +/- 1067.40, N = 3 SE +/- 163.02, N = 3 1799293.0 1890935.3 3598946.3 3782091.8 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
NAS Parallel Benchmarks Test / Class: MG.C EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: MG.C SMT Off SMT On 60K 120K 180K 240K 300K SE +/- 104.98, N = 10 SE +/- 296.36, N = 10 SE +/- 599.41, N = 10 SE +/- 1284.96, N = 11 136942.13 128129.56 268721.05 249109.09 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
ASTC Encoder Preset: Fast EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 4.0 Preset: Fast SMT Off SMT On 300 600 900 1200 1500 SE +/- 1.20, N = 7 SE +/- 1.79, N = 6 SE +/- 1.25, N = 5 SE +/- 1.29, N = 5 1278.65 1190.75 693.40 610.11 1. (CXX) g++ options: -O3 -flto -pthread
OSPRay Benchmark: gravity_spheres_volume/dim_512/ao/real_time EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Items Per Second, More Is Better OSPRay 2.12 Benchmark: gravity_spheres_volume/dim_512/ao/real_time SMT Off SMT On 12 24 36 48 60 SE +/- 0.01, N = 3 SE +/- 0.07, N = 3 SE +/- 0.02, N = 3 SE +/- 0.09, N = 3 25.72 32.68 44.71 53.83
OpenSSL Algorithm: RSA4096 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org sign/s, More Is Better OpenSSL 3.1 Algorithm: RSA4096 SMT Off SMT On 20K 40K 60K 80K 100K SE +/- 1.80, N = 3 SE +/- 16.66, N = 3 SE +/- 3.85, N = 3 SE +/- 3.93, N = 3 56647.3 54195.1 113251.4 108490.5 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
John The Ripper Test: MD5 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: MD5 SMT Off SMT On 7M 14M 21M 28M 35M SE +/- 35950.58, N = 3 SE +/- 52818.35, N = 3 SE +/- 140717.61, N = 3 SE +/- 83819.91, N = 3 16751667 20312667 30221667 34879333 1. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
OSPRay Benchmark: particle_volume/scivis/real_time EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Items Per Second, More Is Better OSPRay 2.12 Benchmark: particle_volume/scivis/real_time SMT Off SMT On 11 22 33 44 55 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 23.72 30.82 41.55 49.23
Blender Blend File: Fishy Cat - Compute: CPU-Only EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: Fishy Cat - Compute: CPU-Only SMT Off SMT On 5 10 15 20 25 SE +/- 0.06, N = 3 SE +/- 0.06, N = 3 SE +/- 0.04, N = 4 SE +/- 0.02, N = 5 20.47 16.49 11.72 9.87
OSPRay Benchmark: particle_volume/ao/real_time EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Items Per Second, More Is Better OSPRay 2.12 Benchmark: particle_volume/ao/real_time SMT Off SMT On 11 22 33 44 55 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 23.73 30.85 41.66 49.13
OSPRay Benchmark: gravity_spheres_volume/dim_512/scivis/real_time EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Items Per Second, More Is Better OSPRay 2.12 Benchmark: gravity_spheres_volume/dim_512/scivis/real_time SMT Off SMT On 12 24 36 48 60 SE +/- 0.03, N = 3 SE +/- 0.09, N = 3 SE +/- 0.07, N = 3 SE +/- 0.03, N = 3 25.51 31.92 44.29 52.73
OpenSSL Algorithm: SHA512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.1 Algorithm: SHA512 SMT Off SMT On 20000M 40000M 60000M 80000M 100000M SE +/- 19506083.54, N = 3 SE +/- 4276543.39, N = 3 SE +/- 660141733.06, N = 3 SE +/- 22577791.79, N = 3 51804333637 53005879330 102232233590 106049655203 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenVINO Model: Face Detection FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Face Detection FP16 - Device: CPU SMT Off SMT On 200 400 600 800 1000 SE +/- 0.07, N = 3 SE +/- 0.05, N = 3 SE +/- 0.28, N = 3 SE +/- 0.18, N = 3 513.23 1049.31 526.46 526.98 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenSSL Algorithm: AES-128-GCM EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.1 Algorithm: AES-128-GCM SMT Off SMT On 500000M 1000000M 1500000M 2000000M 2500000M SE +/- 772883363.35, N = 3 SE +/- 404585301.69, N = 3 SE +/- 4718005576.14, N = 3 SE +/- 4494053056.14, N = 3 1152838928063 1169673735557 2307524535457 2339890700107 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenSSL Algorithm: AES-256-GCM EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.1 Algorithm: AES-256-GCM SMT Off SMT On 400000M 800000M 1200000M 1600000M 2000000M SE +/- 917614258.55, N = 3 SE +/- 2018584981.53, N = 3 SE +/- 2904032106.06, N = 3 SE +/- 676471418.80, N = 3 997034621883 1012537766210 1998234148863 2019317070327 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenVINO Model: Machine Translation EN To DE FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Machine Translation EN To DE FP16 - Device: CPU SMT Off SMT On 20 40 60 80 100 SE +/- 0.45, N = 15 SE +/- 1.12, N = 15 SE +/- 0.21, N = 3 SE +/- 0.35, N = 3 58.71 110.00 54.42 55.12 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenVINO Model: Face Detection FP16-INT8 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Face Detection FP16-INT8 - Device: CPU SMT Off SMT On 120 240 360 480 600 SE +/- 0.06, N = 3 SE +/- 0.09, N = 3 SE +/- 0.06, N = 3 SE +/- 0.07, N = 3 268.82 541.29 271.24 270.93 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenVINO Model: Face Detection FP16-INT8 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Face Detection FP16-INT8 - Device: CPU SMT Off SMT On 50 100 150 200 250 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.08, N = 3 SE +/- 0.09, N = 3 119.01 117.88 235.37 235.65 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenVINO Model: Face Detection FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Face Detection FP16 - Device: CPU SMT Off SMT On 30 60 90 120 150 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 62.18 60.77 121.08 121.03 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
Neural Magic DeepSparse Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.5 Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Stream SMT Off SMT On 600 1200 1800 2400 3000 SE +/- 3.84, N = 3 SE +/- 1.21, N = 3 SE +/- 33.47, N = 15 SE +/- 6.95, N = 3 1780.44 1376.21 2730.50 2627.31
ASTC Encoder Preset: Thorough EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 4.0 Preset: Thorough SMT Off SMT On 30 60 90 120 150 SE +/- 0.01, N = 6 SE +/- 0.02, N = 6 SE +/- 0.05, N = 6 SE +/- 0.22, N = 6 68.31 75.08 127.21 134.89 1. (CXX) g++ options: -O3 -flto -pthread
OpenVINO Model: Weld Porosity Detection FP16-INT8 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Weld Porosity Detection FP16-INT8 - Device: CPU SMT Off SMT On 5K 10K 15K 20K 25K SE +/- 13.11, N = 3 SE +/- 2.13, N = 3 SE +/- 15.58, N = 3 SE +/- 16.03, N = 3 11710.59 11794.32 22955.64 22954.44 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenVINO Model: Weld Porosity Detection FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Weld Porosity Detection FP16 - Device: CPU SMT Off SMT On 2K 4K 6K 8K 10K SE +/- 11.25, N = 3 SE +/- 0.58, N = 3 SE +/- 2.58, N = 3 SE +/- 9.34, N = 3 5837.62 6067.78 11373.95 11299.32 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
NAMD ATPase Simulation - 327,506 Atoms EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org days/ns, Fewer Is Better NAMD 2.14 ATPase Simulation - 327,506 Atoms SMT Off SMT On 0.0466 0.0932 0.1398 0.1864 0.233 SE +/- 0.00018, N = 4 SE +/- 0.00095, N = 4 SE +/- 0.00135, N = 5 SE +/- 0.00040, N = 3 0.20595 0.20702 0.13969 0.10646
Liquid-DSP Threads: 256 - Buffer Length: 256 - Filter Length: 512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 256 - Buffer Length: 256 - Filter Length: 512 SMT Off SMT On 500M 1000M 1500M 2000M 2500M SE +/- 592546.29, N = 3 SE +/- 470224.53, N = 3 SE +/- 1068228.02, N = 3 SE +/- 1877054.43, N = 3 1313933333 1696766667 2542633333 2544400000 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenVINO Model: Person Vehicle Bike Detection FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Person Vehicle Bike Detection FP16 - Device: CPU SMT Off SMT On 2K 4K 6K 8K 10K SE +/- 5.00, N = 3 SE +/- 48.86, N = 15 SE +/- 10.15, N = 3 SE +/- 11.31, N = 3 5118.29 6148.89 9878.88 9889.61 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
Timed Linux Kernel Compilation Build: allmodconfig EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Timed Linux Kernel Compilation 6.1 Build: allmodconfig SMT Off SMT On 50 100 150 200 250 SE +/- 0.48, N = 3 SE +/- 1.35, N = 3 SE +/- 0.55, N = 3 SE +/- 0.51, N = 3 177.76 227.52 118.10 145.93
OpenVINO Model: Weld Porosity Detection FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Weld Porosity Detection FP16 - Device: CPU SMT Off SMT On 3 6 9 12 15 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 5.47 10.53 5.61 5.64 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
Primesieve Length: 1e13 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Primesieve 8.0 Length: 1e13 SMT Off SMT On 5 10 15 20 25 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 5 SE +/- 0.01, N = 5 21.13 21.29 11.15 11.12 1. (CXX) g++ options: -O3
libxsmm M N K: 256 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFLOPS/s, More Is Better libxsmm 2-1.17-3645 M N K: 256 SMT Off SMT On 1400 2800 4200 5600 7000 SE +/- 16.75, N = 3 SE +/- 2.32, N = 3 SE +/- 66.66, N = 9 SE +/- 1.43, N = 3 3813.4 3331.7 6373.0 6112.6 1. (CXX) g++ options: -dynamic -Bstatic -static-libgcc -lgomp -lm -lrt -ldl -lquadmath -lstdc++ -pthread -fPIC -std=c++14 -O2 -fopenmp-simd -funroll-loops -ftree-vectorize -fdata-sections -ffunction-sections -fvisibility=hidden -msse4.2
toyBrot Fractal Generator Implementation: OpenMP EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better toyBrot Fractal Generator 2020-11-18 Implementation: OpenMP SMT Off SMT On 1300 2600 3900 5200 6500 SE +/- 0.20, N = 7 SE +/- 17.08, N = 8 SE +/- 53.03, N = 15 SE +/- 23.29, N = 12 6242 4081 3671 3321 1. (CXX) g++ options: -O3 -lpthread -lm -lgcc -lgcc_s -lc
OpenVINO Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU SMT Off SMT On 30K 60K 90K 120K 150K SE +/- 126.57, N = 3 SE +/- 192.97, N = 3 SE +/- 420.06, N = 3 SE +/- 602.14, N = 3 71673.85 85400.88 118225.55 133931.53 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
miniBUDE Implementation: OpenMP - Input Deck: BM2 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFInst/s, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM2 SMT Off SMT On 2K 4K 6K 8K 10K SE +/- 6.13, N = 3 SE +/- 0.21, N = 3 SE +/- 150.60, N = 12 SE +/- 10.75, N = 4 5903.21 5972.64 10989.94 7888.09 1. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
miniBUDE Implementation: OpenMP - Input Deck: BM2 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Billion Interactions/s, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM2 SMT Off SMT On 100 200 300 400 500 SE +/- 0.25, N = 3 SE +/- 0.01, N = 3 SE +/- 6.02, N = 12 SE +/- 0.43, N = 4 236.13 238.91 439.60 315.52 1. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
NAS Parallel Benchmarks Test / Class: IS.D EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: IS.D SMT Off SMT On 2K 4K 6K 8K 10K SE +/- 29.88, N = 6 SE +/- 27.01, N = 6 SE +/- 105.10, N = 15 SE +/- 104.60, N = 15 5315.15 5300.29 8635.30 9849.01 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
NAS Parallel Benchmarks Test / Class: BT.C EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: BT.C SMT Off SMT On 110K 220K 330K 440K 550K SE +/- 269.21, N = 5 SE +/- 396.74, N = 5 SE +/- 3792.90, N = 12 SE +/- 4903.75, N = 15 298801.44 292243.61 536518.74 491231.83 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
HeFFTe - Highly Efficient FFT for Exascale Test: r2c - Backend: FFTW - Precision: float - X Y Z: 512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFLOP/s, More Is Better HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: r2c - Backend: FFTW - Precision: float - X Y Z: 512 SMT Off SMT On 90 180 270 360 450 SE +/- 0.09, N = 5 SE +/- 0.52, N = 5 SE +/- 0.77, N = 6 SE +/- 0.86, N = 7 248.56 245.54 430.27 433.60 1. (CXX) g++ options: -O3
NAS Parallel Benchmarks Test / Class: SP.C EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.C SMT Off SMT On 50K 100K 150K 200K 250K SE +/- 228.86, N = 4 SE +/- 290.95, N = 4 SE +/- 1880.70, N = 6 SE +/- 1293.24, N = 6 133415.42 131909.91 231041.57 224243.28 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
LuxCoreRender Scene: LuxCore Benchmark - Acceleration: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org M samples/sec, More Is Better LuxCoreRender 2.6 Scene: LuxCore Benchmark - Acceleration: CPU SMT Off SMT On 3 6 9 12 15 SE +/- 0.08, N = 8 SE +/- 0.09, N = 15 SE +/- 0.11, N = 12 SE +/- 0.11, N = 15 8.88 12.18 6.97 9.86
HeFFTe - Highly Efficient FFT for Exascale Test: c2c - Backend: FFTW - Precision: float - X Y Z: 512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFLOP/s, More Is Better HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: c2c - Backend: FFTW - Precision: float - X Y Z: 512 SMT Off SMT On 50 100 150 200 250 SE +/- 0.01, N = 4 SE +/- 0.15, N = 4 SE +/- 1.28, N = 5 SE +/- 1.37, N = 5 128.60 128.12 223.58 221.77 1. (CXX) g++ options: -O3
Primesieve Length: 1e12 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Primesieve 8.0 Length: 1e12 SMT Off SMT On 0.4374 0.8748 1.3122 1.7496 2.187 SE +/- 0.003, N = 11 SE +/- 0.006, N = 11 SE +/- 0.003, N = 12 SE +/- 0.010, N = 14 1.796 1.944 1.160 1.508 1. (CXX) g++ options: -O3
OpenVINO Model: Person Vehicle Bike Detection FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Person Vehicle Bike Detection FP16 - Device: CPU SMT Off SMT On 3 6 9 12 15 SE +/- 0.01, N = 3 SE +/- 0.08, N = 15 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 6.24 10.40 6.47 6.46 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
TensorFlow Device: CPU - Batch Size: 256 - Model: GoogLeNet EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 256 - Model: GoogLeNet SMT Off SMT On 110 220 330 440 550 SE +/- 0.35, N = 3 SE +/- 3.39, N = 3 SE +/- 5.52, N = 4 SE +/- 1.69, N = 3 525.39 504.09 452.99 329.22
Appleseed Scene: Material Tester EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Appleseed 2.0 Beta Scene: Material Tester SMT Off SMT On 60 120 180 240 300 167.90 166.68 265.89
NAS Parallel Benchmarks Test / Class: FT.C EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: FT.C SMT Off SMT On 50K 100K 150K 200K 250K SE +/- 93.32, N = 8 SE +/- 1029.98, N = 8 SE +/- 2978.54, N = 13 SE +/- 1757.80, N = 9 147448.72 140791.52 224178.10 211432.89 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
7-Zip Compression Test: Compression Rating EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 22.01 Test: Compression Rating SMT Off SMT On 200K 400K 600K 800K 1000K SE +/- 4792.89, N = 3 SE +/- 1345.49, N = 3 SE +/- 4721.80, N = 3 SE +/- 5539.49, N = 3 592741 726271 771462 925820 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
OpenVKL Benchmark: vklBenchmark ISPC EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Items / Sec, More Is Better OpenVKL 1.3.1 Benchmark: vklBenchmark ISPC SMT Off SMT On 400 800 1200 1600 2000 SE +/- 0.33, N = 3 SE +/- 1.86, N = 3 SE +/- 1.33, N = 3 SE +/- 6.06, N = 3 1107 1396 1530 1720
TensorFlow Device: CPU - Batch Size: 512 - Model: ResNet-50 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 512 - Model: ResNet-50 SMT Off SMT On 40 80 120 160 200 SE +/- 1.35, N = 3 SE +/- 0.90, N = 3 SE +/- 0.13, N = 3 SE +/- 0.59, N = 3 123.88 122.77 189.40 172.36
OpenVINO Model: Person Detection FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Person Detection FP16 - Device: CPU SMT Off SMT On 9 18 27 36 45 SE +/- 0.23, N = 9 SE +/- 0.20, N = 12 SE +/- 0.29, N = 3 SE +/- 0.14, N = 3 28.74 26.61 41.05 40.67 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenVINO Model: Person Detection FP32 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Person Detection FP32 - Device: CPU SMT Off SMT On 9 18 27 36 45 SE +/- 0.29, N = 5 SE +/- 0.15, N = 15 SE +/- 0.14, N = 3 SE +/- 0.22, N = 3 28.83 26.57 40.20 40.90 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
TensorFlow Device: CPU - Batch Size: 512 - Model: GoogLeNet EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 512 - Model: GoogLeNet SMT Off SMT On 140 280 420 560 700 SE +/- 5.79, N = 12 SE +/- 4.74, N = 12 SE +/- 6.73, N = 3 SE +/- 4.51, N = 15 429.16 416.03 634.13 538.52
OpenVINO Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU SMT Off SMT On 40K 80K 120K 160K 200K SE +/- 202.90, N = 3 SE +/- 390.62, N = 3 SE +/- 158.47, N = 3 SE +/- 676.62, N = 3 113162.29 120515.11 142135.58 168372.89 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
NAS Parallel Benchmarks Test / Class: CG.C EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: CG.C SMT Off SMT On 14K 28K 42K 56K 70K SE +/- 285.31, N = 8 SE +/- 441.00, N = 15 SE +/- 568.39, N = 8 SE +/- 348.18, N = 8 48672.24 45686.88 66822.97 67554.74 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
Timed Linux Kernel Compilation Build: defconfig EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Timed Linux Kernel Compilation 6.1 Build: defconfig SMT Off SMT On 6 12 18 24 30 SE +/- 0.21, N = 7 SE +/- 0.23, N = 7 SE +/- 0.12, N = 13 SE +/- 0.14, N = 13 22.98 26.23 18.47 20.34
LuxCoreRender Scene: DLSC - Acceleration: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org M samples/sec, More Is Better LuxCoreRender 2.6 Scene: DLSC - Acceleration: CPU SMT Off SMT On 5 10 15 20 25 SE +/- 0.05, N = 3 SE +/- 0.10, N = 3 SE +/- 0.10, N = 3 SE +/- 0.20, N = 4 13.27 16.34 14.45 18.61
Neural Magic DeepSparse Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.5 Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream SMT Off SMT On 200 400 600 800 1000 SE +/- 0.16, N = 3 SE +/- 0.14, N = 3 SE +/- 0.02, N = 3 SE +/- 0.27, N = 3 644.98 859.87 676.81 902.86
Neural Magic DeepSparse Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.5 Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream SMT Off SMT On 200 400 600 800 1000 SE +/- 0.12, N = 3 SE +/- 0.49, N = 3 SE +/- 0.21, N = 3 SE +/- 0.45, N = 3 645.00 859.88 676.43 902.78
Neural Magic DeepSparse Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.5 Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream SMT Off SMT On 15 30 45 60 75 SE +/- 0.20, N = 3 SE +/- 0.05, N = 3 SE +/- 0.04, N = 3 SE +/- 0.06, N = 3 49.51 65.97 51.81 68.36
Neural Magic DeepSparse Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.5 Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream SMT Off SMT On 20 40 60 80 100 SE +/- 0.37, N = 3 SE +/- 0.04, N = 3 SE +/- 0.05, N = 3 SE +/- 0.07, N = 3 77.81 102.23 81.20 107.18
Neural Magic DeepSparse Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.5 Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Stream SMT Off SMT On 11 22 33 44 55 SE +/- 0.10, N = 3 SE +/- 0.04, N = 3 SE +/- 0.57, N = 15 SE +/- 0.13, N = 3 35.55 46.43 45.88 48.62
Intel Open Image Denoise Run: RTLightmap.hdr.4096x4096 - Device: CPU-Only EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 2.0 Run: RTLightmap.hdr.4096x4096 - Device: CPU-Only SMT Off SMT On 0.5288 1.0576 1.5864 2.1152 2.644 SE +/- 0.01, N = 4 SE +/- 0.00, N = 4 SE +/- 0.02, N = 5 SE +/- 0.00, N = 5 1.72 1.74 2.09 2.35
Neural Magic DeepSparse Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.5 Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream SMT Off SMT On 50 100 150 200 250 SE +/- 0.81, N = 3 SE +/- 0.54, N = 3 SE +/- 0.09, N = 3 SE +/- 0.18, N = 3 156.06 201.63 162.02 211.48
Neural Magic DeepSparse Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.5 Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Stream SMT Off SMT On 40 80 120 160 200 SE +/- 1.18, N = 15 SE +/- 0.16, N = 3 SE +/- 0.10, N = 3 SE +/- 0.05, N = 3 125.45 152.99 118.13 159.89
MariaDB Clients: 2048 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Queries Per Second, More Is Better MariaDB 11.0.1 Clients: 2048 SMT Off SMT On 200 400 600 800 1000 SE +/- 1.06, N = 3 SE +/- 1.81, N = 3 SE +/- 0.73, N = 3 SE +/- 8.04, N = 3 783 780 591 580 1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lnuma -lpcre2-8 -lcrypt -lz -lm -lssl -lcrypto -lpthread -ldl
Appleseed Scene: Emily EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Appleseed 2.0 Beta Scene: Emily SMT Off SMT On 40 80 120 160 200 123.30 122.59 159.91 164.57
Intel Open Image Denoise Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-Only EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 2.0 Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-Only SMT Off SMT On 1.0755 2.151 3.2265 4.302 5.3775 SE +/- 0.02, N = 15 SE +/- 0.00, N = 7 SE +/- 0.03, N = 15 SE +/- 0.01, N = 7 3.57 3.62 4.31 4.78
OpenVINO Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU SMT Off SMT On 0.1868 0.3736 0.5604 0.7472 0.934 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 0.70 0.83 0.67 0.62 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
Aircrack-ng EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org k/s, More Is Better Aircrack-ng 1.7 SMT Off SMT On 40K 80K 120K 160K 200K SE +/- 1020.90, N = 3 SE +/- 101.44, N = 3 SE +/- 1109.71, N = 3 149056.95 171120.35 128050.22 1. (CXX) g++ options: -std=gnu++17 -O3 -fvisibility=hidden -fcommon -rdynamic -lnl-3 -lnl-genl-3 -lpcre -lpthread -lz -lssl -lcrypto -lhwloc -ldl -lm -pthread
Intel Open Image Denoise Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-Only EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 2.0 Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-Only SMT Off SMT On 1.071 2.142 3.213 4.284 5.355 SE +/- 0.02, N = 7 SE +/- 0.00, N = 7 SE +/- 0.03, N = 7 SE +/- 0.01, N = 7 3.57 3.62 4.35 4.76
TensorFlow Device: CPU - Batch Size: 256 - Model: AlexNet EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 256 - Model: AlexNet SMT Off SMT On 300 600 900 1200 1500 SE +/- 8.29, N = 3 SE +/- 3.81, N = 3 SE +/- 11.49, N = 15 SE +/- 14.23, N = 15 1375.33 1422.08 1581.66 1225.41
MariaDB Clients: 4096 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Queries Per Second, More Is Better MariaDB 11.0.1 Clients: 4096 SMT Off SMT On 150 300 450 600 750 SE +/- 3.99, N = 3 SE +/- 8.08, N = 3 SE +/- 1.48, N = 3 SE +/- 5.45, N = 6 695 655 579 545 1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lnuma -lpcre2-8 -lcrypt -lz -lm -lssl -lcrypto -lpthread -ldl
Timed LLVM Compilation Build System: Ninja EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Timed LLVM Compilation 16.0 Build System: Ninja SMT Off SMT On 30 60 90 120 150 SE +/- 0.18, N = 3 SE +/- 0.29, N = 3 SE +/- 0.70, N = 3 SE +/- 0.91, N = 3 119.58 125.36 99.22 107.72
TensorFlow Device: CPU - Batch Size: 512 - Model: AlexNet EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 512 - Model: AlexNet SMT Off SMT On 400 800 1200 1600 2000 SE +/- 3.13, N = 3 SE +/- 1.79, N = 3 SE +/- 18.22, N = 3 SE +/- 15.22, N = 15 1526.81 1628.80 1908.76 1770.91
Timed Node.js Compilation Time To Compile EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Timed Node.js Compilation 19.8.1 Time To Compile SMT Off SMT On 30 60 90 120 150 SE +/- 0.07, N = 3 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.09, N = 3 116.12 113.70 93.95 93.27
TensorFlow Device: CPU - Batch Size: 256 - Model: ResNet-50 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 256 - Model: ResNet-50 SMT Off SMT On 30 60 90 120 150 SE +/- 1.45, N = 12 SE +/- 1.07, N = 12 SE +/- 0.68, N = 3 SE +/- 1.70, N = 3 121.47 118.45 146.74 124.98
miniFE Problem Size: Small EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org CG Mflops, More Is Better miniFE 2.2 Problem Size: Small SMT Off SMT On 13K 26K 39K 52K 65K SE +/- 25.22, N = 5 SE +/- 52.11, N = 5 SE +/- 478.13, N = 5 SE +/- 411.42, N = 5 51741.0 51784.1 53798.6 62774.2 1. (CXX) g++ options: -O3 -fopenmp -lmpi_cxx -lmpi
OpenVINO Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU SMT Off SMT On 0.2925 0.585 0.8775 1.17 1.4625 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 1.30 1.21 1.13 1.08 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
Appleseed Scene: Disney Material EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Appleseed 2.0 Beta Scene: Disney Material SMT Off SMT On 10 20 30 40 50 38.49 44.30 40.57
Timed Gem5 Compilation Time To Compile EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Timed Gem5 Compilation 21.2 Time To Compile SMT Off SMT On 40 80 120 160 200 SE +/- 0.50, N = 3 SE +/- 0.32, N = 3 SE +/- 1.24, N = 3 SE +/- 1.29, N = 3 148.48 161.65 148.38 152.59
Timed LLVM Compilation Build System: Unix Makefiles EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Timed LLVM Compilation 16.0 Build System: Unix Makefiles SMT Off SMT On 50 100 150 200 250 SE +/- 0.70, N = 3 SE +/- 0.15, N = 3 SE +/- 0.11, N = 3 SE +/- 0.14, N = 3 213.91 211.08 198.75 199.37
Timed Godot Game Engine Compilation Time To Compile EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better Timed Godot Game Engine Compilation 4.0 Time To Compile SMT Off SMT On 20 40 60 80 100 SE +/- 0.17, N = 3 SE +/- 0.14, N = 3 SE +/- 0.30, N = 3 SE +/- 1.01, N = 3 102.59 105.80 100.24 100.79
OpenVINO Model: Weld Porosity Detection FP16-INT8 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Weld Porosity Detection FP16-INT8 - Device: CPU SMT Off SMT On 3 6 9 12 15 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 10.91 10.84 11.10 11.00 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
nekRS Input: Kershaw OpenBenchmarking.org flops/rank, More Is Better nekRS 23.0 Input: Kershaw SMT Off SMT On 1200M 2400M 3600M 4800M 6000M SE +/- 37190783.06, N = 3 SE +/- 8226739.60, N = 3 5734266667 5808636667 1. (CXX) g++ options: -fopenmp -O2 -march=native -mtune=native -ftree-vectorize -rdynamic -lmpi_cxx -lmpi
nekRS CPU Power Consumption Monitor OpenBenchmarking.org Watts, Fewer Is Better nekRS 23.0 CPU Power Consumption Monitor SMT Off SMT On 60 120 180 240 300 Min: 20.65 / Avg: 287.58 / Max: 350.2 Min: 21.11 / Avg: 288.26 / Max: 349.64
nekRS Input: TurboPipe Periodic OpenBenchmarking.org flops/rank Per Watt, More Is Better nekRS 23.0 Input: TurboPipe Periodic SMT Off SMT On 2M 4M 6M 8M 10M 8992432.84 8806106.19
nekRS CPU Power Consumption Monitor OpenBenchmarking.org Watts, Fewer Is Better nekRS 23.0 CPU Power Consumption Monitor SMT Off SMT On 60 120 180 240 300 Min: 20.08 / Avg: 298.09 / Max: 347.49 Min: 21.32 / Avg: 293.57 / Max: 349.54
nekRS Input: Kershaw OpenBenchmarking.org flops/rank Per Watt, More Is Better nekRS 23.0 Input: Kershaw SMT Off SMT On 4M 8M 12M 16M 20M 19236505.51 19786057.99
Appleseed CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Appleseed 2.0 Beta CPU Power Consumption Monitor SMT Off SMT On 50 100 150 200 250 Min: 20.23 / Avg: 148.63 / Max: 164.7 Min: 21.47 / Avg: 150.04 / Max: 165.18 Min: 126.26 / Avg: 258.14 / Max: 281.63
Appleseed CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Appleseed 2.0 Beta CPU Power Consumption Monitor SMT Off SMT On 70 140 210 280 350 Min: 20.19 / Avg: 222.18 / Max: 246.09 Min: 20.47 / Avg: 219.92 / Max: 241.15 Min: 193.93 / Avg: 357.2 / Max: 399.57
Aircrack-ng CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Aircrack-ng 1.7 CPU Power Consumption Monitor SMT Off SMT On 70 140 210 280 350 Min: 20.61 / Avg: 229.46 / Max: 260.41 Min: 21.31 / Avg: 234.62 / Max: 267.15 Min: 42.14 / Avg: 362.8 / Max: 415.41
Aircrack-ng EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org k/s Per Watt, More Is Better Aircrack-ng 1.7 SMT Off SMT On 160 320 480 640 800 649.61 729.35 352.95
CPU Power Consumption Monitor Phoronix Test Suite System Monitoring EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts CPU Power Consumption Monitor Phoronix Test Suite System Monitoring SMT Off SMT On 140 280 420 560 700 Min: 10.61 / Avg: 238.75 / Max: 362.1 Min: 10.52 / Avg: 248.93 / Max: 397.25 Min: 97.83 / Avg: 446.01 / Max: 702.85 Min: 21.61 / Avg: 460.57 / Max: 792.38
Appleseed CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Appleseed 2.0 Beta CPU Power Consumption Monitor SMT Off SMT On 60 120 180 240 300 Min: 20.77 / Avg: 173.02 / Max: 236.67 Min: 21.17 / Avg: 177.28 / Max: 240.78 Min: 199.89 / Avg: 289.71 / Max: 346.75 Min: 44.3 / Avg: 290.19 / Max: 351.48
OpenVINO CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2022.3 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.96 / Avg: 278.47 / Max: 305.64 Min: 21.48 / Avg: 310.01 / Max: 336.95 Min: 196.32 / Avg: 577.61 / Max: 653.4 Min: 42.85 / Avg: 599.24 / Max: 664.03
OpenVINO CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2022.3 CPU Power Consumption Monitor SMT Off SMT On 100 200 300 400 500 Min: 21.12 / Avg: 294.6 / Max: 327.84 Min: 21.1 / Avg: 320.15 / Max: 350.37 Min: 199.33 / Avg: 511.7 / Max: 565.18 Min: 42.63 / Avg: 504.47 / Max: 594.16
OpenVINO CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2022.3 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 21.15 / Avg: 267.68 / Max: 294.4 Min: 21.25 / Avg: 301.97 / Max: 348.43 Min: 199.39 / Avg: 575.68 / Max: 630.57 Min: 44.24 / Avg: 563.32 / Max: 629.29
OpenVINO CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2022.3 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.67 / Avg: 301.27 / Max: 325.39 Min: 20.98 / Avg: 307.35 / Max: 332.52 Min: 198.25 / Avg: 614.16 / Max: 663.84 Min: 43.92 / Avg: 603.75 / Max: 664.52
OpenVINO CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2022.3 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.68 / Avg: 303.27 / Max: 334.77 Min: 21.17 / Avg: 293.93 / Max: 343.74 Min: 197.64 / Avg: 618.57 / Max: 675.82 Min: 44.56 / Avg: 611.86 / Max: 676.4
OpenVINO CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2022.3 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.96 / Avg: 318.56 / Max: 343.94 Min: 20.7 / Avg: 327.23 / Max: 352.86 Min: 197.92 / Avg: 655.04 / Max: 695.42 Min: 44.97 / Avg: 643.58 / Max: 697.31
OpenVINO CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2022.3 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.85 / Avg: 299.11 / Max: 323.89 Min: 21.17 / Avg: 297.12 / Max: 335.54 Min: 195.61 / Avg: 613.08 / Max: 656.04 Min: 42.91 / Avg: 602.26 / Max: 657.52
OpenVINO Model: Vehicle Detection FP16-INT8 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2022.3 Model: Vehicle Detection FP16-INT8 - Device: CPU SMT Off SMT On 3 6 9 12 15 SE +/- 0.00, N = 3 SE +/- 0.20, N = 15 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 4.74 12.07 4.86 4.83 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenVINO Model: Vehicle Detection FP16-INT8 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Vehicle Detection FP16-INT8 - Device: CPU SMT Off SMT On 3K 6K 9K 12K 15K SE +/- 3.11, N = 3 SE +/- 103.73, N = 15 SE +/- 2.74, N = 3 SE +/- 7.09, N = 3 6744.66 5311.48 13141.51 13214.20 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenVINO CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2022.3 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.4 / Avg: 294.88 / Max: 329.14 Min: 20.62 / Avg: 295.03 / Max: 335.3 Min: 195.77 / Avg: 585.67 / Max: 664.94 Min: 45.43 / Avg: 578.64 / Max: 665.53
OpenVINO CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2022.3 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.5 / Avg: 276.2 / Max: 333.13 Min: 10.52 / Avg: 252.92 / Max: 325.99 Min: 193.73 / Avg: 615.75 / Max: 674.39 Min: 42.88 / Avg: 597.65 / Max: 674.79
OpenVINO Model: Vehicle Detection FP16 - Device: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org FPS, More Is Better OpenVINO 2022.3 Model: Vehicle Detection FP16 - Device: CPU SMT Off SMT On 1300 2600 3900 5200 6500 SE +/- 25.69, N = 13 SE +/- 21.50, N = 14 SE +/- 87.31, N = 13 SE +/- 93.68, N = 15 2784.67 1464.29 6242.68 5810.41 1. (CXX) g++ options: -isystem -fsigned-char -ffunction-sections -fdata-sections -msse4.1 -msse4.2 -O3 -fno-strict-overflow -fwrapv -fPIC -fvisibility=hidden -Os -std=c++11 -MD -MT -MF
OpenVINO CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2022.3 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.58 / Avg: 307.41 / Max: 349.24 Min: 20.83 / Avg: 304.46 / Max: 356.65 Min: 195.17 / Avg: 529.98 / Max: 666.79 Min: 44.3 / Avg: 527.63 / Max: 662.1
OpenVINO CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2022.3 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.69 / Avg: 306.74 / Max: 352.51 Min: 21.09 / Avg: 304.04 / Max: 357.06 Min: 199.54 / Avg: 537.68 / Max: 660.12 Min: 43.22 / Avg: 524.02 / Max: 656.42
OpenVINO CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2022.3 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.94 / Avg: 316.32 / Max: 349.13 Min: 21.39 / Avg: 315.89 / Max: 357.76 Min: 193.21 / Avg: 621.51 / Max: 702.85 Min: 45.01 / Avg: 611.1 / Max: 703.01
Blender CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Blender 3.6 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 21.11 / Avg: 297.19 / Max: 330.72 Min: 21.55 / Avg: 303.74 / Max: 348.04 Min: 194.9 / Avg: 571.12 / Max: 665.24 Min: 43.07 / Avg: 552.47 / Max: 701.42
Blender CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Blender 3.6 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.45 / Avg: 315.38 / Max: 335.91 Min: 20.88 / Avg: 319.26 / Max: 346.14 Min: 195.38 / Avg: 605.16 / Max: 674.35 Min: 44.79 / Avg: 604.43 / Max: 701.89
Blender CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Blender 3.6 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.68 / Avg: 257.96 / Max: 331.22 Min: 20.92 / Avg: 259.04 / Max: 346.77 Min: 194.02 / Avg: 484.54 / Max: 664 Min: 42.57 / Avg: 439.22 / Max: 700.81
Blender CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Blender 3.6 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.53 / Avg: 290.08 / Max: 328.71 Min: 20.89 / Avg: 294.4 / Max: 340.14 Min: 195.46 / Avg: 557.64 / Max: 654.85 Min: 41.98 / Avg: 525.55 / Max: 684.07
Blender CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Blender 3.6 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.52 / Avg: 245.21 / Max: 323.25 Min: 21.06 / Avg: 242.87 / Max: 337.46 Min: 196.82 / Avg: 459.78 / Max: 647.78 Min: 42.04 / Avg: 405.53 / Max: 681.36
Neural Magic DeepSparse CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Neural Magic DeepSparse 1.5 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.77 / Avg: 235.91 / Max: 344.85 Min: 21.1 / Avg: 250.83 / Max: 357.61 Min: 192.79 / Avg: 479.5 / Max: 693.71 Min: 44.26 / Avg: 486.69 / Max: 702.22
Neural Magic DeepSparse CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Neural Magic DeepSparse 1.5 CPU Power Consumption Monitor SMT Off SMT On 130 260 390 520 650 Min: 20.7 / Avg: 235.42 / Max: 345.62 Min: 20.88 / Avg: 247.32 / Max: 358.25 Min: 195.28 / Avg: 482.32 / Max: 698.28 Min: 41.6 / Avg: 459.55 / Max: 712.43
Neural Magic DeepSparse CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Neural Magic DeepSparse 1.5 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.67 / Avg: 228.97 / Max: 321.66 Min: 21.09 / Avg: 235.18 / Max: 327.39 Min: 101.92 / Avg: 460.64 / Max: 658.21 Min: 42.89 / Avg: 453.79 / Max: 677.58
Neural Magic DeepSparse Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.5 Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream SMT Off SMT On 110 220 330 440 550 SE +/- 8.17, N = 15 SE +/- 0.68, N = 3 SE +/- 3.79, N = 3 SE +/- 0.26, N = 3 468.78 499.59 426.74 521.31
Neural Magic DeepSparse Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.5 Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream SMT Off SMT On 60 120 180 240 300 SE +/- 2.61, N = 15 SE +/- 0.15, N = 3 SE +/- 2.51, N = 3 SE +/- 0.12, N = 3 134.57 126.93 290.76 242.96
Neural Magic DeepSparse CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Neural Magic DeepSparse 1.5 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.87 / Avg: 257.43 / Max: 318.21 Min: 20.93 / Avg: 256.18 / Max: 318.33 Min: 196.83 / Avg: 534.41 / Max: 647.28 Min: 43.67 / Avg: 535.06 / Max: 668.18
Neural Magic DeepSparse CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Neural Magic DeepSparse 1.5 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.35 / Avg: 278.52 / Max: 329.31 Min: 21.26 / Avg: 279.75 / Max: 331.26 Min: 193.12 / Avg: 553.79 / Max: 663.52 Min: 43.24 / Avg: 556.63 / Max: 669.25
Neural Magic DeepSparse CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Neural Magic DeepSparse 1.5 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.34 / Avg: 281.81 / Max: 336.34 Min: 20.8 / Avg: 291.22 / Max: 348.67 Min: 194.59 / Avg: 576 / Max: 679.21 Min: 43.03 / Avg: 568.51 / Max: 686.75
Neural Magic DeepSparse CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Neural Magic DeepSparse 1.5 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.52 / Avg: 236.78 / Max: 330.52 Min: 20.98 / Avg: 231.96 / Max: 341.24 Min: 192.96 / Avg: 504.41 / Max: 673.83 Min: 42.65 / Avg: 496.12 / Max: 694.42
Neural Magic DeepSparse Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.5 Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream SMT Off SMT On 60 120 180 240 300 SE +/- 6.30, N = 15 SE +/- 6.66, N = 15 SE +/- 3.96, N = 15 SE +/- 2.29, N = 15 259.28 267.71 212.68 229.16
Neural Magic DeepSparse Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.5 Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream SMT Off SMT On 130 260 390 520 650 SE +/- 8.20, N = 15 SE +/- 7.96, N = 15 SE +/- 11.89, N = 15 SE +/- 6.10, N = 15 246.37 240.87 590.88 556.94
Neural Magic DeepSparse CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Neural Magic DeepSparse 1.5 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.36 / Avg: 251.27 / Max: 319.12 Min: 20.92 / Avg: 250.7 / Max: 320.58 Min: 193.54 / Avg: 505.33 / Max: 653.26 Min: 42.64 / Avg: 517.56 / Max: 680.99
Neural Magic DeepSparse CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Neural Magic DeepSparse 1.5 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.07 / Avg: 233.36 / Max: 344.64 Min: 20.51 / Avg: 247.92 / Max: 356.8 Min: 194.94 / Avg: 478.36 / Max: 692.61 Min: 43.81 / Avg: 482.63 / Max: 702.05
TensorFlow CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.12 CPU Power Consumption Monitor SMT Off SMT On 90 180 270 360 450 Min: 20.41 / Avg: 222.86 / Max: 282.49 Min: 20.5 / Avg: 225.8 / Max: 286.84 Min: 194 / Avg: 441.9 / Max: 511.48 Min: 42.05 / Avg: 421.61 / Max: 491.54
TensorFlow Device: CPU - Batch Size: 512 - Model: ResNet-50 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 512 - Model: ResNet-50 SMT Off SMT On 0.1251 0.2502 0.3753 0.5004 0.6255 0.556 0.544 0.429 0.409
TensorFlow CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.12 CPU Power Consumption Monitor SMT Off SMT On 90 180 270 360 450 Min: 20.39 / Avg: 228.42 / Max: 276.77 Min: 20.81 / Avg: 235.45 / Max: 279.59 Min: 192.59 / Avg: 456.01 / Max: 498.3 Min: 21.61 / Avg: 439.04 / Max: 490.17
TensorFlow Device: CPU - Batch Size: 512 - Model: GoogLeNet EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 512 - Model: GoogLeNet SMT Off SMT On 0.4228 0.8456 1.2684 1.6912 2.114 1.879 1.767 1.391 1.227
TensorFlow CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.12 CPU Power Consumption Monitor SMT Off SMT On 80 160 240 320 400 Min: 20.24 / Avg: 222.59 / Max: 275.13 Min: 20.64 / Avg: 220.96 / Max: 274.28 Min: 194.78 / Avg: 406.75 / Max: 462.03 Min: 42.84 / Avg: 391.59 / Max: 458.74
TensorFlow Device: CPU - Batch Size: 256 - Model: ResNet-50 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 256 - Model: ResNet-50 SMT Off SMT On 0.1229 0.2458 0.3687 0.4916 0.6145 0.546 0.536 0.361 0.319
TensorFlow CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.12 CPU Power Consumption Monitor SMT Off SMT On 80 160 240 320 400 Min: 20.23 / Avg: 237.07 / Max: 270.8 Min: 20.72 / Avg: 243.24 / Max: 277.92 Min: 189.77 / Avg: 412.23 / Max: 457.74 Min: 42.83 / Avg: 383.45 / Max: 426.38
TensorFlow Device: CPU - Batch Size: 256 - Model: GoogLeNet EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 256 - Model: GoogLeNet SMT Off SMT On 0.4986 0.9972 1.4958 1.9944 2.493 2.216 2.072 1.099 0.859
TensorFlow CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.12 CPU Power Consumption Monitor SMT Off SMT On 90 180 270 360 450 Min: 19.92 / Avg: 213.28 / Max: 261.74 Min: 20.29 / Avg: 235.29 / Max: 286.59 Min: 192.51 / Avg: 420.4 / Max: 490.99 Min: 42.41 / Avg: 386.28 / Max: 479.62
TensorFlow Device: CPU - Batch Size: 512 - Model: AlexNet EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 512 - Model: AlexNet SMT Off SMT On 2 4 6 8 10 7.159 6.922 4.540 4.584
TensorFlow CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.12 CPU Power Consumption Monitor SMT Off SMT On 80 160 240 320 400 Min: 19.84 / Avg: 191.63 / Max: 248.98 Min: 20.26 / Avg: 201.49 / Max: 263.75 Min: 194.45 / Avg: 376.07 / Max: 452.14 Min: 41.4 / Avg: 334.38 / Max: 420.36
TensorFlow Device: CPU - Batch Size: 256 - Model: AlexNet EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 256 - Model: AlexNet SMT Off SMT On 2 4 6 8 10 7.177 7.058 4.206 3.665
PostgreSQL CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better PostgreSQL 15 CPU Power Consumption Monitor SMT Off SMT On 70 140 210 280 350 Min: 19.47 / Avg: 137.98 / Max: 261.97 Min: 19.92 / Avg: 141 / Max: 209.56 Min: 191.46 / Avg: 297.52 / Max: 385.38 Min: 40.03 / Avg: 294.82 / Max: 404.71
PostgreSQL Scaling Factor: 1000 - Clients: 800 - Mode: Read Only - Average Latency EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 15 Scaling Factor: 1000 - Clients: 800 - Mode: Read Only - Average Latency SMT Off SMT On 0.2291 0.4582 0.6873 0.9164 1.1455 SE +/- 0.020, N = 12 SE +/- 0.040, N = 9 SE +/- 0.006, N = 3 SE +/- 0.025, N = 12 0.917 0.952 1.018 0.974 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
PostgreSQL Scaling Factor: 1000 - Clients: 800 - Mode: Read Only EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org TPS, More Is Better PostgreSQL 15 Scaling Factor: 1000 - Clients: 800 - Mode: Read Only SMT Off SMT On 200K 400K 600K 800K 1000K SE +/- 21185.07, N = 12 SE +/- 45813.37, N = 9 SE +/- 4149.87, N = 3 SE +/- 23693.25, N = 12 877722 855569 785968 827878 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
MariaDB CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better MariaDB 11.0.1 CPU Power Consumption Monitor SMT Off SMT On 50 100 150 200 250 Min: 19.33 / Avg: 123.19 / Max: 137.52 Min: 20.08 / Avg: 122.04 / Max: 137.54 Min: 188.93 / Avg: 257.46 / Max: 279.59 Min: 40.82 / Avg: 250.44 / Max: 275.37
MariaDB Clients: 4096 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Queries Per Second Per Watt, More Is Better MariaDB 11.0.1 Clients: 4096 SMT Off SMT On 1.2695 2.539 3.8085 5.078 6.3475 5.642 5.367 2.249 2.176
MariaDB CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better MariaDB 11.0.1 CPU Power Consumption Monitor SMT Off SMT On 50 100 150 200 250 Min: 20.3 / Avg: 123.06 / Max: 134.69 Min: 20.42 / Avg: 122.43 / Max: 135.29 Min: 188.47 / Avg: 254.51 / Max: 270.43 Min: 43.1 / Avg: 245.22 / Max: 266.18
MariaDB Clients: 2048 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Queries Per Second Per Watt, More Is Better MariaDB 11.0.1 Clients: 2048 SMT Off SMT On 2 4 6 8 10 6.363 6.371 2.322 2.365
Graph500 CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Graph500 3.0 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.89 / Avg: 286.97 / Max: 330.07 Min: 20.27 / Avg: 293.26 / Max: 336.48 Min: 40.03 / Avg: 646.2 / Max: 688.13 Min: 41.39 / Avg: 634.69 / Max: 685.66
Graph500 Scale: 26 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org sssp max_TEPS Per Watt, More Is Better Graph500 3.0 Scale: 26 SMT Off SMT On 400K 800K 1200K 1600K 2000K 1719832.82 1520544.67 1664973.76 1513281.14
ASTC Encoder CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better ASTC Encoder 4.0 CPU Power Consumption Monitor SMT Off SMT On 130 260 390 520 650 Min: 19.61 / Avg: 174.24 / Max: 339.1 Min: 19.98 / Avg: 172.1 / Max: 353.68 Min: 40.28 / Avg: 278.8 / Max: 679.54 Min: 41.13 / Avg: 283.59 / Max: 713.06
ASTC Encoder Preset: Exhaustive EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org MT/s Per Watt, More Is Better ASTC Encoder 4.0 Preset: Exhaustive SMT Off SMT On 0.0126 0.0252 0.0378 0.0504 0.063 0.042 0.047 0.052 0.056
ASTC Encoder CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better ASTC Encoder 4.0 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.58 / Avg: 134.61 / Max: 338.88 Min: 14.08 / Avg: 133.8 / Max: 351.54 Min: 40.08 / Avg: 220.5 / Max: 656.14 Min: 41.01 / Avg: 227.98 / Max: 671.57
ASTC Encoder Preset: Thorough EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org MT/s Per Watt, More Is Better ASTC Encoder 4.0 Preset: Thorough SMT Off SMT On 0.1332 0.2664 0.3996 0.5328 0.666 0.507 0.561 0.577 0.592
ASTC Encoder CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better ASTC Encoder 4.0 CPU Power Consumption Monitor SMT Off SMT On 80 160 240 320 400 Min: 20.06 / Avg: 117.33 / Max: 315.16 Min: 20.41 / Avg: 123.13 / Max: 316.82 Min: 41.7 / Avg: 234.69 / Max: 432.58 Min: 41.99 / Avg: 247.05 / Max: 432.29
ASTC Encoder Preset: Fast EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org MT/s Per Watt, More Is Better ASTC Encoder 4.0 Preset: Fast SMT Off SMT On 3 6 9 12 15 10.898 9.671 2.954 2.470
Liquid-DSP CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 20.3 / Avg: 259.79 / Max: 301.25 Min: 20.83 / Avg: 267.43 / Max: 313.4 Min: 42.64 / Avg: 515.71 / Max: 606.26 Min: 43.17 / Avg: 542.11 / Max: 636.19
Liquid-DSP Threads: 512 - Buffer Length: 256 - Filter Length: 512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 512 - Buffer Length: 256 - Filter Length: 512 SMT Off SMT On 1.4M 2.8M 4.2M 5.6M 7M 5445501.58 6669884.78 5062810.80 6142943.89
Liquid-DSP CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 20.66 / Avg: 260.28 / Max: 300.73 Min: 20.84 / Avg: 271.1 / Max: 312.91 Min: 41.3 / Avg: 524.17 / Max: 605.41 Min: 42.35 / Avg: 526.88 / Max: 607.73
Liquid-DSP Threads: 256 - Buffer Length: 256 - Filter Length: 512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 256 - Buffer Length: 256 - Filter Length: 512 SMT Off SMT On 1.3M 2.6M 3.9M 5.2M 6.5M 5048094.72 6258723.14 4850824.91 4829201.80
OpenSSL CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.1 CPU Power Consumption Monitor SMT Off SMT On 130 260 390 520 650 Min: 20.77 / Avg: 322.28 / Max: 351.12 Min: 21.26 / Avg: 360.04 / Max: 378.01 Min: 42.18 / Avg: 661.5 / Max: 692.49 Min: 42.98 / Avg: 724.1 / Max: 753.79
OpenSSL Algorithm: ChaCha20-Poly1305 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org byte/s Per Watt, More Is Better OpenSSL 3.1 Algorithm: ChaCha20-Poly1305 SMT Off SMT On 300M 600M 900M 1200M 1500M 1218755474.84 1284331827.02 1185929906.41 1256483723.09
OpenSSL CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.1 CPU Power Consumption Monitor SMT Off SMT On 130 260 390 520 650 Min: 20.86 / Avg: 336.86 / Max: 352.15 Min: 21.31 / Avg: 341.99 / Max: 354.9 Min: 42.53 / Avg: 675.04 / Max: 707.9 Min: 45.1 / Avg: 691.78 / Max: 716.83
OpenSSL Algorithm: AES-256-GCM EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org byte/s Per Watt, More Is Better OpenSSL 3.1 Algorithm: AES-256-GCM SMT Off SMT On 600M 1200M 1800M 2400M 3000M 2959773610.69 2960716545.71 2960156985.81 2919004979.80
OpenSSL CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.1 CPU Power Consumption Monitor SMT Off SMT On 130 260 390 520 650 Min: 20.8 / Avg: 341.74 / Max: 355.14 Min: 20.93 / Avg: 343.2 / Max: 354.82 Min: 42.47 / Avg: 687.22 / Max: 716.34 Min: 42.1 / Avg: 693.22 / Max: 715.73
OpenSSL Algorithm: AES-128-GCM EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org byte/s Per Watt, More Is Better OpenSSL 3.1 Algorithm: AES-128-GCM SMT Off SMT On 700M 1400M 2100M 2800M 3500M 3373392798.47 3408107334.73 3357766493.21 3375379415.53
OpenSSL CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.1 CPU Power Consumption Monitor SMT Off SMT On 140 280 420 560 700 Min: 20.75 / Avg: 308.01 / Max: 338.71 Min: 21.23 / Avg: 363.8 / Max: 392.13 Min: 42.08 / Avg: 660.38 / Max: 727.38 Min: 44.02 / Avg: 730.27 / Max: 792.36
OpenSSL Algorithm: ChaCha20 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org byte/s Per Watt, More Is Better OpenSSL 3.1 Algorithm: ChaCha20 SMT Off SMT On 400M 800M 1200M 1600M 2000M 1787936411.25 1812397903.61 1666389521.27 1804203926.06
OpenSSL CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.1 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 21.32 / Avg: 284.75 / Max: 316.61 Min: 21.6 / Avg: 314.07 / Max: 354.72 Min: 42.69 / Avg: 599.56 / Max: 681.84 Min: 44.51 / Avg: 617.73 / Max: 697.4
OpenSSL Algorithm: RSA4096 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org verify/s Per Watt, More Is Better OpenSSL 3.1 Algorithm: RSA4096 SMT Off SMT On 1400 2800 4200 5600 7000 6318.78 6020.72 6002.69 6122.57
OpenSSL CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.1 CPU Power Consumption Monitor SMT Off SMT On 130 260 390 520 650 Min: 20.38 / Avg: 332.69 / Max: 354.84 Min: 21.29 / Avg: 335.75 / Max: 346.67 Min: 41.9 / Avg: 678.87 / Max: 711.36 Min: 44.01 / Avg: 676.25 / Max: 700.91
OpenSSL Algorithm: SHA512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org byte/s Per Watt, More Is Better OpenSSL 3.1 Algorithm: SHA512 SMT Off SMT On 30M 60M 90M 120M 150M 155713911.66 157871758.84 150591857.13 156820493.54
OpenSSL CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.1 CPU Power Consumption Monitor SMT Off SMT On 130 260 390 520 650 Min: 20.37 / Avg: 285.19 / Max: 341.28 Min: 20.72 / Avg: 333.08 / Max: 356.22 Min: 42.55 / Avg: 603.9 / Max: 697.6 Min: 41.96 / Avg: 676 / Max: 716.17
OpenSSL Algorithm: SHA256 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org byte/s Per Watt, More Is Better OpenSSL 3.1 Algorithm: SHA256 SMT Off SMT On 110M 220M 330M 440M 550M 390668971.64 491268818.21 368054429.03 485100025.27
Helsing CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Helsing 1.0-beta CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.68 / Avg: 310.21 / Max: 356.64 Min: 21.03 / Avg: 316.26 / Max: 355.5 Min: 41.78 / Avg: 512.26 / Max: 690.1 Min: 42.81 / Avg: 589.87 / Max: 707.99
OSPRay Studio CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 0.11 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.81 / Avg: 290.35 / Max: 325.63 Min: 21.32 / Avg: 291.25 / Max: 338.09 Min: 42.38 / Avg: 568.08 / Max: 654.97 Min: 42.82 / Avg: 625.58 / Max: 686.92
OSPRay Studio CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 0.11 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.73 / Avg: 274.66 / Max: 324.6 Min: 21.56 / Avg: 282.67 / Max: 337.69 Min: 41.61 / Avg: 566.27 / Max: 654 Min: 44.72 / Avg: 616.86 / Max: 687.3
OSPRay Studio CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 0.11 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.77 / Avg: 284.22 / Max: 324.97 Min: 21.44 / Avg: 307.82 / Max: 338.46 Min: 42.15 / Avg: 573.14 / Max: 653.73 Min: 44.88 / Avg: 575.16 / Max: 684.71
OSPRay Studio CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 0.11 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 10.61 / Avg: 295.55 / Max: 325.63 Min: 21.65 / Avg: 305.15 / Max: 338.21 Min: 41.78 / Avg: 589.05 / Max: 653.82 Min: 43.88 / Avg: 615.69 / Max: 685.2
OSPRay Studio CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 0.11 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 21.33 / Avg: 283.12 / Max: 324.52 Min: 21.29 / Avg: 303.95 / Max: 338.98 Min: 41.88 / Avg: 551.43 / Max: 653.67 Min: 43.58 / Avg: 572.24 / Max: 686.68
OSPRay Studio CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 0.11 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.9 / Avg: 295.01 / Max: 325.94 Min: 21.61 / Avg: 305.02 / Max: 338.57 Min: 42.66 / Avg: 587.43 / Max: 654.45 Min: 42.72 / Avg: 617.3 / Max: 686.55
OSPRay Studio CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 0.11 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.74 / Avg: 281.21 / Max: 323.91 Min: 21.43 / Avg: 287.82 / Max: 335.59 Min: 42.47 / Avg: 564.46 / Max: 651.55 Min: 43.64 / Avg: 585.98 / Max: 682.76
OSPRay Studio CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 0.11 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.75 / Avg: 278.21 / Max: 323.58 Min: 21.25 / Avg: 287.03 / Max: 334.32 Min: 42.03 / Avg: 559.72 / Max: 652.58 Min: 44.18 / Avg: 586.29 / Max: 680.45
OSPRay Studio CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 0.11 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.2 / Avg: 277.45 / Max: 323.23 Min: 20.8 / Avg: 286.61 / Max: 335.12 Min: 41.01 / Avg: 558.64 / Max: 651.98 Min: 41.89 / Avg: 609 / Max: 686.18
Primesieve CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Primesieve 8.0 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.88 / Avg: 261.85 / Max: 316.4 Min: 20.09 / Avg: 268.37 / Max: 324.15 Min: 41.23 / Avg: 464.71 / Max: 636.03 Min: 41.84 / Avg: 485.63 / Max: 670.4
Primesieve CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Primesieve 8.0 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 19.72 / Avg: 104.83 / Max: 309.64 Min: 20.18 / Avg: 113.25 / Max: 312.47 Min: 40.21 / Avg: 164.82 / Max: 618.35 Min: 40.74 / Avg: 202.59 / Max: 640.65
Timed Node.js Compilation CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Timed Node.js Compilation 19.8.1 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.79 / Avg: 195.41 / Max: 330.02 Min: 20.49 / Avg: 192.96 / Max: 340.93 Min: 40.21 / Avg: 335.02 / Max: 662.48 Min: 40.27 / Avg: 330.87 / Max: 680.38
Timed LLVM Compilation CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Timed LLVM Compilation 16.0 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 19.66 / Avg: 159.92 / Max: 323.96 Min: 20.39 / Avg: 159.32 / Max: 327.5 Min: 41.34 / Avg: 283.34 / Max: 646.29 Min: 41.07 / Avg: 282.76 / Max: 651.4
Timed LLVM Compilation CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Timed LLVM Compilation 16.0 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.32 / Avg: 218.43 / Max: 329.81 Min: 20.62 / Avg: 209.56 / Max: 335.43 Min: 41.09 / Avg: 351.2 / Max: 660.92 Min: 42.36 / Avg: 337.35 / Max: 671.56
Timed Linux Kernel Compilation CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Timed Linux Kernel Compilation 6.1 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.38 / Avg: 265.69 / Max: 337.05 Min: 20.05 / Avg: 255.95 / Max: 334.41 Min: 40.59 / Avg: 462.79 / Max: 676.08 Min: 41.29 / Avg: 468.11 / Max: 673.75
Timed Linux Kernel Compilation CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Timed Linux Kernel Compilation 6.1 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.82 / Avg: 149.41 / Max: 323.82 Min: 20.11 / Avg: 158.04 / Max: 312.79 Min: 40.17 / Avg: 267.88 / Max: 656.25 Min: 41.05 / Avg: 285.97 / Max: 638.63
Timed Godot Game Engine Compilation CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Timed Godot Game Engine Compilation 4.0 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.83 / Avg: 154.5 / Max: 328.65 Min: 20.04 / Avg: 148.45 / Max: 334.92 Min: 40.04 / Avg: 275.62 / Max: 665.8 Min: 41.9 / Avg: 275.11 / Max: 667.57
Timed Gem5 Compilation CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Timed Gem5 Compilation 21.2 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.52 / Avg: 131.21 / Max: 323.7 Min: 19.75 / Avg: 126.31 / Max: 329.48 Min: 40.34 / Avg: 250.05 / Max: 652.68 Min: 43.13 / Avg: 253.79 / Max: 657.39
Stockfish CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Stockfish 15 CPU Power Consumption Monitor SMT Off SMT On 130 260 390 520 650 Min: 20.41 / Avg: 309.55 / Max: 341.8 Min: 21.02 / Avg: 326.55 / Max: 356.32 Min: 41.98 / Avg: 598.99 / Max: 681.91 Min: 42.5 / Avg: 649.7 / Max: 710.41
Stockfish Total Time EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Nodes Per Second Per Watt, More Is Better Stockfish 15 Total Time SMT Off SMT On 200K 400K 600K 800K 1000K 881021.76 1117842.33 746293.69 896399.09
Stockfish Total Time EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Nodes Per Second, More Is Better Stockfish 15 Total Time SMT Off SMT On 120M 240M 360M 480M 600M SE +/- 5762415.88, N = 15 SE +/- 7021012.10, N = 12 SE +/- 6859221.31, N = 12 SE +/- 9265130.36, N = 15 272722940 365034349 447023143 582386924 1. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fno-peel-loops -fno-tracer -pedantic -O3 -msse -msse3 -mpopcnt -mavx2 -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto=jobserver
7-Zip Compression CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better 7-Zip Compression 22.01 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.34 / Avg: 247.35 / Max: 330.87 Min: 21.05 / Avg: 266.27 / Max: 350.56 Min: 41.8 / Avg: 472.21 / Max: 667.59 Min: 41.89 / Avg: 501.4 / Max: 701.41
7-Zip Compression Test: Decompression Rating EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org MIPS Per Watt, More Is Better 7-Zip Compression 22.01 Test: Decompression Rating SMT Off SMT On 600 1200 1800 2400 3000 2063.78 2973.58 1933.66 2700.35
OSPRay CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay 2.12 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.36 / Avg: 255.54 / Max: 301.79 Min: 21.29 / Avg: 278.61 / Max: 330.56 Min: 41.49 / Avg: 544.27 / Max: 626.51 Min: 42.48 / Avg: 572.66 / Max: 664.61
OSPRay Benchmark: gravity_spheres_volume/dim_512/scivis/real_time EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Items Per Second Per Watt, More Is Better OSPRay 2.12 Benchmark: gravity_spheres_volume/dim_512/scivis/real_time SMT Off SMT On 0.0259 0.0518 0.0777 0.1036 0.1295 0.100 0.115 0.081 0.092
OSPRay CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay 2.12 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.16 / Avg: 254.24 / Max: 300.93 Min: 20.45 / Avg: 276.76 / Max: 329.91 Min: 40.66 / Avg: 542.85 / Max: 624.55 Min: 44.15 / Avg: 566.91 / Max: 661.42
OSPRay Benchmark: gravity_spheres_volume/dim_512/ao/real_time EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Items Per Second Per Watt, More Is Better OSPRay 2.12 Benchmark: gravity_spheres_volume/dim_512/ao/real_time SMT Off SMT On 0.0266 0.0532 0.0798 0.1064 0.133 0.101 0.118 0.082 0.095
OSPRay CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay 2.12 CPU Power Consumption Monitor SMT Off SMT On 100 200 300 400 500 Min: 20.16 / Avg: 211.09 / Max: 275.73 Min: 20.96 / Avg: 241.99 / Max: 304.45 Min: 42.19 / Avg: 462.03 / Max: 561.22 Min: 44 / Avg: 483.3 / Max: 582.81
OSPRay Benchmark: particle_volume/scivis/real_time EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Items Per Second Per Watt, More Is Better OSPRay 2.12 Benchmark: particle_volume/scivis/real_time SMT Off SMT On 0.0286 0.0572 0.0858 0.1144 0.143 0.112 0.127 0.090 0.102
OSPRay CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OSPRay 2.12 CPU Power Consumption Monitor SMT Off SMT On 100 200 300 400 500 Min: 20.26 / Avg: 261.53 / Max: 278.17 Min: 21.03 / Avg: 291.76 / Max: 307.53 Min: 42.16 / Avg: 534.35 / Max: 564.38 Min: 41.99 / Avg: 553.35 / Max: 585.29
OSPRay Benchmark: particle_volume/ao/real_time EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Items Per Second Per Watt, More Is Better OSPRay 2.12 Benchmark: particle_volume/ao/real_time SMT Off SMT On 0.0239 0.0478 0.0717 0.0956 0.1195 0.091 0.106 0.078 0.089
OpenVKL CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better OpenVKL 1.3.1 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 19.63 / Avg: 223.41 / Max: 305.39 Min: 20.44 / Avg: 238.51 / Max: 321.14 Min: 41.74 / Avg: 435.67 / Max: 614.27 Min: 41.51 / Avg: 467.94 / Max: 643.91
OpenVKL Benchmark: vklBenchmark ISPC EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Items / Sec Per Watt, More Is Better OpenVKL 1.3.1 Benchmark: vklBenchmark ISPC SMT Off SMT On 1.3169 2.6338 3.9507 5.2676 6.5845 4.955 5.853 3.512 3.676
Intel Open Image Denoise CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Intel Open Image Denoise 2.0 CPU Power Consumption Monitor SMT Off SMT On 100 200 300 400 500 Min: 19.75 / Avg: 215.85 / Max: 296.82 Min: 20.37 / Avg: 219.82 / Max: 297.83 Min: 41.02 / Avg: 379.66 / Max: 554.87 Min: 41.62 / Avg: 383.91 / Max: 567.02
Intel Open Image Denoise Run: RTLightmap.hdr.4096x4096 - Device: CPU-Only EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Images / Sec Per Watt, More Is Better Intel Open Image Denoise 2.0 Run: RTLightmap.hdr.4096x4096 - Device: CPU-Only SMT Off SMT On 0.0018 0.0036 0.0054 0.0072 0.009 0.008 0.008 0.006 0.006
Intel Open Image Denoise CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Intel Open Image Denoise 2.0 CPU Power Consumption Monitor SMT Off SMT On 100 200 300 400 500 Min: 19.82 / Avg: 174.13 / Max: 300.84 Min: 20.39 / Avg: 170.54 / Max: 297.28 Min: 40.95 / Avg: 295.21 / Max: 552.75 Min: 40.73 / Avg: 294.58 / Max: 558.39
Intel Open Image Denoise Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-Only EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Images / Sec Per Watt, More Is Better Intel Open Image Denoise 2.0 Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-Only SMT Off SMT On 0.0047 0.0094 0.0141 0.0188 0.0235 0.021 0.021 0.015 0.016
Intel Open Image Denoise CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Intel Open Image Denoise 2.0 CPU Power Consumption Monitor SMT Off SMT On 100 200 300 400 500 Min: 19.89 / Avg: 172.35 / Max: 299.67 Min: 20.3 / Avg: 176.53 / Max: 298.39 Min: 21.72 / Avg: 294.18 / Max: 551.63 Min: 41.67 / Avg: 298.49 / Max: 561.86
Intel Open Image Denoise Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-Only EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Images / Sec Per Watt, More Is Better Intel Open Image Denoise 2.0 Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-Only SMT Off SMT On 0.0047 0.0094 0.0141 0.0188 0.0235 0.021 0.021 0.015 0.016
Embree CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Embree 4.1 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 20.14 / Avg: 183.96 / Max: 309.98 Min: 20.48 / Avg: 170.51 / Max: 326.15 Min: 40.75 / Avg: 312.56 / Max: 630.49 Min: 41.35 / Avg: 286.01 / Max: 649.79
Embree Binary: Pathtracer ISPC - Model: Asian Dragon EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Embree 4.1 Binary: Pathtracer ISPC - Model: Asian Dragon SMT Off SMT On 0.2081 0.4162 0.6243 0.8324 1.0405 0.584 0.925 0.570 0.895
Embree CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Embree 4.1 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.75 / Avg: 204.9 / Max: 312.76 Min: 20.54 / Avg: 191.71 / Max: 332.64 Min: 40.12 / Avg: 347.93 / Max: 636 Min: 41.29 / Avg: 314.33 / Max: 666.86
Embree Binary: Pathtracer ISPC - Model: Crown EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Embree 4.1 Binary: Pathtracer ISPC - Model: Crown SMT Off SMT On 0.1503 0.3006 0.4509 0.6012 0.7515 0.416 0.655 0.420 0.668
LuxCoreRender CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better LuxCoreRender 2.6 CPU Power Consumption Monitor SMT Off SMT On 70 140 210 280 350 Min: 20.14 / Avg: 136.45 / Max: 184.56 Min: 20.9 / Avg: 142.31 / Max: 202.02 Min: 40.47 / Avg: 279.04 / Max: 366.94 Min: 41.26 / Avg: 281.59 / Max: 380.32
LuxCoreRender Scene: Rainbow Colors and Prism - Acceleration: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org M samples/sec Per Watt, More Is Better LuxCoreRender 2.6 Scene: Rainbow Colors and Prism - Acceleration: CPU SMT Off SMT On 0.0331 0.0662 0.0993 0.1324 0.1655 0.105 0.147 0.048 0.068
LuxCoreRender Scene: Rainbow Colors and Prism - Acceleration: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org M samples/sec, More Is Better LuxCoreRender 2.6 Scene: Rainbow Colors and Prism - Acceleration: CPU SMT Off SMT On 5 10 15 20 25 SE +/- 0.04, N = 4 SE +/- 0.03, N = 5 SE +/- 0.35, N = 15 SE +/- 0.03, N = 5 14.37 20.88 13.51 19.02
LuxCoreRender CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better LuxCoreRender 2.6 CPU Power Consumption Monitor SMT Off SMT On 90 180 270 360 450 Min: 20.62 / Avg: 247.27 / Max: 288.16 Min: 21.47 / Avg: 286.22 / Max: 332.7 Min: 41.4 / Avg: 402.3 / Max: 472.3 Min: 42.57 / Avg: 445.45 / Max: 519.63
LuxCoreRender Scene: LuxCore Benchmark - Acceleration: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org M samples/sec Per Watt, More Is Better LuxCoreRender 2.6 Scene: LuxCore Benchmark - Acceleration: CPU SMT Off SMT On 0.0097 0.0194 0.0291 0.0388 0.0485 0.036 0.043 0.017 0.022
LuxCoreRender CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better LuxCoreRender 2.6 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.39 / Avg: 293.46 / Max: 329.9 Min: 21.26 / Avg: 298.35 / Max: 337.28 Min: 41.2 / Avg: 501.47 / Max: 651.41 Min: 42.69 / Avg: 558.11 / Max: 673.06
LuxCoreRender Scene: Orange Juice - Acceleration: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org M samples/sec Per Watt, More Is Better LuxCoreRender 2.6 Scene: Orange Juice - Acceleration: CPU SMT Off SMT On 0.0185 0.037 0.0555 0.074 0.0925 0.071 0.082 0.050 0.062
LuxCoreRender Scene: Orange Juice - Acceleration: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org M samples/sec, More Is Better LuxCoreRender 2.6 Scene: Orange Juice - Acceleration: CPU SMT Off SMT On 8 16 24 32 40 SE +/- 0.03, N = 3 SE +/- 0.30, N = 15 SE +/- 0.60, N = 15 SE +/- 1.35, N = 15 20.98 24.47 25.11 34.45
LuxCoreRender CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better LuxCoreRender 2.6 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.7 / Avg: 285.26 / Max: 319.31 Min: 20.94 / Avg: 303.49 / Max: 332.56 Min: 41.49 / Avg: 468.61 / Max: 635.46 Min: 41.78 / Avg: 510.83 / Max: 659.11
LuxCoreRender Scene: DLSC - Acceleration: CPU EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org M samples/sec Per Watt, More Is Better LuxCoreRender 2.6 Scene: DLSC - Acceleration: CPU SMT Off SMT On 0.0122 0.0244 0.0366 0.0488 0.061 0.047 0.054 0.031 0.036
John The Ripper CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better John The Ripper 2023.03.14 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.4 / Avg: 296.91 / Max: 362.1 Min: 20.42 / Avg: 291.74 / Max: 363.73 Min: 40.66 / Avg: 526.99 / Max: 698.52 Min: 41.25 / Avg: 517.18 / Max: 684.78
John The Ripper Test: MD5 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Real C/S Per Watt, More Is Better John The Ripper 2023.03.14 Test: MD5 SMT Off SMT On 15K 30K 45K 60K 75K 56420.63 69625.90 57347.31 67440.80
John The Ripper Test: Blowfish EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Real C/S Per Watt, More Is Better John The Ripper 2023.03.14 Test: Blowfish SMT Off SMT On 170 340 510 680 850 632.58 768.10 582.13 726.35
John The Ripper CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better John The Ripper 2023.03.14 CPU Power Consumption Monitor SMT Off SMT On 140 280 420 560 700 Min: 20.23 / Avg: 301.84 / Max: 349.66 Min: 20.68 / Avg: 348.64 / Max: 397.25 Min: 41.73 / Avg: 635.7 / Max: 734.09 Min: 43.21 / Avg: 681.87 / Max: 792.38
John The Ripper Test: WPA PSK EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Real C/S Per Watt, More Is Better John The Ripper 2023.03.14 Test: WPA PSK SMT Off SMT On 500 1000 1500 2000 2500 2240.31 2324.37 2047.36 2235.81
John The Ripper CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better John The Ripper 2023.03.14 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 20 / Avg: 254.23 / Max: 291.94 Min: 20.76 / Avg: 258.09 / Max: 293.57 Min: 20.17 / Avg: 280.29 / Max: 320.86 Min: 20.74 / Avg: 281.36 / Max: 322.03 Min: 40.81 / Avg: 541.52 / Max: 627.33 Min: 42.35 / Avg: 551.23 / Max: 631.68 Min: 42.39 / Avg: 564.26 / Max: 642.19 Min: 42.38 / Avg: 553.47 / Max: 644.02
John The Ripper Test: bcrypt EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Real C/S Per Watt, More Is Better John The Ripper 2023.03.14 Test: bcrypt SMT Off SMT On 170 340 510 680 850 642.02 770.87 585.47 736.92
srsRAN Project CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better srsRAN Project 23.5 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.63 / Avg: 260.57 / Max: 317.08 Min: 20.98 / Avg: 261.42 / Max: 327.2 Min: 41.48 / Avg: 517.72 / Max: 634.9 Min: 43.09 / Avg: 560.78 / Max: 670.82
srsRAN Project Test: PUSCH Processor Benchmark, Throughput Total EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Mbps Per Watt, More Is Better srsRAN Project 23.5 Test: PUSCH Processor Benchmark, Throughput Total SMT Off SMT On 20 40 60 80 100 78.41 32.09 70.64 31.90
srsRAN Project Test: PUSCH Processor Benchmark, Throughput Total EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Mbps, More Is Better srsRAN Project 23.5 Test: PUSCH Processor Benchmark, Throughput Total SMT Off SMT On 8K 16K 24K 32K 40K SE +/- 54.33, N = 3 SE +/- 50.60, N = 3 SE +/- 211.99, N = 3 SE +/- 831.80, N = 15 20430.9 8389.0 36573.8 17891.4 1. (CXX) g++ options: -march=native -mfma -O3 -fno-trapping-math -fno-math-errno -lgtest
Xmrig CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Xmrig 6.18.1 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.64 / Avg: 267.92 / Max: 342.62 Min: 20.54 / Avg: 267.22 / Max: 350.81 Min: 41.49 / Avg: 447.09 / Max: 663.47 Min: 43.21 / Avg: 439.66 / Max: 708.61
Xmrig Variant: Wownero - Hash Count: 1M EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org H/s Per Watt, More Is Better Xmrig 6.18.1 Variant: Wownero - Hash Count: 1M SMT Off SMT On 70 140 210 280 350 235.83 279.93 225.36 323.17
Xmrig CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better Xmrig 6.18.1 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.83 / Avg: 268.02 / Max: 330.69 Min: 20.63 / Avg: 221.93 / Max: 329.48 Min: 40.87 / Avg: 462.3 / Max: 653.34 Min: 41.43 / Avg: 488.1 / Max: 682.33
Xmrig Variant: Monero - Hash Count: 1M EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org H/s Per Watt, More Is Better Xmrig 6.18.1 Variant: Monero - Hash Count: 1M SMT Off SMT On 40 80 120 160 200 191.10 109.99 185.91 177.29
Xmrig Variant: Monero - Hash Count: 1M EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org H/s, More Is Better Xmrig 6.18.1 Variant: Monero - Hash Count: 1M SMT Off SMT On 20K 40K 60K 80K 100K SE +/- 513.99, N = 3 SE +/- 587.76, N = 15 SE +/- 83.35, N = 4 SE +/- 871.70, N = 4 51218.9 24409.4 85946.0 86533.7 1. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
nekRS Input: TurboPipe Periodic OpenBenchmarking.org flops/rank, More Is Better nekRS 23.0 Input: TurboPipe Periodic SMT Off SMT On 600M 1200M 1800M 2400M 3000M SE +/- 87729075.59, N = 15 SE +/- 62315945.32, N = 13 2586083333 2538406923 1. (CXX) g++ options: -fopenmp -O2 -march=native -mtune=native -ftree-vectorize -rdynamic -lmpi_cxx -lmpi
SPECFEM3D CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better SPECFEM3D 4.0 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.04 / Avg: 228.45 / Max: 329.73 Min: 20.6 / Avg: 249.37 / Max: 338.57 Min: 41.07 / Avg: 358.64 / Max: 670.26 Min: 41.18 / Avg: 415.4 / Max: 678.2
SPECFEM3D CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better SPECFEM3D 4.0 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.06 / Avg: 188.68 / Max: 325.98 Min: 20.79 / Avg: 218.63 / Max: 337.15 Min: 40.89 / Avg: 314 / Max: 658.8 Min: 41.07 / Avg: 338.57 / Max: 663.27
SPECFEM3D Model: Homogeneous Halfspace EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.0 Model: Homogeneous Halfspace SMT Off SMT On 3 6 9 12 15 SE +/- 0.044108994, N = 15 SE +/- 0.048294894, N = 4 SE +/- 0.065109589, N = 12 SE +/- 0.120152278, N = 15 6.092750299 9.404348500 3.451830169 4.727830148 1. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
SPECFEM3D CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better SPECFEM3D 4.0 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.14 / Avg: 175.44 / Max: 323.55 Min: 20.85 / Avg: 203.93 / Max: 337.24 Min: 41.07 / Avg: 299.85 / Max: 655.45 Min: 42.2 / Avg: 321.36 / Max: 664.42
SPECFEM3D Model: Tomographic Model EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.0 Model: Tomographic Model SMT Off SMT On 2 4 6 8 10 SE +/- 0.048140615, N = 15 SE +/- 0.080164897, N = 5 SE +/- 0.014728964, N = 6 SE +/- 0.086611165, N = 15 4.937265506 7.480674706 2.709205428 3.782741798 1. (F9X) gfortran options: -O2 -fopenmp -std=f2003 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
SPECFEM3D CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better SPECFEM3D 4.0 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.36 / Avg: 226.26 / Max: 322.57 Min: 20.77 / Avg: 234.87 / Max: 334.16 Min: 41.53 / Avg: 355.84 / Max: 655.14 Min: 41.53 / Avg: 396.49 / Max: 677.63
SPECFEM3D CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better SPECFEM3D 4.0 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.51 / Avg: 173.78 / Max: 324.55 Min: 21.64 / Avg: 185.39 / Max: 336.6 Min: 43.56 / Avg: 283.61 / Max: 651.99 Min: 42.65 / Avg: 310.22 / Max: 679.55
HeFFTe - Highly Efficient FFT for Exascale CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better HeFFTe - Highly Efficient FFT for Exascale 2.3 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 20.21 / Avg: 229.75 / Max: 280.78 Min: 20.59 / Avg: 233.52 / Max: 280.97 Min: 41.26 / Avg: 404.72 / Max: 596.32 Min: 43.55 / Avg: 408.39 / Max: 622.11
HeFFTe - Highly Efficient FFT for Exascale Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFLOP/s Per Watt, More Is Better HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512 SMT Off SMT On 0.1172 0.2344 0.3516 0.4688 0.586 0.296 0.284 0.521 0.507
HeFFTe - Highly Efficient FFT for Exascale Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFLOP/s, More Is Better HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512 SMT Off SMT On 50 100 150 200 250 SE +/- 3.41, N = 15 SE +/- 3.14, N = 15 SE +/- 1.62, N = 5 SE +/- 0.28, N = 5 67.99 66.34 210.93 207.20 1. (CXX) g++ options: -O3
HeFFTe - Highly Efficient FFT for Exascale CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better HeFFTe - Highly Efficient FFT for Exascale 2.3 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 20.26 / Avg: 246.69 / Max: 279.29 Min: 20.58 / Avg: 248.88 / Max: 282.9 Min: 40.61 / Avg: 446.09 / Max: 590.49 Min: 41.88 / Avg: 458.33 / Max: 608.28
HeFFTe - Highly Efficient FFT for Exascale Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFLOP/s Per Watt, More Is Better HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512 SMT Off SMT On 0.0567 0.1134 0.1701 0.2268 0.2835 0.144 0.140 0.252 0.239
HeFFTe - Highly Efficient FFT for Exascale Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFLOP/s, More Is Better HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512 SMT Off SMT On 30 60 90 120 150 SE +/- 2.10, N = 15 SE +/- 2.09, N = 15 SE +/- 1.48, N = 3 SE +/- 0.77, N = 3 35.40 34.85 112.41 109.65 1. (CXX) g++ options: -O3
HeFFTe - Highly Efficient FFT for Exascale CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better HeFFTe - Highly Efficient FFT for Exascale 2.3 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 20.14 / Avg: 164.96 / Max: 252.95 Min: 20.52 / Avg: 169.96 / Max: 279.94 Min: 40.74 / Avg: 340.8 / Max: 617.42 Min: 42.06 / Avg: 336.38 / Max: 638.26
HeFFTe - Highly Efficient FFT for Exascale Test: r2c - Backend: FFTW - Precision: float - X Y Z: 512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFLOP/s Per Watt, More Is Better HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: r2c - Backend: FFTW - Precision: float - X Y Z: 512 SMT Off SMT On 0.3391 0.6782 1.0173 1.3564 1.6955 1.507 1.445 1.263 1.289
HeFFTe - Highly Efficient FFT for Exascale CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better HeFFTe - Highly Efficient FFT for Exascale 2.3 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 11.69 / Avg: 185.74 / Max: 241.77 Min: 20.78 / Avg: 187.92 / Max: 271.2 Min: 41.14 / Avg: 392.95 / Max: 599.94 Min: 44.94 / Avg: 404.61 / Max: 614.02
HeFFTe - Highly Efficient FFT for Exascale Test: c2c - Backend: FFTW - Precision: float - X Y Z: 512 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFLOP/s Per Watt, More Is Better HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: c2c - Backend: FFTW - Precision: float - X Y Z: 512 SMT Off SMT On 0.1557 0.3114 0.4671 0.6228 0.7785 0.692 0.682 0.569 0.548
libxsmm CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better libxsmm 2-1.17-3645 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.05 / Avg: 210.44 / Max: 332.07 Min: 20.42 / Avg: 211.17 / Max: 339.24 Min: 40.86 / Avg: 319.88 / Max: 674.35 Min: 41.94 / Avg: 321.06 / Max: 680.36
libxsmm M N K: 256 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFLOPS/s Per Watt, More Is Better libxsmm 2-1.17-3645 M N K: 256 SMT Off SMT On 5 10 15 20 25 18.12 15.78 19.92 19.04
libxsmm CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better libxsmm 2-1.17-3645 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.02 / Avg: 204.17 / Max: 323.63 Min: 20.34 / Avg: 205.01 / Max: 330.64 Min: 41.28 / Avg: 324.27 / Max: 648.28 Min: 42.27 / Avg: 322.92 / Max: 664.29
libxsmm M N K: 128 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFLOPS/s Per Watt, More Is Better libxsmm 2-1.17-3645 M N K: 128 SMT Off SMT On 4 8 12 16 20 13.21 13.24 13.89 15.41
libxsmm M N K: 128 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFLOPS/s, More Is Better libxsmm 2-1.17-3645 M N K: 128 SMT Off SMT On 1100 2200 3300 4400 5500 SE +/- 19.26, N = 3 SE +/- 0.99, N = 3 SE +/- 114.55, N = 9 SE +/- 62.14, N = 4 2696.5 2713.4 4505.5 4976.7 1. (CXX) g++ options: -dynamic -Bstatic -static-libgcc -lgomp -lm -lrt -ldl -lquadmath -lstdc++ -pthread -fPIC -std=c++14 -O2 -fopenmp-simd -funroll-loops -ftree-vectorize -fdata-sections -ffunction-sections -fvisibility=hidden -msse4.2
toyBrot Fractal Generator CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better toyBrot Fractal Generator 2020-11-18 CPU Power Consumption Monitor SMT Off SMT On 100 200 300 400 500 Min: 19.84 / Avg: 154.08 / Max: 274.28 Min: 20.45 / Avg: 148.84 / Max: 302.49 Min: 40.79 / Avg: 251.07 / Max: 574.98 Min: 40.87 / Avg: 242.46 / Max: 590.56
toyBrot Fractal Generator CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better toyBrot Fractal Generator 2020-11-18 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 20.03 / Avg: 161.83 / Max: 283.89 Min: 20.91 / Avg: 150.09 / Max: 303.56 Min: 40.87 / Avg: 257.24 / Max: 584.5 Min: 41.02 / Avg: 232.68 / Max: 605.7
NAMD CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better NAMD 2.14 CPU Power Consumption Monitor SMT Off SMT On 130 260 390 520 650 Min: 20.91 / Avg: 242.43 / Max: 333.11 Min: 21.34 / Avg: 258.84 / Max: 353.57 Min: 42.59 / Avg: 421.91 / Max: 639.89 Min: 43.91 / Avg: 386.17 / Max: 712.38
CP2K Molecular Dynamics CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better CP2K Molecular Dynamics 2023.1 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 20.12 / Avg: 290.26 / Max: 349.59 Min: 20.44 / Avg: 291.52 / Max: 350.26 Min: 40.71 / Avg: 623.68 / Max: 701.99 Min: 41.56 / Avg: 614.44 / Max: 701.65
CloverLeaf CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better CloverLeaf CPU Power Consumption Monitor SMT Off SMT On 70 140 210 280 350 Min: 19.99 / Avg: 142.48 / Max: 202.51 Min: 20.74 / Avg: 156.83 / Max: 210.59 Min: 40.69 / Avg: 293.85 / Max: 373.73 Min: 41.07 / Avg: 322.81 / Max: 382.87
miniBUDE CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better miniBUDE 20210901 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 19.67 / Avg: 268.1 / Max: 319.7 Min: 20.25 / Avg: 270.52 / Max: 323.64 Min: 39.97 / Avg: 463.17 / Max: 639.3 Min: 40.87 / Avg: 385.07 / Max: 627.6
miniBUDE Implementation: OpenMP - Input Deck: BM2 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Billion Interactions/s Per Watt, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM2 SMT Off SMT On 0.2135 0.427 0.6405 0.854 1.0675 0.881 0.883 0.949 0.819
miniBUDE CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better miniBUDE 20210901 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 19.77 / Avg: 154.01 / Max: 315.25 Min: 20.2 / Avg: 155.92 / Max: 319.62 Min: 40.2 / Avg: 223.4 / Max: 628.17 Min: 40.84 / Avg: 269.38 / Max: 547.76
miniBUDE Implementation: OpenMP - Input Deck: BM1 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Billion Interactions/s Per Watt, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM1 SMT Off SMT On 0.4421 0.8842 1.3263 1.7684 2.2105 1.524 1.525 1.965 0.940
miniBUDE Implementation: OpenMP - Input Deck: BM1 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Billion Interactions/s, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM1 SMT Off SMT On 100 200 300 400 500 SE +/- 0.10, N = 9 SE +/- 0.05, N = 9 SE +/- 7.21, N = 15 SE +/- 0.21, N = 9 234.68 237.76 439.07 253.12 1. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
miniBUDE Implementation: OpenMP - Input Deck: BM1 EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org GFInst/s, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM1 SMT Off SMT On 2K 4K 6K 8K 10K SE +/- 2.53, N = 9 SE +/- 1.30, N = 9 SE +/- 180.32, N = 15 SE +/- 5.35, N = 9 5867.11 5944.06 10976.68 6328.07 1. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
NAS Parallel Benchmarks CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.6 / Avg: 202.94 / Max: 285.46 Min: 20.01 / Avg: 210.74 / Max: 311.07 Min: 41.29 / Avg: 383.27 / Max: 632.69 Min: 40.07 / Avg: 394.92 / Max: 657.14
NAS Parallel Benchmarks Test / Class: SP.C EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.C SMT Off SMT On 140 280 420 560 700 657.40 625.95 602.82 567.82
NAS Parallel Benchmarks CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.52 / Avg: 119.85 / Max: 307.11 Min: 19.9 / Avg: 127.5 / Max: 326.83 Min: 39.68 / Avg: 227.42 / Max: 648.76 Min: 41.07 / Avg: 228.71 / Max: 655.77
NAS Parallel Benchmarks Test / Class: SP.B EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.B SMT Off SMT On 300 600 900 1200 1500 1347.26 1171.42 1082.90 1034.03
NAS Parallel Benchmarks Test / Class: SP.B EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.B SMT Off SMT On 50K 100K 150K 200K 250K SE +/- 885.17, N = 9 SE +/- 1110.64, N = 9 SE +/- 8884.57, N = 15 SE +/- 732.52, N = 9 161475.24 149355.54 246272.79 236490.76 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
miniFE CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better miniFE 2.2 CPU Power Consumption Monitor SMT Off SMT On 80 160 240 320 400 Min: 19.39 / Avg: 120.31 / Max: 246.66 Min: 20.05 / Avg: 123.43 / Max: 252.15 Min: 39.96 / Avg: 227.79 / Max: 468.68 Min: 39.79 / Avg: 233.98 / Max: 478.7
miniFE Problem Size: Small EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org CG Mflops Per Watt, More Is Better miniFE 2.2 Problem Size: Small SMT Off SMT On 90 180 270 360 450 430.07 419.54 236.18 268.29
NAS Parallel Benchmarks CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 19.66 / Avg: 91.16 / Max: 281.3 Min: 20.16 / Avg: 88.42 / Max: 270.94 Min: 40.59 / Avg: 172.91 / Max: 610.41 Min: 41.54 / Avg: 173.73 / Max: 526.19
NAS Parallel Benchmarks Test / Class: MG.C EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: MG.C SMT Off SMT On 300 600 900 1200 1500 1502.14 1449.06 1554.08 1433.88
NAS Parallel Benchmarks CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.72 / Avg: 192.37 / Max: 313.77 Min: 20.31 / Avg: 193.05 / Max: 313.96 Min: 40.61 / Avg: 310.37 / Max: 663.69 Min: 40.56 / Avg: 324.76 / Max: 663.39
NAS Parallel Benchmarks Test / Class: LU.C EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C SMT Off SMT On 500 1000 1500 2000 2500 1504.97 1448.63 2122.50 1821.35
NAS Parallel Benchmarks CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.77 / Avg: 150.15 / Max: 264.6 Min: 20.3 / Avg: 148.47 / Max: 271.57 Min: 40.25 / Avg: 307.61 / Max: 652.71 Min: 41.72 / Avg: 315.26 / Max: 633.61
NAS Parallel Benchmarks Test / Class: IS.D EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: IS.D SMT Off SMT On 8 16 24 32 40 35.40 35.70 28.07 31.24
NAS Parallel Benchmarks CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 19.73 / Avg: 133.34 / Max: 299.42 Min: 20.47 / Avg: 134.84 / Max: 312.66 Min: 40.17 / Avg: 249.94 / Max: 637.03 Min: 40.97 / Avg: 256.9 / Max: 647.79
NAS Parallel Benchmarks Test / Class: FT.C EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: FT.C SMT Off SMT On 200 400 600 800 1000 1105.80 1044.17 896.92 823.02
NAS Parallel Benchmarks CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.6 / Avg: 229.32 / Max: 332.77 Min: 20.22 / Avg: 238.61 / Max: 342.73 Min: 40.26 / Avg: 375.87 / Max: 680.44 Min: 41.42 / Avg: 399.82 / Max: 682.45
NAS Parallel Benchmarks Test / Class: EP.D EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.D SMT Off SMT On 15 30 45 60 75 62.25 55.59 69.13 59.29
NAS Parallel Benchmarks Test / Class: EP.D EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.D SMT Off SMT On 6K 12K 18K 24K 30K SE +/- 54.02, N = 5 SE +/- 214.86, N = 15 SE +/- 940.02, N = 12 SE +/- 413.32, N = 15 14274.53 13264.79 25983.74 23705.40 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.2
NAS Parallel Benchmarks CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor SMT Off SMT On 110 220 330 440 550 Min: 19.7 / Avg: 143.06 / Max: 313.46 Min: 20.14 / Avg: 149.82 / Max: 321.08 Min: 40.04 / Avg: 259.96 / Max: 632.64 Min: 41.39 / Avg: 274.22 / Max: 646.25
NAS Parallel Benchmarks Test / Class: CG.C EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: CG.C SMT Off SMT On 70 140 210 280 350 340.22 304.95 257.05 246.35
NAS Parallel Benchmarks CPU Power Consumption Monitor EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor SMT Off SMT On 120 240 360 480 600 Min: 19.51 / Avg: 216.16 / Max: 309.45 Min: 19.79 / Avg: 220.03 / Max: 326.76 Min: 40.15 / Avg: 380.97 / Max: 653.61 Min: 40.66 / Avg: 397.37 / Max: 663.29
NAS Parallel Benchmarks Test / Class: BT.C EPYC 9754 1P EPYC 9754 2P OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: BT.C SMT Off SMT On 300 600 900 1200 1500 1382.30 1328.18 1408.28 1236.19
Phoronix Test Suite v10.8.5