Google Cloud c3 Sapphire Rapids vs. AMD Milan

Benchmarks by Michael Larabel for a future article.

HTML result view exported from: https://openbenchmarking.org/result/2303280-NE-2303286NE24&sgm=1&ppd_YzMtaGlnaGNwdS04IFNQUg=0.3448&ppd_YzItc3RhbmRhcmQtOCBDTFg=0.4176&ppd_bjItc3RhbmRhcmQtOCBDTFg=0.388472&ppd_bjItaGlnaGNwdS04IENMWA=0.286784&ppd_dDJkLXN0YW5kYXJkLTggQU1E=0.337968&ppd_YzJkLWhpZ2hjcHUtOCBBTUQ=0.2998&ppt=DPH&sgm=1&sts=1&ppd_YzMtaGlnaGNwdS04IFNQUg=0.3448&ppd_YzItc3RhbmRhcmQtOCBDTFg=0.4176&ppd_bjItc3RhbmRhcmQtOCBDTFg=0.388472&ppd_bjItaGlnaGNwdS04IENMWA=0.286784&ppd_dDJkLXN0YW5kYXJkLTggQU1E=0.337968&ppd_YzJkLWhpZ2hjcHUtOCBBTUQ=0.2998&ppt=DPH&grs&sor&rro.

Google Cloud c3 Sapphire Rapids vs. AMD MilanProcessorMotherboardChipsetMemoryDiskNetworkOSKernelVulkanCompilerFile-SystemSystem Layerc3-highcpu-8 SPRc2-standard-8 CLXn2-standard-8 CLXn2-highcpu-8 CLXt2d-standard-8 AMDc2d-highcpu-8 AMDIntel Xeon Platinum 8481C (4 Cores / 8 Threads)Google Compute Engine c3-highcpu-8Intel 440FX 82441FX PMC16GB322GB nvme_card-pdGoogle Compute Engine VirtualUbuntu 22.105.19.0-1015-gcp (x86_64)1.3.224GCC 12.2.0ext4KVMIntel Xeon (4 Cores / 8 Threads)Google Compute Engine c2-standard-832GB322GB PersistentDiskRed Hat Virtio deviceGoogle Compute Engine n2-standard-8Google Compute Engine n2-highcpu-88GBAMD EPYC 7B13 (8 Cores)Google Compute Engine t2d-standard-832GBAMD EPYC 7B13 (4 Cores / 8 Threads)Google Compute Engine c2d-highcpu-816GBOpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler Details- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Processor Details- CPU Microcode: 0xffffffffPython Details- Python 3.10.7Security Details- c3-highcpu-8 SPR: 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 Enhanced IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected - c2-standard-8 CLX: itlb_multihit: Not affected + l1tf: Not affected + mds: Mitigation of Clear buffers; SMT Host state unknown + meltdown: Not affected + mmio_stale_data: Vulnerable: Clear buffers attempted no microcode; SMT Host state unknown + retbleed: Mitigation of Enhanced IBRS + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Mitigation of Clear buffers; SMT Host state unknown - n2-standard-8 CLX: itlb_multihit: Not affected + l1tf: Not affected + mds: Mitigation of Clear buffers; SMT Host state unknown + meltdown: Not affected + mmio_stale_data: Vulnerable: Clear buffers attempted no microcode; SMT Host state unknown + retbleed: Mitigation of Enhanced IBRS + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Mitigation of Clear buffers; SMT Host state unknown - n2-highcpu-8 CLX: itlb_multihit: Not affected + l1tf: Not affected + mds: Mitigation of Clear buffers; SMT Host state unknown + meltdown: Not affected + mmio_stale_data: Vulnerable: Clear buffers attempted no microcode; SMT Host state unknown + retbleed: Mitigation of Enhanced IBRS + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Mitigation of Clear buffers; SMT Host state unknown - t2d-standard-8 AMD: 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 - c2d-highcpu-8 AMD: 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: conditional RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Google Cloud c3 Sapphire Rapids vs. AMD Milanonednn: Convolution Batch Shapes Auto - bf16bf16bf16 - CPUonednn: Matrix Multiply Batch Shapes Transformer - bf16bf16bf16 - CPUopenssl: SHA256onednn: Deconvolution Batch shapes_1d - bf16bf16bf16 - CPUopenssl: AES-256-GCMopenssl: AES-128-GCMopenradioss: Rubber O-Ring Seal Installationopenradioss: Bumper Beamopenssl: SHA512ospray-studio: 3 - 4K - 1 - Path Traceropenradioss: Cell Phone Drop Testdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamminibude: OpenMP - BM1minibude: OpenMP - BM1brl-cad: VGR Performance Metricpgbench: 100 - 1000 - Read Only - Average Latencypgbench: 100 - 1000 - Read Onlyuvg266: Bosphorus 4K - Super Fastuvg266: Bosphorus 4K - Very Fastuvg266: Bosphorus 1080p - Super Fastuvg266: Bosphorus 1080p - Very Fastuvg266: Bosphorus 4K - Ultra Fastpgbench: 100 - 800 - Read Only - Average Latencypgbench: 100 - 800 - Read Onlyspecfem3d: Homogeneous Halfspaceopenradioss: Bird Strike on Windshieldgromacs: MPI CPU - water_GMX50_bareopencv: Object Detectionuvg266: Bosphorus 1080p - Ultra Fastspecfem3d: Tomographic Modeldeepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Streammemcached: 1:100openssl: RSA4096specfem3d: Mount St. Helensonednn: IP Shapes 1D - bf16bf16bf16 - CPUonednn: Deconvolution Batch shapes_3d - bf16bf16bf16 - CPUdeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamcompress-zstd: 19 - Decompression Speedmemcached: 1:10openfoam: drivaerFastback, Small Mesh Size - Execution Timeopenfoam: drivaerFastback, Small Mesh Size - Mesh Timespecfem3d: Water-layered Halfspacebuild-ffmpeg: Time To Compileincompact3d: input.i3d 129 Cells Per Directiondeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamoidn: RTLightmap.hdr.4096x4096oidn: RT.hdr_alb_nrm.3840x2160specfem3d: Layered Halfspacetensorflow: CPU - 64 - ResNet-50openssl: ChaCha20-Poly1305compress-zstd: 19, Long Mode - Decompression Speeddeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamjohn-the-ripper: Blowfishjohn-the-ripper: bcryptnekrs: TurboPipe Periodicbuild-linux-kernel: defconfignginx: 500deepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streamnginx: 1000john-the-ripper: MD5tensorflow: CPU - 32 - ResNet-50draco: Church Facadecompress-7zip: Compression Ratingnginx: 4000ospray-studio: 1 - 4K - 1 - Path Tracerembree: Pathtracer ISPC - Crowncockroach: KV, 95% Reads - 128nginx: 200lczero: BLASlczero: Eigentensorflow: CPU - 16 - ResNet-50namd: ATPase Simulation - 327,506 Atomsonednn: Recurrent Neural Network Training - bf16bf16bf16 - CPUnginx: 100onednn: IP Shapes 3D - bf16bf16bf16 - CPUembree: Pathtracer ISPC - Asian Dragondeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streamdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamonednn: Recurrent Neural Network Inference - bf16bf16bf16 - CPUopenvkl: vklBenchmark ISPCopencv: Image Processingmysqlslap: 4096cockroach: KV, 50% Reads - 128openssl: ChaCha20mysqlslap: 2048opencv: Stitchingdraco: Lionopencv: Graph APIdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamcompress-zstd: 19 - Compression Speedcompress-zstd: 19, Long Mode - Compression Speeddeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Streamblender: BMW27 - CPU-Onlyopencv: Corec3-highcpu-8 SPRc2-standard-8 CLXn2-standard-8 CLXn2-highcpu-8 CLXt2d-standard-8 AMDc2d-highcpu-8 AMD4.177070.96898642838739871.471454800836157357594077823367.00303.38156857292025819219.733.78733.7693188.6517.546710723.4052937257.486.9934.5032.399.122.565311942179.545901627595.140.7773899942.24143.93702287516.21941030937.282062.7139.5244696141.500043.5494433.1064905.21044947.13422.6783662.044277321.625251668120.43832.394316330.79520.120.24372.08948085715.6915970781140907.264.87476.53726930693230667900000244.80034672.6519.167732118.5876571314.9375733530632814.75209525.847524960.135602.101272122114.203.357794660.7036310.355.342187.3642104.3102305.89912337.319812816331719321.6220915576373322147606250219931528.0193530.610010.36.560.3858123.2931315.248737234.18937.43260131819353052.94211694500063023237312603523.45390.61146581522737892291.492.93082.9352152.7986.112503145.8951696436.035.6927.7226.307.454.616173317190.794140810724.800.5795805634.47150.91529263313.9932702291.931156.6145.37932851919.528735.916429.1783701.1715723.53560.6074485.518264347.812814070139.03937.757567133.29240.110.22374.75687720414.7910919151843713.859.99146.18936684668725100766667289.66221957.1715.917921446.2768354614.11114623098921594.94309113.934016184.224695.9190990213.334.445345767.7025148.288.024165.1542125.4746323.08282998.037014723423713684.5213468118132482508337437236186682.3748681.34499.245.8668.5076142.889214277038.87178.50678117851387360.59251512923800320765256453591.71436.27130815402761337322.222.65872.6597133.9515.358445636.4501551965.284.9724.1522.816.515.055158267213.892383575811.760.5286428430.05166.37762270512.7443624980.661028.6162.42533230822.025240.654225.9736649.5641713.60574.2360295.062233380.129762523154.89940.539009136.34470.100.20415.18307543413.749576513353660.354.95905.74915968597022963566667322.38120146.1015.362119683.6059963712.95123322806219947.56358843.458414976.822570.1881080812.185.067066078.8122830.198.518644.5768130.0183347.82993172.346317870420712699.1190551302172202784938355267250752.1925751.93838.245.1376.9661156.902620927738.96768.44515117733361360.55491513385492320753550177584.09433.87130622372061150320.362.62682.6570133.9505.358445566.8621457905.174.9324.1522.876.415.274152093215.274502309800.020.5296626929.97167.34805189112.8809604363.371028.8160.29615908022.035640.778725.6937640.8614742.62563.8655691.505362374.194312999154.90340.032824236.64260.090.20416.41313733813.709567474300650.054.51595.648159595962327.75519806.6015.113519418.0859897112.99128822809319477.31351933.471414884.721431.4277080112.185.055446112.6522444.408.626374.5994132.2178354.04873157.388117284221412635.1190654859032202813388391289162761.3379752.67638.295.0477.8099155.227820819968761277303144747272033948315373208.81176.83315865481730299135.996.16116.1956304.05312.162991683.11832082711.3610.6852.2549.1913.672.475323353102.336558318386.351.0983201461.8581.80646531826.05921224550.681765.081.18567235251.35411129.01200527.42294.344448.749592197.52049013380.64921.500117857.98850.150.30223.10985342810.48175926000771075.868.85518.1654106111058540756366667184.68335127.2623.808334079.6273821010.6187544844933592.77253065.492822148.835225.041202103910.743.108224004.6835257.776.4173167.6700489.07582107.397511441028019110.4263252869103011958236073210214645.4229642.979110.95.4277.7789153.15629044358241798831696893865318522423023364.33286.02186335206737227213.883.41943.3909186.0807.443674374.0962441988.147.9037.7936.459.823.206249490163.999040336621.880.6903766645.53128.16573835114.3549835405.061025.7125.62369412028.48951253.6842830.77403.8335463.91618287.423691806119.64026.521487652.41500.080.16324.8525828018.46117593264571183.838.12424.58427619762027115200000259.33524265.3113.546623758.474365858.5574423920323747.52313704.923120571.824677.038567438.664.457656503.6325019.865.8350147.4795436.25083335.706211761829217168.4172665177903051874005617199470584.5598588.613211.06.7170.1289139.202280771OpenBenchmarking.org

oneDNN

Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.0Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPUn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR918273645SE +/- 0.00800, N = 3SE +/- 0.03521, N = 3SE +/- 0.00489, N = 3SE +/- 0.00870, N = 338.9676038.8717034.189304.17707MIN: 38.58MIN: 38.45MIN: 33.771. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

oneDNN

Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.0Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPUn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR246810SE +/- 0.005782, N = 3SE +/- 0.006708, N = 3SE +/- 0.007542, N = 3SE +/- 0.013114, N = 38.5067808.4451507.4326000.968986MIN: 8.27MIN: 8.22MIN: 7.251. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

OpenSSL

Algorithm: SHA256

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: SHA256n2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD1500M3000M4500M6000M7500MSE +/- 191767.32, N = 3SE +/- 1532515.57, N = 3SE +/- 28303.98, N = 3SE +/- 2722265.85, N = 3SE +/- 2236930.48, N = 3SE +/- 1526902.57, N = 31177333613117851387313181935304283873987582417988368761277301. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

oneDNN

Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.0Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPUn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR1428425670SE +/- 0.04947, N = 3SE +/- 0.01683, N = 3SE +/- 0.00209, N = 3SE +/- 0.00472, N = 360.5925060.5549052.942101.47145MIN: 60.11MIN: 60.16MIN: 52.71. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

OpenSSL

Algorithm: AES-256-GCM

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: AES-256-GCMn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPR10000M20000M30000M40000M50000MSE +/- 5869541.98, N = 3SE +/- 6267687.89, N = 3SE +/- 3855504.36, N = 3SE +/- 2631212.92, N = 3SE +/- 31574149.68, N = 3SE +/- 48074910.93, N = 31512923800315133854923169450006301696893865331447472720480083615731. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenSSL

Algorithm: AES-128-GCM

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: AES-128-GCMc2d-highcpu-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR12000M24000M36000M48000M60000MSE +/- 281398.00, N = 3SE +/- 2887046.99, N = 3SE +/- 1982730.14, N = 3SE +/- 6165180.91, N = 3SE +/- 13818265.47, N = 3SE +/- 33921495.47, N = 31852242302320753550177207652564532323731260333948315373575940778231. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenRadioss

Model: Rubber O-Ring Seal Installation

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2022.10.13Model: Rubber O-Ring Seal Installationn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD130260390520650SE +/- 7.04, N = 3SE +/- 1.46, N = 3SE +/- 0.45, N = 3SE +/- 0.20, N = 3SE +/- 0.93, N = 3SE +/- 0.50, N = 3591.71584.09523.45367.00364.33208.81

OpenRadioss

Model: Bumper Beam

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2022.10.13Model: Bumper Beamn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD90180270360450SE +/- 2.04, N = 3SE +/- 2.21, N = 3SE +/- 0.90, N = 3SE +/- 0.51, N = 3SE +/- 0.29, N = 3SE +/- 0.34, N = 3436.27433.87390.61303.38286.02176.83

OpenSSL

Algorithm: SHA512

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: SHA512n2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD700M1400M2100M2800M3500MSE +/- 2213714.99, N = 3SE +/- 3181911.70, N = 3SE +/- 2965153.70, N = 3SE +/- 1152941.98, N = 3SE +/- 256893.72, N = 3SE +/- 1798677.56, N = 31306223720130815402714658152271568572920186335206731586548171. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OSPRay Studio

Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracern2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPR13K26K39K52K65KSE +/- 134.74, N = 3SE +/- 131.72, N = 3SE +/- 39.94, N = 3SE +/- 47.59, N = 3SE +/- 224.15, N = 3SE +/- 364.36, N = 36133761150378923722730299258191. (CXX) g++ options: -O3 -lm -ldl

OpenRadioss

Model: Cell Phone Drop Test

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2022.10.13Model: Cell Phone Drop Testn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD70140210280350SE +/- 0.36, N = 3SE +/- 0.68, N = 3SE +/- 0.39, N = 3SE +/- 0.38, N = 3SE +/- 2.62, N = 4SE +/- 0.30, N = 3322.22320.36291.49219.73213.88135.99

Neural Magic DeepSparse

Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Streamn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD246810SE +/- 0.0005, N = 3SE +/- 0.0040, N = 3SE +/- 0.0001, N = 3SE +/- 0.0030, N = 3SE +/- 0.0130, N = 3SE +/- 0.0107, N = 32.62682.65872.93083.41943.78736.1611

Neural Magic DeepSparse

Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Streamn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD246810SE +/- 0.0045, N = 3SE +/- 0.0025, N = 3SE +/- 0.0038, N = 3SE +/- 0.0121, N = 3SE +/- 0.0239, N = 3SE +/- 0.0070, N = 32.65702.65972.93523.39093.76936.1956

miniBUDE

Implementation: OpenMP - Input Deck: BM1

OpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1n2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD70140210280350SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3133.95133.95152.80186.08188.65304.051. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

miniBUDE

Implementation: OpenMP - Input Deck: BM1

OpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1n2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD3691215SE +/- 0.002, N = 3SE +/- 0.002, N = 3SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.001, N = 35.3585.3586.1127.4437.54612.1621. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

BRL-CAD

VGR Performance Metric

OpenBenchmarking.orgVGR Performance Metric, More Is BetterBRL-CAD 7.34VGR Performance Metricn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD20K40K60K80K100K4455644563503146743771072991681. (CXX) g++ options: -std=c++14 -pipe -fvisibility=hidden -fno-strict-aliasing -fno-common -fexceptions -ftemplate-depth-128 -m64 -ggdb3 -O3 -fipa-pta -fstrength-reduce -finline-functions -flto -ltcl8.6 -lregex_brl -lz_brl -lnetpbm -ldl -lm -ltk8.6

PostgreSQL

Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latency

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 15Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latencyn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD246810SE +/- 0.097, N = 3SE +/- 0.066, N = 12SE +/- 0.035, N = 3SE +/- 0.036, N = 3SE +/- 0.028, N = 3SE +/- 0.026, N = 36.8626.4505.8954.0963.4053.1181. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

PostgreSQL

Scaling Factor: 100 - Clients: 1000 - Mode: Read Only

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 15Scaling Factor: 100 - Clients: 1000 - Mode: Read Onlyn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD70K140K210K280K350KSE +/- 2071.56, N = 3SE +/- 1490.26, N = 12SE +/- 1014.66, N = 3SE +/- 2132.82, N = 3SE +/- 2414.33, N = 3SE +/- 2738.70, N = 31457901551961696432441982937253208271. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

uvg266

Video Input: Bosphorus 4K - Video Preset: Super Fast

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Super Fastn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD3691215SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 35.175.286.037.488.1411.36

uvg266

Video Input: Bosphorus 4K - Video Preset: Very Fast

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Very Fastn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD3691215SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 34.934.975.696.997.9010.68

uvg266

Video Input: Bosphorus 1080p - Video Preset: Super Fast

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Super Fastn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD1224364860SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.12, N = 324.1524.1527.7234.5037.7952.25

uvg266

Video Input: Bosphorus 1080p - Video Preset: Very Fast

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Very Fastn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD1122334455SE +/- 0.15, N = 3SE +/- 0.12, N = 3SE +/- 0.16, N = 3SE +/- 0.25, N = 3SE +/- 0.37, N = 3SE +/- 0.51, N = 322.8122.8726.3032.3936.4549.19

uvg266

Video Input: Bosphorus 4K - Video Preset: Ultra Fast

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Ultra Fastn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD48121620SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.07, N = 36.416.517.459.129.8213.67

PostgreSQL

Scaling Factor: 100 - Clients: 800 - Mode: Read Only - Average Latency

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 15Scaling Factor: 100 - Clients: 800 - Mode: Read Only - Average Latencyn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD1.18672.37343.56014.74685.9335SE +/- 0.088, N = 12SE +/- 0.008, N = 3SE +/- 0.022, N = 3SE +/- 0.009, N = 3SE +/- 0.028, N = 3SE +/- 0.035, N = 35.2745.0554.6163.2062.5652.4751. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

PostgreSQL

Scaling Factor: 100 - Clients: 800 - Mode: Read Only

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 15Scaling Factor: 100 - Clients: 800 - Mode: Read Onlyn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD70K140K210K280K350KSE +/- 2195.04, N = 12SE +/- 259.80, N = 3SE +/- 822.93, N = 3SE +/- 664.60, N = 3SE +/- 3369.27, N = 3SE +/- 4510.85, N = 31520931582671733172494903119423233531. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

SPECFEM3D

Model: Homogeneous Halfspace

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Homogeneous Halfspacen2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD50100150200250SE +/- 2.61, N = 3SE +/- 1.45, N = 3SE +/- 1.87, N = 3SE +/- 0.09, N = 3SE +/- 0.71, N = 3SE +/- 0.29, N = 3215.27213.89190.79179.55164.00102.341. (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

OpenRadioss

Model: Bird Strike on Windshield

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2022.10.13Model: Bird Strike on Windshieldn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD2004006008001000SE +/- 5.07, N = 3SE +/- 1.32, N = 3SE +/- 1.51, N = 3SE +/- 0.64, N = 3SE +/- 0.30, N = 3SE +/- 0.51, N = 3811.76800.02724.80621.88595.14386.35

GROMACS

Implementation: MPI CPU - Input: water_GMX50_bare

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2023Implementation: MPI CPU - Input: water_GMX50_baren2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.24710.49420.74130.98841.2355SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.009, N = 30.5280.5290.5790.6900.7771.0981. (CXX) g++ options: -O3

OpenCV

Test: Object Detection

OpenBenchmarking.orgms, Fewer Is BetterOpenCV 4.7Test: Object Detectionn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD14K28K42K56K70KSE +/- 736.69, N = 4SE +/- 842.44, N = 3SE +/- 751.87, N = 3SE +/- 384.74, N = 5SE +/- 509.28, N = 3SE +/- 424.20, N = 156626964284580563899937666320141. (CXX) g++ options: -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -ldl -lm -lpthread -lrt

uvg266

Video Input: Bosphorus 1080p - Video Preset: Ultra Fast

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Ultra Fastn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD1428425670SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.08, N = 3SE +/- 0.33, N = 329.9730.0534.4742.2445.5361.85

SPECFEM3D

Model: Tomographic Model

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Tomographic Modeln2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD4080120160200SE +/- 1.83, N = 3SE +/- 0.66, N = 3SE +/- 1.30, N = 3SE +/- 1.65, N = 3SE +/- 0.71, N = 3SE +/- 0.35, N = 3167.35166.38150.92143.94128.1781.811. (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

Neural Magic DeepSparse

Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Streamn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD612182430SE +/- 0.07, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.13, N = 3SE +/- 0.06, N = 312.7412.8813.9914.3516.2226.06

Memcached

Set To Get Ratio: 1:100

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.18Set To Get Ratio: 1:100n2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD300K600K900K1200K1500KSE +/- 1810.97, N = 3SE +/- 4342.69, N = 3SE +/- 3476.80, N = 3SE +/- 2694.56, N = 3SE +/- 11157.73, N = 3SE +/- 15912.40, N = 3604363.37624980.66702291.93835405.061030937.281224550.681. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

OpenSSL

Algorithm: RSA4096

OpenBenchmarking.orgsign/s, More Is BetterOpenSSL 3.1Algorithm: RSA4096c2d-highcpu-8 AMDn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR400800120016002000SE +/- 0.28, N = 3SE +/- 2.24, N = 3SE +/- 3.47, N = 3SE +/- 2.05, N = 3SE +/- 0.38, N = 3SE +/- 1.17, N = 31025.71028.61028.81156.61765.02062.71. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

SPECFEM3D

Model: Mount St. Helens

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Mount St. Helensn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD4080120160200SE +/- 0.57, N = 3SE +/- 0.25, N = 3SE +/- 0.56, N = 3SE +/- 0.10, N = 3SE +/- 0.89, N = 3SE +/- 0.46, N = 3162.43160.30145.38139.52125.6281.191. (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

oneDNN

Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.0Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPUn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR510152025SE +/- 0.02769, N = 3SE +/- 0.02550, N = 3SE +/- 0.01518, N = 3SE +/- 0.00340, N = 322.0356022.0252019.528701.50004MIN: 21.37MIN: 21.41MIN: 18.981. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

oneDNN

Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.0Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPUn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR918273645SE +/- 0.03843, N = 3SE +/- 0.03334, N = 3SE +/- 0.01009, N = 3SE +/- 0.01072, N = 340.7787040.6542035.916403.54944MIN: 40.38MIN: 40.35MIN: 35.711. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

Neural Magic DeepSparse

Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Streamn2-highcpu-8 CLXn2-standard-8 CLXc2d-highcpu-8 AMDc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD1224364860SE +/- 0.20, N = 3SE +/- 0.12, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.15, N = 3SE +/- 0.02, N = 325.6925.9728.4929.1833.1151.35

Zstd Compression

Compression Level: 19 - Decompression Speed

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19 - Decompression Speedn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD30060090012001500SE +/- 1.15, N = 5SE +/- 3.67, N = 3SE +/- 1.45, N = 3SE +/- 1.50, N = 3SE +/- 4.02, N = 3SE +/- 3.36, N = 3640.8649.5701.1905.21129.01253.6-llzma1. (CC) gcc options: -O3 -pthread -lz

Memcached

Set To Get Ratio: 1:10

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.18Set To Get Ratio: 1:10n2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD300K600K900K1200K1500KSE +/- 2135.86, N = 3SE +/- 3190.98, N = 3SE +/- 2213.84, N = 3SE +/- 5221.50, N = 3SE +/- 3280.78, N = 3SE +/- 10218.36, N = 3614742.62641713.60715723.53842830.771044947.131200527.421. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

OpenFOAM

Input: drivaerFastback, Small Mesh Size - Execution Time

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Small Mesh Size - Execution Timen2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD120240360480600574.24563.87560.61422.68403.83294.34-lphysicalProperties -lspecie -lfiniteVolume -lfvModels -lmeshTools -lsampling-ldynamicMesh-ldynamicMesh-lphysicalProperties -lspecie -lfiniteVolume -lfvModels -lmeshTools -lsampling-lphysicalProperties -lspecie -lfiniteVolume -lfvModels -lmeshTools -lsampling-lphysicalProperties -lspecie -lfiniteVolume -lfvModels -lmeshTools -lsampling1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lgenericPatchFields -lOpenFOAM -ldl -lm

OpenFOAM

Input: drivaerFastback, Small Mesh Size - Mesh Time

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Small Mesh Size - Mesh Timen2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD2040608010095.0691.5185.5263.9262.0448.75-lphysicalProperties -lspecie -lfiniteVolume -lfvModels -lmeshTools -lsampling-ldynamicMesh-ldynamicMesh-lphysicalProperties -lspecie -lfiniteVolume -lfvModels -lmeshTools -lsampling-lphysicalProperties -lspecie -lfiniteVolume -lfvModels -lmeshTools -lsampling-lphysicalProperties -lspecie -lfiniteVolume -lfvModels -lmeshTools -lsampling1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lgenericPatchFields -lOpenFOAM -ldl -lm

SPECFEM3D

Model: Water-layered Halfspace

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Water-layered Halfspacen2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD80160240320400SE +/- 1.21, N = 3SE +/- 2.86, N = 3SE +/- 0.85, N = 3SE +/- 0.31, N = 3SE +/- 0.44, N = 3SE +/- 1.86, N = 12380.13374.19347.81321.63287.42197.521. (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

Timed FFmpeg Compilation

Time To Compile

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed FFmpeg Compilation 6.0Time To Compilen2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD306090120150SE +/- 0.08, N = 3SE +/- 1.17, N = 3SE +/- 0.03, N = 3SE +/- 0.10, N = 3SE +/- 0.01, N = 3SE +/- 0.17, N = 3154.90154.90139.04120.44119.6480.65

Xcompact3d Incompact3d

Input: input.i3d 129 Cells Per Direction

OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 129 Cells Per Directionn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD918273645SE +/- 0.12, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.24, N = 440.5440.0337.7632.3926.5221.501. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

Neural Magic DeepSparse

Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Streamt2d-standard-8 AMDc2d-highcpu-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR1326395265SE +/- 0.08, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 357.9952.4236.6436.3433.2930.80

Intel Open Image Denoise

Run: RTLightmap.hdr.4096x4096

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.4.0Run: RTLightmap.hdr.4096x4096c2d-highcpu-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD0.03380.06760.10140.13520.169SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.080.090.100.110.120.15

Intel Open Image Denoise

Run: RT.hdr_alb_nrm.3840x2160

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.4.0Run: RT.hdr_alb_nrm.3840x2160c2d-highcpu-8 AMDn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD0.06750.1350.20250.270.3375SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.160.200.200.220.240.30

SPECFEM3D

Model: Layered Halfspace

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.0Model: Layered Halfspacen2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD90180270360450SE +/- 0.99, N = 3SE +/- 0.67, N = 3SE +/- 2.51, N = 3SE +/- 0.64, N = 3SE +/- 1.43, N = 3SE +/- 3.19, N = 3416.41415.18374.76372.09324.85223.111. (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

TensorFlow

Device: CPU - Batch Size: 64 - Model: ResNet-50

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.10Device: CPU - Batch Size: 64 - Model: ResNet-50c2d-highcpu-8 AMDt2d-standard-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR48121620SE +/- 0.00, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.18, N = 3SE +/- 0.01, N = 38.4610.4813.7013.7414.7915.69

OpenSSL

Algorithm: ChaCha20-Poly1305

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: ChaCha20-Poly1305n2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD4000M8000M12000M16000M20000MSE +/- 3854876.66, N = 3SE +/- 2582153.15, N = 3SE +/- 1592104.52, N = 3SE +/- 378013.69, N = 3SE +/- 21990039.72, N = 3SE +/- 35085963.78, N = 395674743009576513353109191518431175932645715970781140175926000771. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

Zstd Compression

Compression Level: 19, Long Mode - Decompression Speed

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19, Long Mode - Decompression Speedn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD30060090012001500SE +/- 0.09, N = 3SE +/- 1.88, N = 15SE +/- 2.02, N = 3SE +/- 1.48, N = 3SE +/- 7.21, N = 3SE +/- 9.51, N = 3650.0660.3713.8907.21075.81183.8-llzma1. (CC) gcc options: -O3 -pthread -lz

Neural Magic DeepSparse

Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Streamc2d-highcpu-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD1530456075SE +/- 0.04, N = 3SE +/- 0.07, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.10, N = 338.1254.5254.9659.9964.8768.86

Neural Magic DeepSparse

Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Streamc2d-highcpu-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD246810SE +/- 0.0013, N = 3SE +/- 0.0036, N = 3SE +/- 0.0137, N = 3SE +/- 0.0034, N = 3SE +/- 0.0078, N = 3SE +/- 0.0795, N = 34.58425.64815.74916.18936.53728.1654

John The Ripper

Test: Blowfish

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: Blowfishn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD2K4K6K8K10KSE +/- 1.45, N = 3SE +/- 1.00, N = 3SE +/- 2.73, N = 3SE +/- 2.08, N = 3SE +/- 1.67, N = 3SE +/- 13.28, N = 359595968668469307619106111. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt

John The Ripper

Test: bcrypt

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: bcryptn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD2K4K6K8K10KSE +/- 1.45, N = 3SE +/- 0.67, N = 3SE +/- 0.67, N = 3SE +/- 0.67, N = 3SE +/- 1.00, N = 3SE +/- 8.14, N = 359625970668769327620105851. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt

nekRS

Input: TurboPipe Periodic

OpenBenchmarking.orgFLOP/s, More Is BetternekRS 22.0Input: TurboPipe Periodicn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD9000M18000M27000M36000M45000MSE +/- 62364769.26, N = 3SE +/- 56849518.71, N = 3SE +/- 48942108.66, N = 3SE +/- 92013205.57, N = 3SE +/- 64451851.11, N = 322963566667251007666672711520000030667900000407563666671. (CXX) g++ options: -fopenmp -O2 -march=native -mtune=native -ftree-vectorize -lmpi_cxx -lmpi

Timed Linux Kernel Compilation

Build: defconfig

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.1Build: defconfign2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD70140210280350SE +/- 0.77, N = 3SE +/- 1.15, N = 3SE +/- 0.85, N = 3SE +/- 0.57, N = 3SE +/- 0.64, N = 3SE +/- 0.35, N = 3327.76322.38289.66259.34244.80184.68

nginx

Connections: 500

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 500n2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD8K16K24K32K40KSE +/- 34.93, N = 3SE +/- 51.39, N = 3SE +/- 17.44, N = 3SE +/- 81.15, N = 3SE +/- 321.85, N = 3SE +/- 56.19, N = 319806.6020146.1021957.1724265.3134672.6535127.261. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

Neural Magic DeepSparse

Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Streamc2d-highcpu-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD612182430SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 313.5515.1115.3615.9219.1723.81

nginx

Connections: 1000

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 1000n2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD7K14K21K28K35KSE +/- 102.00, N = 3SE +/- 19.22, N = 3SE +/- 84.22, N = 3SE +/- 10.88, N = 3SE +/- 22.42, N = 3SE +/- 47.86, N = 319418.0819683.6021446.2723758.4732118.5834079.621. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

John The Ripper

Test: MD5

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: MD5c2d-highcpu-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR160K320K480K640K800KSE +/- 56.60, N = 3SE +/- 168.31, N = 3SE +/- 266.70, N = 3SE +/- 162.53, N = 3SE +/- 385.43, N = 3SE +/- 1472.13, N = 34365855989715996376835467382107657131. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt

TensorFlow

Device: CPU - Batch Size: 32 - Model: ResNet-50

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.10Device: CPU - Batch Size: 32 - Model: ResNet-50c2d-highcpu-8 AMDt2d-standard-8 AMDn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR48121620SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.07, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 38.5510.6112.9512.9914.1114.93

Google Draco

Model: Church Facade

OpenBenchmarking.orgms, Fewer Is BetterGoogle Draco 1.5.6Model: Church Facaden2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPRc2d-highcpu-8 AMD3K6K9K12K15KSE +/- 141.00, N = 4SE +/- 21.39, N = 3SE +/- 16.76, N = 3SE +/- 40.60, N = 3SE +/- 10.48, N = 3SE +/- 34.49, N = 31288212332114628754757374421. (CXX) g++ options: -O3

7-Zip Compression

Test: Compression Rating

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Compression Ratingn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD10K20K30K40K50KSE +/- 125.77, N = 3SE +/- 324.52, N = 4SE +/- 248.15, N = 3SE +/- 246.17, N = 15SE +/- 371.22, N = 3SE +/- 126.17, N = 32806228093309893530639203484491. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

nginx

Connections: 4000

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 4000n2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD7K14K21K28K35KSE +/- 58.80, N = 3SE +/- 55.22, N = 3SE +/- 11.52, N = 3SE +/- 84.28, N = 3SE +/- 27.93, N = 3SE +/- 461.85, N = 319477.3119947.5621594.9423747.5232814.7533592.771. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

OSPRay Studio

Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracern2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR8K16K24K32K40KSE +/- 166.18, N = 3SE +/- 93.88, N = 3SE +/- 54.89, N = 3SE +/- 32.33, N = 3SE +/- 93.60, N = 3SE +/- 4.91, N = 33588435193313703091125306209521. (CXX) g++ options: -O3 -lm -ldl

Embree

Binary: Pathtracer ISPC - Model: Crown

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.0.1Binary: Pathtracer ISPC - Model: Crownn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPR1.31572.63143.94715.26286.5785SE +/- 0.0212, N = 3SE +/- 0.0114, N = 3SE +/- 0.0030, N = 3SE +/- 0.0095, N = 3SE +/- 0.0419, N = 3SE +/- 0.0120, N = 33.45843.47143.93404.92315.49285.8475MIN: 3.38 / MAX: 3.58MIN: 3.4 / MAX: 3.58MIN: 3.91 / MAX: 3.99MIN: 4.87 / MAX: 5.01MIN: 5.31 / MAX: 5.65MIN: 5.81 / MAX: 5.92

CockroachDB

Workload: KV, 95% Reads - Concurrency: 128

OpenBenchmarking.orgops/s, More Is BetterCockroachDB 22.2Workload: KV, 95% Reads - Concurrency: 128n2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPR5K10K15K20K25KSE +/- 99.60, N = 3SE +/- 39.57, N = 3SE +/- 79.71, N = 3SE +/- 44.80, N = 3SE +/- 53.72, N = 3SE +/- 127.78, N = 314884.714976.816184.220571.822148.824960.1

nginx

Connections: 200

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 200n2-highcpu-8 CLXn2-standard-8 CLXc2d-highcpu-8 AMDc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR8K16K24K32K40KSE +/- 136.11, N = 3SE +/- 116.88, N = 3SE +/- 298.83, N = 3SE +/- 37.62, N = 3SE +/- 259.52, N = 3SE +/- 83.52, N = 321431.4222570.1824677.0324695.9135225.0435602.101. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

LeelaChessZero

Backend: BLAS

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: BLASn2-highcpu-8 CLXn2-standard-8 CLXc2d-highcpu-8 AMDc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR30060090012001500SE +/- 6.44, N = 3SE +/- 3.18, N = 3SE +/- 4.70, N = 3SE +/- 7.36, N = 3SE +/- 1.45, N = 3SE +/- 16.59, N = 3770810856909120212721. (CXX) g++ options: -flto -pthread

LeelaChessZero

Backend: Eigen

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: Eigenc2d-highcpu-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR30060090012001500SE +/- 3.48, N = 3SE +/- 8.29, N = 3SE +/- 8.78, N = 5SE +/- 9.00, N = 3SE +/- 3.06, N = 3SE +/- 7.69, N = 3743801808902103912211. (CXX) g++ options: -flto -pthread

TensorFlow

Device: CPU - Batch Size: 16 - Model: ResNet-50

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.10Device: CPU - Batch Size: 16 - Model: ResNet-50c2d-highcpu-8 AMDt2d-standard-8 AMDn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR48121620SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.08, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 38.6610.7412.1812.1813.3314.20

NAMD

ATPase Simulation - 327,506 Atoms

OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.14ATPase Simulation - 327,506 Atomsn2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD1.14012.28023.42034.56045.7005SE +/- 0.03361, N = 3SE +/- 0.01808, N = 3SE +/- 0.00235, N = 3SE +/- 0.00771, N = 3SE +/- 0.00254, N = 3SE +/- 0.00017, N = 35.067065.055444.457654.445343.357793.10822

oneDNN

Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.0Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPUc2d-highcpu-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD14002800420056007000SE +/- 4.35, N = 3SE +/- 9.18, N = 3SE +/- 2.66, N = 3SE +/- 8.38, N = 3SE +/- 0.86, N = 3SE +/- 17.70, N = 36503.636112.656078.815767.704660.704004.68MIN: 6453.16MIN: 6059.51MIN: 6032.59MIN: 5732.12MIN: 4648.25MIN: 3904.561. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

nginx

Connections: 100

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 100n2-highcpu-8 CLXn2-standard-8 CLXc2d-highcpu-8 AMDc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR8K16K24K32K40KSE +/- 51.88, N = 3SE +/- 51.85, N = 3SE +/- 22.78, N = 3SE +/- 33.32, N = 3SE +/- 123.11, N = 3SE +/- 23.95, N = 322444.4022830.1925019.8625148.2835257.7736310.351. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2

oneDNN

Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.0Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPUn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR246810SE +/- 0.03681, N = 3SE +/- 0.02441, N = 3SE +/- 0.01190, N = 3SE +/- 0.00535, N = 38.626378.518648.024165.34218MIN: 8.15MIN: 8.13MIN: 7.86MIN: 4.941. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

Embree

Binary: Pathtracer ISPC - Model: Asian Dragon

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.0.1Binary: Pathtracer ISPC - Model: Asian Dragonn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPR246810SE +/- 0.0155, N = 3SE +/- 0.0079, N = 3SE +/- 0.0119, N = 3SE +/- 0.0077, N = 3SE +/- 0.0540, N = 3SE +/- 0.0079, N = 34.57684.59945.15425.83506.41737.3642MIN: 4.49 / MAX: 4.73MIN: 4.54 / MAX: 4.72MIN: 5.12 / MAX: 5.22MIN: 5.78 / MAX: 5.94MIN: 6.23 / MAX: 6.66MIN: 7.33 / MAX: 7.44

Neural Magic DeepSparse

Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Streamt2d-standard-8 AMDc2d-highcpu-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR4080120160200SE +/- 0.31, N = 3SE +/- 0.37, N = 3SE +/- 0.19, N = 3SE +/- 0.52, N = 3SE +/- 0.40, N = 3SE +/- 0.14, N = 3167.67147.48132.22130.02125.47104.31

Neural Magic DeepSparse

Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Streamt2d-standard-8 AMDc2d-highcpu-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR110220330440550SE +/- 4.26, N = 3SE +/- 0.12, N = 3SE +/- 0.22, N = 3SE +/- 0.83, N = 3SE +/- 0.18, N = 3SE +/- 0.37, N = 3489.08436.25354.05347.83323.08305.90

oneDNN

Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.0Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPUc2d-highcpu-8 AMDn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD7001400210028003500SE +/- 3.60, N = 3SE +/- 31.87, N = 3SE +/- 3.96, N = 3SE +/- 0.83, N = 3SE +/- 1.76, N = 3SE +/- 7.71, N = 33335.703172.343157.382998.032337.312107.39MIN: 3294.12MIN: 3085.07MIN: 3110.9MIN: 2977.98MIN: 2326.77MIN: 2012.961. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

OpenVKL

Benchmark: vklBenchmark ISPC

OpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 1.3.1Benchmark: vklBenchmark ISPCc2d-highcpu-8 AMDn2-standard-8 CLXc2-standard-8 CLXt2d-standard-8 AMDn2-highcpu-8 CLXc3-highcpu-8 SPR20406080100SE +/- 0.58, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.58, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3626370758198MIN: 7 / MAX: 923MIN: 7 / MAX: 1024MIN: 8 / MAX: 1119MIN: 9 / MAX: 997MIN: 8 / MAX: 1030MIN: 11 / MAX: 1579

OpenCV

Test: Image Processing

OpenBenchmarking.orgms, Fewer Is BetterOpenCV 4.7Test: Image Processingn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD40K80K120K160K200KSE +/- 2291.38, N = 3SE +/- 2082.92, N = 3SE +/- 1527.37, N = 4SE +/- 1624.35, N = 12SE +/- 1358.27, N = 3SE +/- 1246.60, N = 41787041728421472341281631176181144101. (CXX) g++ options: -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -ldl -lm -lpthread -lrt

MariaDB

Clients: 4096

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.0.1Clients: 4096n2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc2d-highcpu-8 AMDc3-highcpu-8 SPR70140210280350SE +/- 2.13, N = 3SE +/- 1.01, N = 3SE +/- 2.55, N = 3SE +/- 2.99, N = 4SE +/- 1.97, N = 3SE +/- 2.62, N = 32072142372802923171. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lnuma -lcrypt -lz -lm -lssl -lcrypto -lpthread -ldl

CockroachDB

Workload: KV, 50% Reads - Concurrency: 128

OpenBenchmarking.orgops/s, More Is BetterCockroachDB 22.2Workload: KV, 50% Reads - Concurrency: 128n2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPR4K8K12K16K20KSE +/- 40.52, N = 3SE +/- 70.13, N = 3SE +/- 68.65, N = 3SE +/- 73.34, N = 3SE +/- 5.78, N = 3SE +/- 40.86, N = 312635.112699.113684.517168.419110.419321.6

OpenSSL

Algorithm: ChaCha20

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: ChaCha20c2d-highcpu-8 AMDn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD6000M12000M18000M24000M30000MSE +/- 9826068.49, N = 3SE +/- 15025960.55, N = 3SE +/- 2986147.97, N = 3SE +/- 1497684.67, N = 3SE +/- 35781789.46, N = 3SE +/- 5497165.35, N = 31726651779019055130217190654859032134681181322091557637263252869101. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

MariaDB

Clients: 2048

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.0.1Clients: 2048n2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc2d-highcpu-8 AMDc3-highcpu-8 SPR70140210280350SE +/- 2.04, N = 7SE +/- 1.89, N = 3SE +/- 3.50, N = 3SE +/- 7.31, N = 6SE +/- 2.57, N = 3SE +/- 3.01, N = 32202202483013053321. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lnuma -lcrypt -lz -lm -lssl -lcrypto -lpthread -ldl

OpenCV

Test: Stitching

OpenBenchmarking.orgms, Fewer Is BetterOpenCV 4.7Test: Stitchingn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD60K120K180K240K300KSE +/- 2338.09, N = 3SE +/- 2268.99, N = 3SE +/- 1856.06, N = 3SE +/- 1973.06, N = 7SE +/- 2789.72, N = 3SE +/- 1971.58, N = 52813382784932508332147601958231874001. (CXX) g++ options: -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -ldl -lm -lpthread -lrt

Google Draco

Model: Lion

OpenBenchmarking.orgms, Fewer Is BetterGoogle Draco 1.5.6Model: Lionn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD2K4K6K8K10KSE +/- 21.31, N = 3SE +/- 75.43, N = 3SE +/- 18.52, N = 3SE +/- 80.23, N = 15SE +/- 32.94, N = 3SE +/- 18.18, N = 38391835574376250607356171. (CXX) g++ options: -O3

OpenCV

Test: Graph API

OpenBenchmarking.orgms, Fewer Is BetterOpenCV 4.7Test: Graph APIn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD60K120K180K240K300KSE +/- 3363.59, N = 9SE +/- 727.87, N = 3SE +/- 1570.24, N = 3SE +/- 931.36, N = 3SE +/- 529.01, N = 3SE +/- 2222.25, N = 32891622672502361862199312102141994701. (CXX) g++ options: -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -ldl -lm -lpthread -lrt

Neural Magic DeepSparse

Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Streamn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc2d-highcpu-8 AMDc3-highcpu-8 SPR160320480640800SE +/- 0.15, N = 3SE +/- 1.14, N = 3SE +/- 0.03, N = 3SE +/- 0.74, N = 3SE +/- 0.39, N = 3SE +/- 1.78, N = 3761.34752.19682.37645.42584.56528.02

Neural Magic DeepSparse

Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Streamn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc2d-highcpu-8 AMDc3-highcpu-8 SPR160320480640800SE +/- 1.28, N = 3SE +/- 0.71, N = 3SE +/- 0.89, N = 3SE +/- 0.40, N = 3SE +/- 2.04, N = 3SE +/- 3.34, N = 3752.68751.94681.34642.98588.61530.61

Zstd Compression

Compression Level: 19 - Compression Speed

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19 - Compression Speedn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD3691215SE +/- 0.06, N = 3SE +/- 0.08, N = 5SE +/- 0.08, N = 3SE +/- 0.12, N = 3SE +/- 0.12, N = 3SE +/- 0.06, N = 38.248.299.2410.3010.9011.00-llzma1. (CC) gcc options: -O3 -pthread -lz

Zstd Compression

Compression Level: 19, Long Mode - Compression Speed

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19, Long Mode - Compression Speedn2-highcpu-8 CLXn2-standard-8 CLXt2d-standard-8 AMDc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMD246810SE +/- 0.00, N = 3SE +/- 0.04, N = 15SE +/- 0.06, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.04, N = 35.045.135.425.866.506.71-llzma1. (CC) gcc options: -O3 -pthread -lz

Neural Magic DeepSparse

Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Streamn2-highcpu-8 CLXt2d-standard-8 AMDn2-standard-8 CLXc2d-highcpu-8 AMDc2-standard-8 CLXc3-highcpu-8 SPR20406080100SE +/- 0.60, N = 3SE +/- 0.02, N = 3SE +/- 0.37, N = 3SE +/- 0.07, N = 3SE +/- 0.08, N = 3SE +/- 0.28, N = 377.8177.7876.9770.1368.5160.39

Neural Magic DeepSparse

Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Streamn2-standard-8 CLXn2-highcpu-8 CLXt2d-standard-8 AMDc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPR306090120150SE +/- 0.88, N = 3SE +/- 0.06, N = 3SE +/- 0.34, N = 3SE +/- 0.29, N = 3SE +/- 0.08, N = 3SE +/- 0.95, N = 3156.90155.23153.16142.89139.20123.29

Blender

Blend File: BMW27 - Compute: CPU-Only

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.4Blend File: BMW27 - Compute: CPU-Onlyc3-highcpu-8 SPR70140210280350SE +/- 1.13, N = 3315.24

OpenCV

Test: Core

OpenBenchmarking.orgms, Fewer Is BetterOpenCV 4.7Test: Coren2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPRc2d-highcpu-8 AMD40K80K120K160K200KSE +/- 3894.18, N = 12SE +/- 2878.10, N = 9SE +/- 2578.21, N = 12SE +/- 3138.88, N = 15SE +/- 280.31, N = 3SE +/- 678.61, N = 32092772081991427709044387372807711. (CXX) g++ options: -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -ldl -lm -lpthread -lrt

Geometric Mean Of All Test Results

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgGeometric Mean, More Is BetterGeometric Mean Of All Test ResultsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD142842567037.3637.5641.9647.0554.7662.19

Geometric Mean Of Timed Code Compilation Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Timed Code Compilation TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.41510.83021.24531.66042.07551.0001.0071.1221.2791.3111.845

Geometric Mean Of C/C++ Compiler Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of C/C++ Compiler TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.41270.82541.23811.65082.06351.0001.0151.1241.3501.6481.834

Geometric Mean Of Compression Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Compression TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD0.34740.69481.04221.38961.7371.0001.0081.1141.3181.4801.544

Geometric Mean Of CPU Massive Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of CPU Massive TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.36860.73721.10581.47441.8431.0001.0121.1201.2271.5701.638

Geometric Mean Of Creator Workloads Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Creator Workloads TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.38480.76961.15441.53921.9241.0001.0041.1431.2911.4841.710

Geometric Mean Of Cryptography Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Cryptography TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.43990.87981.31971.75962.19951.0001.0011.1251.2501.7421.955

Geometric Mean Of Database Test Suite

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Database Test SuiteResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.40410.80821.21231.61642.02051.0001.0261.1341.4821.7601.796

Geometric Mean Of Fortran Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Fortran TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD0.44190.88381.32571.76762.20951.0001.0041.1021.1751.3341.964

Geometric Mean Of Game Development Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Game Development TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD0.31230.62460.93691.24921.56151.0001.0181.0911.1031.3761.388

Geometric Mean Of HPC - High Performance Computing Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of HPC - High Performance Computing TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.34090.68181.02271.36361.70451.0001.0041.1141.1491.3671.515

Geometric Mean Of Common Kernel Benchmarks Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Common Kernel Benchmarks TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.48420.96841.45261.93682.4211.0001.0191.1341.4492.0412.152

Geometric Mean Of Machine Learning Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Machine Learning TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2d-highcpu-8 AMDc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR0.30220.60440.90661.20881.5111.0001.0101.0571.1201.3121.343

Geometric Mean Of Molecular Dynamics Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Molecular Dynamics TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.42620.85241.27861.70482.1311.0001.0151.0891.3691.4211.894

Geometric Mean Of MPI Benchmarks Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of MPI Benchmarks TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD0.44570.89141.33711.78282.22851.0001.0081.0851.3581.4141.981

Geometric Mean Of Multi-Core Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Multi-Core TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.38250.7651.14751.531.91251.0001.0031.1331.3161.4661.700

Geometric Mean Of NVIDIA GPU Compute Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of NVIDIA GPU Compute TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2d-highcpu-8 AMDc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD0.36320.72641.08961.45281.8161.0001.0191.1041.1331.5461.614

Geometric Mean Of Intel oneAPI Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Intel oneAPI TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR0.34270.68541.02811.37081.71351.0001.0201.0781.1571.5101.523

Geometric Mean Of OpenMPI Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of OpenMPI TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD0.47410.94821.42231.89642.37051.0001.0101.1021.3241.3972.107

Geometric Mean Of Programmer / Developer System Benchmarks Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Programmer / Developer System Benchmarks TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD0.34940.69881.04821.39761.7471.0001.0101.1191.3261.4751.553

Geometric Mean Of Python Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Python TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milanc2d-highcpu-8 AMDn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR0.30350.6070.91051.2141.51751.0001.0701.0771.1741.2671.349

Geometric Mean Of Renderers Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Renderers TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2d-highcpu-8 AMDc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR0.450.91.351.82.251.0001.0001.5001.5002.0002.000

Geometric Mean Of Scientific Computing Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Scientific Computing TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.42620.85241.27861.70482.1311.0001.0151.0891.3691.4211.894

Geometric Mean Of Server Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Server TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.42190.84381.26571.68762.10951.0001.0181.1271.3651.8251.875

Geometric Mean Of Server CPU Tests

Result Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milan

OpenBenchmarking.orgRelative Performance, More Is BetterGeometric Mean Of Server CPU TestsResult Composite - Google Cloud c3 Sapphire Rapids vs. AMD Milann2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.38230.76461.14691.52921.91151.0001.0011.1161.2861.5361.699

oneDNN

Performance / Cost - Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms x Dollar, Fewer Is BetteroneDNN 3.0Performance / Cost - Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPUn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR0.00880.01760.02640.03520.0440.0390.0340.0041. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 6 seconds per run: cost approximately 0.001 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 6 seconds per run: cost approximately 0.001 dollars. 3. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 6 seconds per run: cost approximately 0.001 dollars.

oneDNN

Performance / Cost - Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms x Dollar, Fewer Is BetteroneDNN 3.0Performance / Cost - Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPUn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR0.0020.0040.0060.0080.010.0090.0080.0070.0011. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 12 seconds per run: cost approximately 0.001 dollars. 2. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 12 seconds per run: cost approximately 0.001 dollars. 3. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 12 seconds per run: cost approximately 0.001 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 12 seconds per run: cost approximately 0.001 dollars.

OpenSSL

Performance / Cost - Algorithm: SHA256

OpenBenchmarking.orgbyte/s Per Dollar, More Is BetterOpenSSL 3.1Performance / Cost - Algorithm: SHA256n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD90000M180000M270000M360000M450000M62027045947.3762771120476.1984095258071.43251992587470.59388278658866.67404478101764.711. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.019 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.021 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.014 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.017 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.015 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.017 dollars.

oneDNN

Performance / Cost - Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms x Dollar, Fewer Is BetteroneDNN 3.0Performance / Cost - Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPUn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR0.02720.05440.08160.10880.1360.1210.1210.1060.0031. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 21 seconds per run: cost approximately 0.002 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 21 seconds per run: cost approximately 0.002 dollars. 3. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 21 seconds per run: cost approximately 0.002 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 21 seconds per run: cost approximately 0.002 dollars.

OpenSSL

Performance / Cost - Algorithm: AES-256-GCM

OpenBenchmarking.orgbyte/s Per Dollar, More Is BetterOpenSSL 3.1Performance / Cost - Algorithm: AES-256-GCMn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPR600000M1200000M1800000M2400000M3000000M796275684368.42806904791904.761080989637357.141131262576866.671849851336470.592824021269000.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.019 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.021 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.014 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.015 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.017 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.017 dollars.

OpenSSL

Performance / Cost - Algorithm: AES-128-GCM

OpenBenchmarking.orgbyte/s Per Dollar, More Is BetterOpenSSL 3.1Performance / Cost - Algorithm: AES-128-GCMn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDn2-highcpu-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR700000M1400000M2100000M2800000M3500000M1092908234368.421106538695380.951234828201533.331482396441214.291996959727823.533387886930764.711. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.019 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.021 dollars. 3. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.015 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.014 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.017 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.017 dollars.

OpenRadioss

Performance / Cost - Model: Rubber O-Ring Seal Installation

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterOpenRadioss 2022.10.13Performance / Cost - Model: Rubber O-Ring Seal Installationn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD91827364538.46131.93027.45212.84511.2944.1761. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 9 minutes, 58 seconds per run: cost approximately 0.065 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 8 minutes, 49 seconds per run: cost approximately 0.061 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 9 minutes, 51 seconds per run: cost approximately 0.047 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 6 minutes, 10 seconds per run: cost approximately 0.035 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 6 minutes, 7 seconds per run: cost approximately 0.031 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 212 seconds per run: cost approximately 0.02 dollars.

OpenRadioss

Performance / Cost - Model: Bumper Beam

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterOpenRadioss 2022.10.13Performance / Cost - Model: Bumper Beamn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD51015202520.50517.57715.1858.7986.8643.0061. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 7 minutes, 18 seconds per run: cost approximately 0.047 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 6 minutes, 32 seconds per run: cost approximately 0.045 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 7 minutes, 16 seconds per run: cost approximately 0.035 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 5 minutes, 5 seconds per run: cost approximately 0.029 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 4 minutes, 47 seconds per run: cost approximately 0.024 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 178 seconds per run: cost approximately 0.017 dollars.

OpenSSL

Performance / Cost - Algorithm: SHA512

OpenBenchmarking.orgbyte/s Per Dollar, More Is BetterOpenSSL 3.1Performance / Cost - Algorithm: SHA512n2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRn2-highcpu-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMD40000M80000M120000M160000M200000M68850211947.3769800725095.2492268995294.1293301694285.71124223471133.33185803224529.411. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.019 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.021 dollars. 3. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.017 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.014 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.015 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.017 dollars.

OSPRay Studio

Performance / Cost - Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer

OpenBenchmarking.orgms x Dollar, Fewer Is BetterOSPRay Studio 0.11Performance / Cost - Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracerc2-standard-8 CLXn2-standard-8 CLXt2d-standard-8 AMDc2d-highcpu-8 AMDn2-highcpu-8 CLXc3-highcpu-8 SPR130260390520650606.27490.70393.89372.27366.90309.831. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 142 seconds per run: cost approximately 0.016 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 76 seconds per run: cost approximately 0.008 dollars. 3. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 137 seconds per run: cost approximately 0.013 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 126 seconds per run: cost approximately 0.01 dollars. 5. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 75 seconds per run: cost approximately 0.006 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 121 seconds per run: cost approximately 0.012 dollars.

OpenRadioss

Performance / Cost - Model: Cell Phone Drop Test

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterOpenRadioss 2022.10.13Performance / Cost - Model: Cell Phone Drop Testn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD369121511.60010.2028.3294.8343.8501.7681. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 5 minutes, 29 seconds per run: cost approximately 0.036 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 4 minutes, 58 seconds per run: cost approximately 0.035 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 5 minutes, 28 seconds per run: cost approximately 0.026 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 225 seconds per run: cost approximately 0.022 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 219 seconds per run: cost approximately 0.018 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 142 seconds per run: cost approximately 0.013 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec Per Dollar, More Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Streamc2-standard-8 CLXn2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD30060090012001500418.69443.12525.36757.46854.851232.221. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 57 seconds per run: cost approximately 0.007 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.006 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 59 seconds per run: cost approximately 0.005 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 54 seconds per run: cost approximately 0.005 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 52 seconds per run: cost approximately 0.004 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 55 seconds per run: cost approximately 0.005 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec Per Dollar, More Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Streamn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD30060090012001500443.28489.20531.40753.86847.731239.121. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 58 seconds per run: cost approximately 0.006 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 55 seconds per run: cost approximately 0.006 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 57 seconds per run: cost approximately 0.005 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 53 seconds per run: cost approximately 0.005 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 50 seconds per run: cost approximately 0.004 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 53 seconds per run: cost approximately 0.005 dollars.

miniBUDE

Performance / Cost - Implementation: OpenMP - Input Deck: BM1

OpenBenchmarking.orgGFInst/s Per Dollar, More Is BetterminiBUDE 20210901Performance / Cost - Implementation: OpenMP - Input Deck: BM1n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD11K22K33K44K55K8371.9410186.5311162.5018865.1020675.5650675.501. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 149 seconds per run: cost approximately 0.016 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 131 seconds per run: cost approximately 0.015 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 149 seconds per run: cost approximately 0.012 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 106 seconds per run: cost approximately 0.01 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 108 seconds per run: cost approximately 0.009 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.006 dollars.

miniBUDE

Performance / Cost - Implementation: OpenMP - Input Deck: BM1

OpenBenchmarking.orgBillion Interactions/s Per Dollar, More Is BetterminiBUDE 20210901Performance / Cost - Implementation: OpenMP - Input Deck: BM1n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD400800120016002000334.88407.47446.50754.60827.002027.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 149 seconds per run: cost approximately 0.016 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 131 seconds per run: cost approximately 0.015 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 149 seconds per run: cost approximately 0.012 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 106 seconds per run: cost approximately 0.01 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 108 seconds per run: cost approximately 0.009 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.006 dollars.

BRL-CAD

Performance / Cost - VGR Performance Metric

OpenBenchmarking.orgVGR Performance Metric Per Dollar, More Is BetterBRL-CAD 7.34Performance / Cost - VGR Performance Metricn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD200K400K600K800K1000K512218.39541010.75655235.29923794.521076848.491166682.351. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 13 minutes, 24 seconds per run: cost approximately 0.087 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 13 minutes, 19 seconds per run: cost approximately 0.093 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 14 minutes, 15 seconds per run: cost approximately 0.068 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 14 minutes, 41 seconds per run: cost approximately 0.073 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 11 minutes, 32 seconds per run: cost approximately 0.066 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 15 minutes, 5 seconds per run: cost approximately 0.085 dollars.

PostgreSQL

Performance / Cost - Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latency

OpenBenchmarking.orgms x Dollar, Fewer Is BetterPostgreSQL 15Performance / Cost - Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latencyn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.02320.04640.06960.09280.1160.1030.1000.0820.0490.0480.0411. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 145 seconds per run: cost approximately 0.016 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 147 seconds per run: cost approximately 0.017 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 146 seconds per run: cost approximately 0.012 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 144 seconds per run: cost approximately 0.012 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 145 seconds per run: cost approximately 0.014 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 143 seconds per run: cost approximately 0.013 dollars.

PostgreSQL

Performance / Cost - Scaling Factor: 100 - Clients: 1000 - Mode: Read Only

OpenBenchmarking.orgTPS Per Dollar, More Is BetterPostgreSQL 15Performance / Cost - Scaling Factor: 100 - Clients: 1000 - Mode: Read Onlyn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD5M10M15M20M25M9699750.009979000.0012149166.6720349833.3320980357.1424679000.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 145 seconds per run: cost approximately 0.016 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 147 seconds per run: cost approximately 0.017 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 146 seconds per run: cost approximately 0.012 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 144 seconds per run: cost approximately 0.012 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 145 seconds per run: cost approximately 0.014 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 143 seconds per run: cost approximately 0.013 dollars.

uvg266

Performance / Cost - Video Input: Bosphorus 4K - Video Preset: Super Fast

OpenBenchmarking.orgFrames Per Second Per Dollar, More Is Betteruvg266 0.4.1Performance / Cost - Video Input: Bosphorus 4K - Video Preset: Super Fastn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD5001000150020002500440.00502.50574.44935.001356.672272.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 114 seconds per run: cost approximately 0.012 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 100 seconds per run: cost approximately 0.012 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 116 seconds per run: cost approximately 0.009 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 80 seconds per run: cost approximately 0.008 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 74 seconds per run: cost approximately 0.006 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 53 seconds per run: cost approximately 0.005 dollars.

uvg266

Performance / Cost - Video Input: Bosphorus 4K - Video Preset: Very Fast

OpenBenchmarking.orgFrames Per Second Per Dollar, More Is Betteruvg266 0.4.1Performance / Cost - Video Input: Bosphorus 4K - Video Preset: Very Fastn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD5001000150020002500382.31474.17493.00873.751316.672136.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 121 seconds per run: cost approximately 0.013 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 106 seconds per run: cost approximately 0.012 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 122 seconds per run: cost approximately 0.01 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 86 seconds per run: cost approximately 0.008 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 76 seconds per run: cost approximately 0.006 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 56 seconds per run: cost approximately 0.005 dollars.

uvg266

Performance / Cost - Video Input: Bosphorus 1080p - Video Preset: Super Fast

OpenBenchmarking.orgFrames Per Second Per Dollar, More Is Betteruvg266 0.4.1Performance / Cost - Video Input: Bosphorus 1080p - Video Preset: Super Fastn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD11K22K33K44K55K8050.009240.0012075.0017250.0037790.0052250.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 25 seconds per run: cost approximately 0.003 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 22 seconds per run: cost approximately 0.003 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 25 seconds per run: cost approximately 0.002 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 17 seconds per run: cost approximately 0.002 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 16 seconds per run: cost approximately 0.001 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 12 seconds per run: cost approximately 0.001 dollars.

uvg266

Performance / Cost - Video Input: Bosphorus 1080p - Video Preset: Very Fast

OpenBenchmarking.orgFrames Per Second Per Dollar, More Is Betteruvg266 0.4.1Performance / Cost - Video Input: Bosphorus 1080p - Video Preset: Very Fastn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD11K22K33K44K55K7603.338766.6711435.0016195.0036450.0049190.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 26 seconds per run: cost approximately 0.003 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 23 seconds per run: cost approximately 0.003 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 26 seconds per run: cost approximately 0.002 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 19 seconds per run: cost approximately 0.002 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 16 seconds per run: cost approximately 0.001 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 12 seconds per run: cost approximately 0.001 dollars.

uvg266

Performance / Cost - Video Input: Bosphorus 4K - Video Preset: Ultra Fast

OpenBenchmarking.orgFrames Per Second Per Dollar, More Is Betteruvg266 0.4.1Performance / Cost - Video Input: Bosphorus 4K - Video Preset: Ultra Fastn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD7001400210028003500651.00827.78915.711520.001964.003417.501. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 92 seconds per run: cost approximately 0.01 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 81 seconds per run: cost approximately 0.009 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 94 seconds per run: cost approximately 0.007 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.006 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 61 seconds per run: cost approximately 0.005 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 44 seconds per run: cost approximately 0.004 dollars.

PostgreSQL

Performance / Cost - Scaling Factor: 100 - Clients: 800 - Mode: Read Only - Average Latency

OpenBenchmarking.orgms x Dollar, Fewer Is BetterPostgreSQL 15Performance / Cost - Scaling Factor: 100 - Clients: 800 - Mode: Read Only - Average Latencyn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.01820.03640.05460.07280.0910.0810.0780.0630.0380.0360.0321. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 146 seconds per run: cost approximately 0.016 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 147 seconds per run: cost approximately 0.017 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 145 seconds per run: cost approximately 0.012 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 145 seconds per run: cost approximately 0.012 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 146 seconds per run: cost approximately 0.014 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 143 seconds per run: cost approximately 0.013 dollars.

PostgreSQL

Performance / Cost - Scaling Factor: 100 - Clients: 800 - Mode: Read Only

OpenBenchmarking.orgTPS Per Dollar, More Is BetterPostgreSQL 15Performance / Cost - Scaling Factor: 100 - Clients: 800 - Mode: Read Onlyn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD5M10M15M20M25M9891687.5010195117.6512674416.6720790833.3322281571.4324873307.691. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 146 seconds per run: cost approximately 0.016 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 147 seconds per run: cost approximately 0.017 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 145 seconds per run: cost approximately 0.012 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 145 seconds per run: cost approximately 0.012 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 146 seconds per run: cost approximately 0.014 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 143 seconds per run: cost approximately 0.013 dollars.

SPECFEM3D

Performance / Cost - Model: Homogeneous Halfspace

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterSPECFEM3D 4.0Performance / Cost - Model: Homogeneous Halfspacen2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD1.15492.30983.46474.61965.77455.1334.3883.8753.2322.2961.0231. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 221 seconds per run: cost approximately 0.024 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 198 seconds per run: cost approximately 0.023 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 223 seconds per run: cost approximately 0.018 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 185 seconds per run: cost approximately 0.018 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 169 seconds per run: cost approximately 0.014 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 106 seconds per run: cost approximately 0.01 dollars.

OpenRadioss

Performance / Cost - Model: Bird Strike on Windshield

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterOpenRadioss 2022.10.13Performance / Cost - Model: Bird Strike on Windshieldn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD163248648071.4460.8851.2033.9232.3413.911. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 13 minutes, 34 seconds per run: cost approximately 0.088 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 12 minutes, 7 seconds per run: cost approximately 0.084 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 13 minutes, 22 seconds per run: cost approximately 0.064 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 9 minutes, 57 seconds per run: cost approximately 0.057 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 10 minutes, 24 seconds per run: cost approximately 0.052 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 6 minutes, 28 seconds per run: cost approximately 0.036 dollars.

GROMACS

Performance / Cost - Implementation: MPI CPU - Input: water_GMX50_bare

OpenBenchmarking.orgNs Per Day Per Dollar, More Is BetterGROMACS 2023Performance / Cost - Implementation: MPI CPU - Input: water_GMX50_baren2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD163248648014.6716.5419.5932.8635.3273.201. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 5 minutes, 35 seconds per run: cost approximately 0.036 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 5 minutes, 5 seconds per run: cost approximately 0.035 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 5 minutes, 34 seconds per run: cost approximately 0.027 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 4 minutes, 16 seconds per run: cost approximately 0.021 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 229 seconds per run: cost approximately 0.022 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 163 seconds per run: cost approximately 0.015 dollars.

OpenCV

Performance / Cost - Test: Object Detection

OpenBenchmarking.orgms x Dollar, Fewer Is BetterOpenCV 4.7Performance / Cost - Test: Object Detectionn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD100200300400500449.99406.39331.35156.00113.0096.041. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 64 seconds per run: cost approximately 0.007 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 58 seconds per run: cost approximately 0.007 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.005 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 39 seconds per run: cost approximately 0.004 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 38 seconds per run: cost approximately 0.003 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 32 seconds per run: cost approximately 0.003 dollars.

uvg266

Performance / Cost - Video Input: Bosphorus 1080p - Video Preset: Ultra Fast

OpenBenchmarking.orgFrames Per Second Per Dollar, More Is Betteruvg266 0.4.1Performance / Cost - Video Input: Bosphorus 1080p - Video Preset: Ultra Fastn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD13K26K39K52K65K14985.0015025.0017235.0042240.0045530.0061850.001. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 20 seconds per run: cost approximately 0.002 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 20 seconds per run: cost approximately 0.002 dollars. 3. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 17 seconds per run: cost approximately 0.002 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 14 seconds per run: cost approximately 0.001 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 13 seconds per run: cost approximately 0.001 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 10 seconds per run: cost approximately 0.001 dollars.

SPECFEM3D

Performance / Cost - Model: Tomographic Model

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterSPECFEM3D 4.0Performance / Cost - Model: Tomographic Modeln2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD0.67391.34782.02172.69563.36952.9952.7162.3432.0151.4100.6541. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 169 seconds per run: cost approximately 0.018 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 154 seconds per run: cost approximately 0.018 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 170 seconds per run: cost approximately 0.014 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 147 seconds per run: cost approximately 0.014 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 131 seconds per run: cost approximately 0.011 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 84 seconds per run: cost approximately 0.008 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec Per Dollar, More Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Streamn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD110022003300440055002124.052332.202576.183243.883588.735211.841. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.006 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 55 seconds per run: cost approximately 0.006 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 57 seconds per run: cost approximately 0.005 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 53 seconds per run: cost approximately 0.005 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 49 seconds per run: cost approximately 0.004 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 52 seconds per run: cost approximately 0.005 dollars.

Memcached

Performance / Cost - Set To Get Ratio: 1:100

OpenBenchmarking.orgOps/sec Per Dollar, More Is BetterMemcached 1.6.18Performance / Cost - Set To Get Ratio: 1:100c2-standard-8 CLXn2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD40M80M120M160M200M87786491.2589282951.43120872674.00167081012.00171822880.00204091780.001. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.008 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.007 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.005 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.005 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.006 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.006 dollars.

OpenSSL

Performance / Cost - Algorithm: RSA4096

OpenBenchmarking.orgsign/s Per Dollar, More Is BetterOpenSSL 3.1Performance / Cost - Algorithm: RSA4096c2-standard-8 CLXn2-standard-8 CLXc2d-highcpu-8 AMDn2-highcpu-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR70K140K210K280K350K165228.57171433.33205140.00205760.00294166.67343783.331. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.007 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.006 dollars. 3. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.005 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.005 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.006 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.006 dollars.

SPECFEM3D

Performance / Cost - Model: Mount St. Helens

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterSPECFEM3D 4.0Performance / Cost - Model: Mount St. Helensn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD0.65791.31581.97372.63163.28952.9242.4712.0841.9531.3820.6491. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 165 seconds per run: cost approximately 0.018 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 147 seconds per run: cost approximately 0.017 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 163 seconds per run: cost approximately 0.013 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 142 seconds per run: cost approximately 0.014 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 127 seconds per run: cost approximately 0.011 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 83 seconds per run: cost approximately 0.008 dollars.

oneDNN

Performance / Cost - Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms x Dollar, Fewer Is BetteroneDNN 3.0Performance / Cost - Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPUn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPR0.00990.01980.02970.03960.04950.0440.0390.0220.0021. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 15 seconds per run: cost approximately 0.002 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 15 seconds per run: cost approximately 0.002 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 15 seconds per run: cost approximately 0.001 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 15 seconds per run: cost approximately 0.001 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec Per Dollar, More Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Streamn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRn2-highcpu-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMD3K6K9K12K15K5194.725835.668276.608564.579496.5012838.531. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 42 seconds per run: cost approximately 0.005 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 45 seconds per run: cost approximately 0.005 dollars. 3. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 40 seconds per run: cost approximately 0.004 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 42 seconds per run: cost approximately 0.003 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 40 seconds per run: cost approximately 0.003 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 42 seconds per run: cost approximately 0.004 dollars.

Zstd Compression

Performance / Cost - Compression Level: 19 - Decompression Speed

OpenBenchmarking.orgMB/s Per Dollar, More Is BetterZstd Compression 1.5.4Performance / Cost - Compression Level: 19 - Decompression Speedn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD40K80K120K160K200K72166.6777900.0091542.86129314.29161285.71208933.331. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 84 seconds per run: cost approximately 0.009 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 80 seconds per run: cost approximately 0.009 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 84 seconds per run: cost approximately 0.007 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 74 seconds per run: cost approximately 0.007 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 72 seconds per run: cost approximately 0.007 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 71 seconds per run: cost approximately 0.006 dollars.

Memcached

Performance / Cost - Set To Get Ratio: 1:10

OpenBenchmarking.orgOps/sec Per Dollar, More Is BetterMemcached 1.6.18Performance / Cost - Set To Get Ratio: 1:10c2-standard-8 CLXn2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD40M80M120M160M200M89465441.2591673371.43122948524.00168566154.00174157855.00200087903.331. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.008 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.007 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.005 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.005 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.006 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 66 seconds per run: cost approximately 0.006 dollars.

OpenFOAM

Performance / Cost - Input: drivaerFastback, Small Mesh Size - Execution Time

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterOpenFOAM 10Performance / Cost - Input: drivaerFastback, Small Mesh Size - Execution Timec2-standard-8 CLXn2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD102030405042.04641.91929.88519.86615.7509.7131. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 10 minutes, 50 seconds per run: cost approximately 0.075 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 11 minutes, 13 seconds per run: cost approximately 0.073 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 11 minutes, 1 second per run: cost approximately 0.053 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 8 minutes, 9 seconds per run: cost approximately 0.047 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 7 minutes, 52 seconds per run: cost approximately 0.039 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 5 minutes, 49 seconds per run: cost approximately 0.033 dollars.

OpenFOAM

Performance / Cost - Input: drivaerFastback, Small Mesh Size - Mesh Time

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterOpenFOAM 10Performance / Cost - Input: drivaerFastback, Small Mesh Size - Mesh Timen2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD2468106.9406.4144.8502.9162.4931.6091. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 11 minutes, 13 seconds per run: cost approximately 0.073 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 10 minutes, 50 seconds per run: cost approximately 0.075 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 11 minutes, 1 second per run: cost approximately 0.053 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 8 minutes, 9 seconds per run: cost approximately 0.047 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 7 minutes, 52 seconds per run: cost approximately 0.039 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 5 minutes, 49 seconds per run: cost approximately 0.033 dollars.

SPECFEM3D

Performance / Cost - Model: Water-layered Halfspace

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterSPECFEM3D 4.0Performance / Cost - Model: Water-layered Halfspacen2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD4812162016.34614.60811.60010.2927.1863.7531. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 6 minutes, 36 seconds per run: cost approximately 0.043 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 6 minutes, 2 seconds per run: cost approximately 0.042 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 6 minutes, 30 seconds per run: cost approximately 0.031 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 5 minutes, 32 seconds per run: cost approximately 0.032 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 4 minutes, 58 seconds per run: cost approximately 0.025 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 205 seconds per run: cost approximately 0.019 dollars.

Timed FFmpeg Compilation

Performance / Cost - Time To Compile

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterTimed FFmpeg Compilation 6.0Performance / Cost - Time To Compilen2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD0.59241.18481.77722.36962.9622.6332.2251.8591.3251.1960.6451. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 155 seconds per run: cost approximately 0.017 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 139 seconds per run: cost approximately 0.016 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 155 seconds per run: cost approximately 0.012 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 120 seconds per run: cost approximately 0.011 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 120 seconds per run: cost approximately 0.01 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 81 seconds per run: cost approximately 0.008 dollars.

Xcompact3d Incompact3d

Performance / Cost - Input: input.i3d 129 Cells Per Direction

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Performance / Cost - Input: input.i3d 129 Cells Per Directionn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD0.03650.0730.10950.1460.18250.1620.1510.1200.0970.0530.0431. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 41 seconds per run: cost approximately 0.004 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 38 seconds per run: cost approximately 0.004 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 41 seconds per run: cost approximately 0.003 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 33 seconds per run: cost approximately 0.003 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 27 seconds per run: cost approximately 0.002 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 22 seconds per run: cost approximately 0.002 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch x Dollar, Fewer Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Streamt2d-standard-8 AMDc2d-highcpu-8 AMDn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRn2-highcpu-8 CLX0.05220.10440.15660.20880.2610.2320.1570.1450.1330.1230.1101. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 38 seconds per run: cost approximately 0.004 dollars. 2. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 36 seconds per run: cost approximately 0.003 dollars. 3. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 39 seconds per run: cost approximately 0.004 dollars. 4. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 38 seconds per run: cost approximately 0.004 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 37 seconds per run: cost approximately 0.004 dollars. 6. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 38 seconds per run: cost approximately 0.003 dollars.

Intel Open Image Denoise

Performance / Cost - Run: RTLightmap.hdr.4096x4096

OpenBenchmarking.orgImages / Sec Per Dollar, More Is BetterIntel Open Image Denoise 1.4.0Performance / Cost - Run: RTLightmap.hdr.4096x4096c2d-highcpu-8 AMDn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD2468102.6672.9413.4383.6004.8007.8951. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 5 minutes, 59 seconds per run: cost approximately 0.03 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 5 minutes, 12 seconds per run: cost approximately 0.034 dollars. 3. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 4 minutes, 40 seconds per run: cost approximately 0.032 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 5 minutes, 17 seconds per run: cost approximately 0.025 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 4 minutes, 17 seconds per run: cost approximately 0.025 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 200 seconds per run: cost approximately 0.019 dollars.

Intel Open Image Denoise

Performance / Cost - Run: RT.hdr_alb_nrm.3840x2160

OpenBenchmarking.orgImages / Sec Per Dollar, More Is BetterIntel Open Image Denoise 1.4.0Performance / Cost - Run: RT.hdr_alb_nrm.3840x2160c2d-highcpu-8 AMDn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD81624324010.6712.5014.6716.6720.0033.331. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.015 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 148 seconds per run: cost approximately 0.016 dollars. 3. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 133 seconds per run: cost approximately 0.015 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 148 seconds per run: cost approximately 0.012 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 123 seconds per run: cost approximately 0.012 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 101 seconds per run: cost approximately 0.009 dollars.

SPECFEM3D

Performance / Cost - Model: Layered Halfspace

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterSPECFEM3D 4.0Performance / Cost - Model: Layered Halfspacen2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMDt2d-standard-8 AMD51015202518.68316.48913.74213.3958.7714.6851. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 6 minutes, 59 seconds per run: cost approximately 0.045 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 6 minutes, 18 seconds per run: cost approximately 0.044 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 7 minutes per run: cost approximately 0.033 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 6 minutes, 16 seconds per run: cost approximately 0.036 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 5 minutes, 28 seconds per run: cost approximately 0.027 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 226 seconds per run: cost approximately 0.021 dollars.

TensorFlow

Performance / Cost - Device: CPU - Batch Size: 64 - Model: ResNet-50

OpenBenchmarking.orgimages/sec Per Dollar, More Is BetterTensorFlow 2.10Performance / Cost - Device: CPU - Batch Size: 64 - Model: ResNet-50c2d-highcpu-8 AMDt2d-standard-8 AMDn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPR80160240320400120.86163.75245.36264.11326.19364.881. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 13 minutes, 56 seconds per run: cost approximately 0.07 dollars. 2. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 11 minutes, 18 seconds per run: cost approximately 0.064 dollars. 3. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 8 minutes, 41 seconds per run: cost approximately 0.056 dollars. 4. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 8 minutes, 3 seconds per run: cost approximately 0.056 dollars. 5. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 8 minutes, 42 seconds per run: cost approximately 0.042 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 7 minutes, 34 seconds per run: cost approximately 0.043 dollars.

OpenSSL

Performance / Cost - Algorithm: ChaCha20-Poly1305

OpenBenchmarking.orgbyte/s Per Dollar, More Is BetterOpenSSL 3.1Performance / Cost - Algorithm: ChaCha20-Poly1305n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD200000M400000M600000M800000M1000000M504027018578.95519959611571.43683391021428.57783955097133.33939457714117.651034858828058.821. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.019 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.021 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.014 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.015 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.017 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.017 dollars.

Zstd Compression

Performance / Cost - Compression Level: 19, Long Mode - Decompression Speed

OpenBenchmarking.orgMB/s Per Dollar, More Is BetterZstd Compression 1.5.4Performance / Cost - Compression Level: 19, Long Mode - Decompression Speedn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD50K100K150K200K250K82537.5089225.00108333.33151200.00153685.71236760.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 74 seconds per run: cost approximately 0.008 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 68 seconds per run: cost approximately 0.008 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 74 seconds per run: cost approximately 0.006 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 65 seconds per run: cost approximately 0.006 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 71 seconds per run: cost approximately 0.007 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 63 seconds per run: cost approximately 0.005 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec Per Dollar, More Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Streamc2d-highcpu-8 AMDn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDn2-highcpu-8 CLX4K8K12K16K20K12708.0713739.7514997.8516218.6817213.7818171.971. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 36 seconds per run: cost approximately 0.003 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 39 seconds per run: cost approximately 0.004 dollars. 3. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 38 seconds per run: cost approximately 0.004 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 37 seconds per run: cost approximately 0.004 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 38 seconds per run: cost approximately 0.004 dollars. 6. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 38 seconds per run: cost approximately 0.003 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec Per Dollar, More Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Streamn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD400800120016002000821.30884.191129.621146.051307.441633.081. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 62 seconds per run: cost approximately 0.007 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.007 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 62 seconds per run: cost approximately 0.005 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 50 seconds per run: cost approximately 0.004 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 56 seconds per run: cost approximately 0.005 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 52 seconds per run: cost approximately 0.005 dollars.

John The Ripper

Performance / Cost - Test: Blowfish

OpenBenchmarking.orgReal C/S Per Dollar, More Is BetterJohn The Ripper 2023.03.14Performance / Cost - Test: Blowfishn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRn2-highcpu-8 CLXt2d-standard-8 AMDc2d-highcpu-8 AMD800K1600K2400K3200K4000K1989333.332228000.002310000.002979500.003537000.003809500.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 30 seconds per run: cost approximately 0.003 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 30 seconds per run: cost approximately 0.003 dollars. 3. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 30 seconds per run: cost approximately 0.003 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 30 seconds per run: cost approximately 0.002 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 30 seconds per run: cost approximately 0.003 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 30 seconds per run: cost approximately 0.002 dollars.

John The Ripper

Performance / Cost - Test: bcrypt

OpenBenchmarking.orgReal C/S Per Dollar, More Is BetterJohn The Ripper 2023.03.14Performance / Cost - Test: bcryptn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRn2-highcpu-8 CLXt2d-standard-8 AMDc2d-highcpu-8 AMD800K1600K2400K3200K4000K1990000.002229000.002310666.672981000.003528333.333810000.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 30 seconds per run: cost approximately 0.003 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 30 seconds per run: cost approximately 0.003 dollars. 3. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 30 seconds per run: cost approximately 0.003 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 30 seconds per run: cost approximately 0.002 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 30 seconds per run: cost approximately 0.003 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 30 seconds per run: cost approximately 0.002 dollars.

nekRS

Performance / Cost - Input: TurboPipe Periodic

OpenBenchmarking.orgFLOP/s Per Dollar, More Is BetternekRS 22.0Performance / Cost - Input: TurboPipe Periodicn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD150000M300000M450000M600000M750000M194606497177.97216385919543.10352145454545.46393178205128.21702695977017.241. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 18 minutes, 11 seconds per run: cost approximately 0.118 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 16 minutes, 38 seconds per run: cost approximately 0.116 dollars. 3. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 15 minutes, 24 seconds per run: cost approximately 0.077 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 13 minutes, 37 seconds per run: cost approximately 0.078 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 10 minutes, 21 seconds per run: cost approximately 0.058 dollars.

Timed Linux Kernel Compilation

Performance / Cost - Build: defconfig

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterTimed Linux Kernel Compilation 6.1Performance / Cost - Build: defconfign2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD369121511.2839.8498.5225.7055.6303.1401. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 5 minutes, 22 seconds per run: cost approximately 0.035 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 4 minutes, 50 seconds per run: cost approximately 0.034 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 5 minutes, 28 seconds per run: cost approximately 0.026 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 4 minutes, 19 seconds per run: cost approximately 0.022 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 4 minutes, 5 seconds per run: cost approximately 0.023 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 185 seconds per run: cost approximately 0.017 dollars.

nginx

Performance / Cost - Connections: 500

OpenBenchmarking.orgRequests Per Second Per Dollar, More Is Betternginx 1.23.2Performance / Cost - Connections: 500n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD900K1800K2700K3600K4500K2014610.002195717.002829514.293466472.863852516.674390907.501. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.01 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.01 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.007 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.007 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.009 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.008 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgitems/sec Per Dollar, More Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Streamc2-standard-8 CLXn2-standard-8 CLXc2d-highcpu-8 AMDn2-highcpu-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR100020003000400050002652.983072.423386.653778.384761.664791.931. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 48 seconds per run: cost approximately 0.006 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 49 seconds per run: cost approximately 0.005 dollars. 3. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 53 seconds per run: cost approximately 0.004 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 49 seconds per run: cost approximately 0.004 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 57 seconds per run: cost approximately 0.005 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 44 seconds per run: cost approximately 0.004 dollars.

nginx

Performance / Cost - Connections: 1000

OpenBenchmarking.orgRequests Per Second Per Dollar, More Is Betternginx 1.23.2Performance / Cost - Connections: 1000n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD900K1800K2700K3600K4500K1968360.002144627.002774011.433394067.143568731.114259952.501. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.01 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.01 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.007 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.007 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.009 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.008 dollars.

John The Ripper

Performance / Cost - Test: MD5

OpenBenchmarking.orgReal C/S Per Dollar, More Is BetterJohn The Ripper 2023.03.14Performance / Cost - Test: MD5c2d-highcpu-8 AMDc2-standard-8 CLXn2-standard-8 CLXn2-highcpu-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR30M60M90M120M150M87317000.0097649428.5799939500.00119794200.00123035000.00127618833.331. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.005 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.007 dollars. 3. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.006 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.005 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.006 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.006 dollars.

TensorFlow

Performance / Cost - Device: CPU - Batch Size: 32 - Model: ResNet-50

OpenBenchmarking.orgimages/sec Per Dollar, More Is BetterTensorFlow 2.10Performance / Cost - Device: CPU - Batch Size: 32 - Model: ResNet-50c2d-highcpu-8 AMDt2d-standard-8 AMDn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPR140280420560700244.29331.56431.67470.33590.46649.131. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 6 minutes, 56 seconds per run: cost approximately 0.035 dollars. 2. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 5 minutes, 37 seconds per run: cost approximately 0.032 dollars. 3. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 4 minutes, 40 seconds per run: cost approximately 0.03 dollars. 4. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 4 minutes, 16 seconds per run: cost approximately 0.03 dollars. 5. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 4 minutes, 38 seconds per run: cost approximately 0.022 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 4 minutes, 1 second per run: cost approximately 0.023 dollars.

Google Draco

Performance / Cost - Model: Church Facade

OpenBenchmarking.orgms x Dollar, Fewer Is BetterGoogle Draco 1.5.6Performance / Cost - Model: Church Facaden2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPRc2d-highcpu-8 AMD61218243024.66422.92412.8828.7547.5737.4421. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 15 seconds per run: cost approximately 0.002 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 14 seconds per run: cost approximately 0.002 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 15 seconds per run: cost approximately 0.001 dollars. 4. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 11 seconds per run: cost approximately 0.001 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 9 seconds per run: cost approximately 0.001 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 9 seconds per run: cost approximately 0.001 dollars.

7-Zip Compression

Performance / Cost - Test: Compression Rating

OpenBenchmarking.orgMIPS Per Dollar, More Is Better7-Zip Compression 22.01Performance / Cost - Test: Compression Ratingn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRn2-highcpu-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMD5M10M15M20M25M7015500.007747250.008826500.009364333.3319601500.0024224500.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 38 seconds per run: cost approximately 0.004 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 35 seconds per run: cost approximately 0.004 dollars. 3. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 37 seconds per run: cost approximately 0.004 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 39 seconds per run: cost approximately 0.003 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 27 seconds per run: cost approximately 0.002 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 24 seconds per run: cost approximately 0.002 dollars.

nginx

Performance / Cost - Connections: 4000

OpenBenchmarking.orgRequests Per Second Per Dollar, More Is Betternginx 1.23.2Performance / Cost - Connections: 4000n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD900K1800K2700K3600K4500K1994756.002159494.002782472.863392502.863646083.334199096.251. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.01 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.01 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.007 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.007 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.009 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.008 dollars.

OSPRay Studio

Performance / Cost - Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer

OpenBenchmarking.orgms x Dollar, Fewer Is BetterOSPRay Studio 0.11Performance / Cost - Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracern2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPR120240360480600538.26432.75387.12282.33278.37209.521. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 140 seconds per run: cost approximately 0.015 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 121 seconds per run: cost approximately 0.014 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 137 seconds per run: cost approximately 0.011 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 109 seconds per run: cost approximately 0.009 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 118 seconds per run: cost approximately 0.011 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 102 seconds per run: cost approximately 0.01 dollars.

Embree

Performance / Cost - Binary: Pathtracer ISPC - Model: Crown

OpenBenchmarking.orgFrames Per Second Per Dollar, More Is BetterEmbree 4.0.1Performance / Cost - Binary: Pathtracer ISPC - Model: Crownn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPR110220330440550172.92207.05231.43447.56499.35531.591. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 187 seconds per run: cost approximately 0.02 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 164 seconds per run: cost approximately 0.019 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 186 seconds per run: cost approximately 0.015 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 131 seconds per run: cost approximately 0.011 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 118 seconds per run: cost approximately 0.011 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 110 seconds per run: cost approximately 0.011 dollars.

CockroachDB

Performance / Cost - Workload: KV, 95% Reads - Concurrency: 128

OpenBenchmarking.orgops/s Per Dollar, More Is BetterCockroachDB 22.2Performance / Cost - Workload: KV, 95% Reads - Concurrency: 128n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXt2d-standard-8 AMDc2d-highcpu-8 AMDc3-highcpu-8 SPR500K1000K1500K2000K2500K1248066.671348683.331653855.562214880.002285755.562496010.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 107 seconds per run: cost approximately 0.012 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 106 seconds per run: cost approximately 0.012 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 107 seconds per run: cost approximately 0.009 dollars. 4. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 106 seconds per run: cost approximately 0.01 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 106 seconds per run: cost approximately 0.009 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 106 seconds per run: cost approximately 0.01 dollars.

nginx

Performance / Cost - Connections: 200

OpenBenchmarking.orgRequests Per Second Per Dollar, More Is Betternginx 1.23.2Performance / Cost - Connections: 200n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD900K1800K2700K3600K4500K2257018.002469591.003061631.433525290.003955788.894403130.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.01 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.01 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.007 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.007 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.009 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.008 dollars.

LeelaChessZero

Performance / Cost - Backend: BLAS

OpenBenchmarking.orgNodes Per Second Per Dollar, More Is BetterLeelaChessZero 0.28Performance / Cost - Backend: BLASn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPR8K16K24K32K40K20769.2321642.8626551.7228533.3335352.9437411.771. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 6 minutes, 2 seconds per run: cost approximately 0.039 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 6 minutes, 1 second per run: cost approximately 0.042 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 6 minutes, 2 seconds per run: cost approximately 0.029 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 6 minutes, 1 second per run: cost approximately 0.03 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 6 minutes, 2 seconds per run: cost approximately 0.034 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 6 minutes per run: cost approximately 0.034 dollars.

LeelaChessZero

Performance / Cost - Backend: Eigen

OpenBenchmarking.orgNodes Per Second Per Dollar, More Is BetterLeelaChessZero 0.28Performance / Cost - Backend: Eigenn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDn2-highcpu-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPR8K16K24K32K40K20717.9521476.1924766.6727620.6930558.8235911.771. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 6 minutes, 3 seconds per run: cost approximately 0.039 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 6 minutes, 1 second per run: cost approximately 0.042 dollars. 3. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 6 minutes, 5 seconds per run: cost approximately 0.03 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 6 minutes, 2 seconds per run: cost approximately 0.029 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 6 minutes, 5 seconds per run: cost approximately 0.034 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 6 minutes per run: cost approximately 0.034 dollars.

TensorFlow

Performance / Cost - Device: CPU - Batch Size: 16 - Model: ResNet-50

OpenBenchmarking.orgimages/sec Per Dollar, More Is BetterTensorFlow 2.10Performance / Cost - Device: CPU - Batch Size: 16 - Model: ResNet-50c2d-highcpu-8 AMDt2d-standard-8 AMDn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPR30060090012001500509.41671.25761.25833.131015.001183.331. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 209 seconds per run: cost approximately 0.017 dollars. 2. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 169 seconds per run: cost approximately 0.016 dollars. 3. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 152 seconds per run: cost approximately 0.016 dollars. 4. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 139 seconds per run: cost approximately 0.016 dollars. 5. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 153 seconds per run: cost approximately 0.012 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 130 seconds per run: cost approximately 0.012 dollars.

NAMD

Performance / Cost - ATPase Simulation - 327,506 Atoms

OpenBenchmarking.orgdays/ns x Dollar, Fewer Is BetterNAMD 2.14Performance / Cost - ATPase Simulation - 327,506 Atomsn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD0.03080.06160.09240.12320.1540.1370.1160.1010.0800.0540.0441. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 4 minutes, 13 seconds per run: cost approximately 0.027 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 224 seconds per run: cost approximately 0.026 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 4 minutes, 12 seconds per run: cost approximately 0.02 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 215 seconds per run: cost approximately 0.018 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 168 seconds per run: cost approximately 0.016 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 154 seconds per run: cost approximately 0.014 dollars.

oneDNN

Performance / Cost - Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms x Dollar, Fewer Is BetteroneDNN 3.0Performance / Cost - Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPUn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDn2-highcpu-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD153045607566.8763.4552.0348.9041.9536.041. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 102 seconds per run: cost approximately 0.011 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 99 seconds per run: cost approximately 0.011 dollars. 3. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 102 seconds per run: cost approximately 0.008 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 102 seconds per run: cost approximately 0.008 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 94 seconds per run: cost approximately 0.009 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 92 seconds per run: cost approximately 0.009 dollars.

nginx

Performance / Cost - Connections: 100

OpenBenchmarking.orgRequests Per Second Per Dollar, More Is Betternginx 1.23.2Performance / Cost - Connections: 100n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRt2d-standard-8 AMD900K1800K2700K3600K4500K2283019.002514828.003206342.863574265.714034483.334407221.251. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.01 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.01 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.007 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.007 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.009 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 90 seconds per run: cost approximately 0.008 dollars.

oneDNN

Performance / Cost - Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms x Dollar, Fewer Is BetteroneDNN 3.0Performance / Cost - Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPUn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPR0.0020.0040.0060.0080.010.0090.0090.0080.0051. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 10 seconds per run: cost approximately 0.001 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 10 seconds per run: cost approximately 0.001 dollars. 3. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 10 seconds per run: cost approximately 0.001 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 9 seconds per run: cost approximately 0.001 dollars.

Embree

Performance / Cost - Binary: Pathtracer ISPC - Model: Asian Dragon

OpenBenchmarking.orgFrames Per Second Per Dollar, More Is BetterEmbree 4.0.1Performance / Cost - Binary: Pathtracer ISPC - Model: Asian Dragonn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPR2004006008001000305.12343.61418.13648.33713.03920.531. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 142 seconds per run: cost approximately 0.015 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 126 seconds per run: cost approximately 0.015 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 141 seconds per run: cost approximately 0.011 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 111 seconds per run: cost approximately 0.009 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 101 seconds per run: cost approximately 0.009 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 88 seconds per run: cost approximately 0.008 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch x Dollar, Fewer Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Streamt2d-standard-8 AMDc2-standard-8 CLXn2-standard-8 CLXc2d-highcpu-8 AMDn2-highcpu-8 CLXc3-highcpu-8 SPR0.18860.37720.56580.75440.9430.8380.7530.6500.5900.5290.4171. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 57 seconds per run: cost approximately 0.005 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 48 seconds per run: cost approximately 0.006 dollars. 3. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 49 seconds per run: cost approximately 0.005 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 53 seconds per run: cost approximately 0.004 dollars. 5. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 49 seconds per run: cost approximately 0.004 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 44 seconds per run: cost approximately 0.004 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch x Dollar, Fewer Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Streamt2d-standard-8 AMDn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPR0.55011.10021.65032.20042.75052.4452.4352.2621.7701.7451.5291. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 52 seconds per run: cost approximately 0.005 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 62 seconds per run: cost approximately 0.007 dollars. 3. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.007 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 62 seconds per run: cost approximately 0.005 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 50 seconds per run: cost approximately 0.004 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 56 seconds per run: cost approximately 0.005 dollars.

oneDNN

Performance / Cost - Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.orgms x Dollar, Fewer Is BetteroneDNN 3.0Performance / Cost - Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPUc2-standard-8 CLXn2-standard-8 CLXc2d-highcpu-8 AMDn2-highcpu-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMD71421283529.9828.5523.3522.1018.7016.861. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 86 seconds per run: cost approximately 0.01 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 87 seconds per run: cost approximately 0.009 dollars. 3. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 87 seconds per run: cost approximately 0.007 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 87 seconds per run: cost approximately 0.007 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 84 seconds per run: cost approximately 0.008 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 84 seconds per run: cost approximately 0.008 dollars.

OpenVKL

Performance / Cost - Benchmark: vklBenchmark ISPC

OpenBenchmarking.orgItems / Sec Per Dollar, More Is BetterOpenVKL 1.3.1Performance / Cost - Benchmark: vklBenchmark ISPCn2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPRn2-highcpu-8 CLX70014002100280035001369.571458.331823.531973.682450.003375.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 7 minutes, 5 seconds per run: cost approximately 0.046 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 6 minutes, 50 seconds per run: cost approximately 0.048 dollars. 3. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 6 minutes, 50 seconds per run: cost approximately 0.034 dollars. 4. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 6 minutes, 43 seconds per run: cost approximately 0.038 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 6 minutes, 53 seconds per run: cost approximately 0.04 dollars. 6. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 4 minutes, 55 seconds per run: cost approximately 0.024 dollars.

OpenCV

Performance / Cost - Test: Image Processing

OpenBenchmarking.orgms x Dollar, Fewer Is BetterOpenCV 4.7Performance / Cost - Test: Image Processingn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD70014002100280035003395.382502.982419.791537.961258.511176.181. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 179 seconds per run: cost approximately 0.019 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 147 seconds per run: cost approximately 0.017 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 173 seconds per run: cost approximately 0.014 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 128 seconds per run: cost approximately 0.012 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 115 seconds per run: cost approximately 0.011 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 118 seconds per run: cost approximately 0.01 dollars.

MariaDB

Performance / Cost - Clients: 4096

OpenBenchmarking.orgQueries Per Second Per Dollar, More Is BetterMariaDB 11.0.1Performance / Cost - Clients: 4096n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXt2d-standard-8 AMDc2d-highcpu-8 AMDc3-highcpu-8 SPR2K4K6K8K10K3833.334647.065487.188000.009733.339906.251. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 8 minutes, 20 seconds per run: cost approximately 0.054 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 7 minutes, 17 seconds per run: cost approximately 0.051 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 8 minutes, 4 seconds per run: cost approximately 0.039 dollars. 4. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 6 minutes, 13 seconds per run: cost approximately 0.035 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 5 minutes, 58 seconds per run: cost approximately 0.03 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 5 minutes, 30 seconds per run: cost approximately 0.032 dollars.

CockroachDB

Performance / Cost - Workload: KV, 50% Reads - Concurrency: 128

OpenBenchmarking.orgops/s Per Dollar, More Is BetterCockroachDB 22.2Performance / Cost - Workload: KV, 50% Reads - Concurrency: 128n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc2d-highcpu-8 AMDt2d-standard-8 AMDc3-highcpu-8 SPR400K800K1200K1600K2000K1058258.331140375.001403900.001907600.001911040.001932160.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 107 seconds per run: cost approximately 0.012 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 107 seconds per run: cost approximately 0.012 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 107 seconds per run: cost approximately 0.009 dollars. 4. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 106 seconds per run: cost approximately 0.009 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 106 seconds per run: cost approximately 0.01 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 106 seconds per run: cost approximately 0.01 dollars.

OpenSSL

Performance / Cost - Algorithm: ChaCha20

OpenBenchmarking.orgbyte/s Per Dollar, More Is BetterOpenSSL 3.1Performance / Cost - Algorithm: ChaCha20n2-standard-8 CLXc2-standard-8 CLXc2d-highcpu-8 AMDc3-highcpu-8 SPRn2-highcpu-8 CLXt2d-standard-8 AMD300000M600000M900000M1200000M1500000M1002901590368.421016514848238.101151101186000.001299503390411.761361820421642.861548546288823.531. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.019 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.021 dollars. 3. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.015 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.017 dollars. 5. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.014 dollars. 6. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 180 seconds per run: cost approximately 0.017 dollars.

MariaDB

Performance / Cost - Clients: 2048

OpenBenchmarking.orgQueries Per Second Per Dollar, More Is BetterMariaDB 11.0.1Performance / Cost - Clients: 2048n2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXt2d-standard-8 AMDc2d-highcpu-8 AMDc3-highcpu-8 SPR2K4K6K8K10K4313.735166.675945.959406.2510892.8611066.671. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 7 minutes, 48 seconds per run: cost approximately 0.051 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 6 minutes, 53 seconds per run: cost approximately 0.048 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 7 minutes, 48 seconds per run: cost approximately 0.037 dollars. 4. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 5 minutes, 46 seconds per run: cost approximately 0.032 dollars. 5. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 5 minutes, 39 seconds per run: cost approximately 0.028 dollars. 6. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 5 minutes, 11 seconds per run: cost approximately 0.03 dollars.

OpenCV

Performance / Cost - Test: Stitching

OpenBenchmarking.orgms x Dollar, Fewer Is BetterOpenCV 4.7Performance / Cost - Test: Stitchingn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD2K4K6K8K10K8354.797274.166189.444509.963524.812998.401. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 4 minutes, 39 seconds per run: cost approximately 0.03 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 4 minutes, 11 seconds per run: cost approximately 0.029 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 4 minutes, 41 seconds per run: cost approximately 0.022 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 215 seconds per run: cost approximately 0.021 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 196 seconds per run: cost approximately 0.018 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 187 seconds per run: cost approximately 0.016 dollars.

Google Draco

Performance / Cost - Model: Lion

OpenBenchmarking.orgms x Dollar, Fewer Is BetterGoogle Draco 1.5.6Performance / Cost - Model: Lionn2-highcpu-8 CLXn2-standard-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD2468108.3918.3557.4376.2506.0735.6171. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 11 seconds per run: cost approximately 0.001 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 11 seconds per run: cost approximately 0.001 dollars. 3. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 10 seconds per run: cost approximately 0.001 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 8 seconds per run: cost approximately 0.001 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 8 seconds per run: cost approximately 0.001 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 7 seconds per run: cost approximately 0.001 dollars.

OpenCV

Performance / Cost - Test: Graph API

OpenBenchmarking.orgms x Dollar, Fewer Is BetterOpenCV 4.7Performance / Cost - Test: Graph APIn2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD170034005100680085007750.256650.736377.024618.554204.283390.991. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 4 minutes, 28 seconds per run: cost approximately 0.029 dollars. 2. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 4 minutes, 50 seconds per run: cost approximately 0.023 dollars. 3. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 237 seconds per run: cost approximately 0.027 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 221 seconds per run: cost approximately 0.021 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 211 seconds per run: cost approximately 0.02 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 200 seconds per run: cost approximately 0.017 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch x Dollar, Fewer Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Streamc2-standard-8 CLXn2-standard-8 CLXn2-highcpu-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPRc2d-highcpu-8 AMD1.07482.14963.22444.29925.3744.7774.5133.8073.2272.6402.3381. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 57 seconds per run: cost approximately 0.007 dollars. 2. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.006 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 59 seconds per run: cost approximately 0.005 dollars. 4. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 55 seconds per run: cost approximately 0.005 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 54 seconds per run: cost approximately 0.005 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 52 seconds per run: cost approximately 0.004 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch x Dollar, Fewer Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Streamn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPRc2d-highcpu-8 AMD1.01522.03043.04564.06085.0764.5124.0883.7633.2152.6532.3541. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 58 seconds per run: cost approximately 0.006 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 55 seconds per run: cost approximately 0.006 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 57 seconds per run: cost approximately 0.005 dollars. 4. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 53 seconds per run: cost approximately 0.005 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 53 seconds per run: cost approximately 0.005 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 50 seconds per run: cost approximately 0.004 dollars.

Zstd Compression

Performance / Cost - Compression Level: 19 - Compression Speed

OpenBenchmarking.orgMB/s Per Dollar, More Is BetterZstd Compression 1.5.4Performance / Cost - Compression Level: 19 - Compression Speedn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXc3-highcpu-8 SPRt2d-standard-8 AMDc2d-highcpu-8 AMD400800120016002000915.561026.671184.291471.431557.141833.331. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 84 seconds per run: cost approximately 0.009 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 80 seconds per run: cost approximately 0.009 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 84 seconds per run: cost approximately 0.007 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 74 seconds per run: cost approximately 0.007 dollars. 5. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 72 seconds per run: cost approximately 0.007 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 71 seconds per run: cost approximately 0.006 dollars.

Zstd Compression

Performance / Cost - Compression Level: 19, Long Mode - Compression Speed

OpenBenchmarking.orgMB/s Per Dollar, More Is BetterZstd Compression 1.5.4Performance / Cost - Compression Level: 19, Long Mode - Compression Speedn2-standard-8 CLXc2-standard-8 CLXt2d-standard-8 AMDn2-highcpu-8 CLXc3-highcpu-8 SPRc2d-highcpu-8 AMD30060090012001500641.25732.50774.29840.001083.331342.001. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 74 seconds per run: cost approximately 0.008 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 68 seconds per run: cost approximately 0.008 dollars. 3. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 71 seconds per run: cost approximately 0.007 dollars. 4. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 74 seconds per run: cost approximately 0.006 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 65 seconds per run: cost approximately 0.006 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 63 seconds per run: cost approximately 0.005 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch x Dollar, Fewer Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Streamn2-standard-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPRn2-highcpu-8 CLXc2d-highcpu-8 AMD0.08660.17320.25980.34640.4330.3850.3430.3110.2420.2330.2101. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 42 seconds per run: cost approximately 0.005 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 45 seconds per run: cost approximately 0.005 dollars. 3. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 42 seconds per run: cost approximately 0.004 dollars. 4. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 40 seconds per run: cost approximately 0.004 dollars. 5. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 42 seconds per run: cost approximately 0.003 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 40 seconds per run: cost approximately 0.003 dollars.

Neural Magic DeepSparse

Performance / Cost - Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream

OpenBenchmarking.orgms/batch x Dollar, Fewer Is BetterNeural Magic DeepSparse 1.3.2Performance / Cost - Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Streamn2-standard-8 CLXc2-standard-8 CLXn2-highcpu-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPRc2d-highcpu-8 AMD0.21170.42340.63510.84681.05850.9410.8570.7760.7660.6160.5571. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 60 seconds per run: cost approximately 0.006 dollars. 2. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 55 seconds per run: cost approximately 0.006 dollars. 3. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 57 seconds per run: cost approximately 0.005 dollars. 4. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 52 seconds per run: cost approximately 0.005 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 53 seconds per run: cost approximately 0.005 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 49 seconds per run: cost approximately 0.004 dollars.

Blender

Performance / Cost - Blend File: BMW27 - Compute: CPU-Only

OpenBenchmarking.orgSeconds x Dollar, Fewer Is BetterBlender 3.4Performance / Cost - Blend File: BMW27 - Compute: CPU-Onlyc3-highcpu-8 SPR36912159.4571. $0.3448 reported cost per hour, test took average of 5 minutes, 16 seconds per run: cost approximately 0.03 dollars.

OpenCV

Performance / Cost - Test: Core

OpenBenchmarking.orgms x Dollar, Fewer Is BetterOpenCV 4.7Performance / Cost - Test: Coren2-standard-8 CLXn2-highcpu-8 CLXc2-standard-8 CLXt2d-standard-8 AMDc3-highcpu-8 SPRc2d-highcpu-8 AMD100020003000400050004813.373539.382427.09813.99698.98565.401. n2-standard-8 CLX: $0.388472 reported cost per hour, test took average of 209 seconds per run: cost approximately 0.023 dollars. 2. n2-highcpu-8 CLX: $0.286784 reported cost per hour, test took average of 208 seconds per run: cost approximately 0.017 dollars. 3. c2-standard-8 CLX: $0.4176 reported cost per hour, test took average of 143 seconds per run: cost approximately 0.017 dollars. 4. t2d-standard-8 AMD: $0.337968 reported cost per hour, test took average of 91 seconds per run: cost approximately 0.009 dollars. 5. c3-highcpu-8 SPR: $0.3448 reported cost per hour, test took average of 87 seconds per run: cost approximately 0.008 dollars. 6. c2d-highcpu-8 AMD: $0.2998 reported cost per hour, test took average of 81 seconds per run: cost approximately 0.007 dollars.

Phoronix Test Suite v10.8.5