eoy2024 Benchmarks for a future article. AMD EPYC 4124P 4-Core testing with a Supermicro AS-3015A-I H13SAE-MF v1.00 (2.1 BIOS) and ASPEED on Ubuntu 24.04 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2412201-NE-EOY20246700&grr&sro .
eoy2024 Processor Motherboard Chipset Memory Disk Graphics Audio Monitor Network OS Kernel Desktop Display Server Compiler File-System Screen Resolution a 4484PX px 4464p 4464p epyc 4364P 4584PX EPYC 4584PX amd 45 41 41 b AMD EPYC 4564P 16-Core @ 5.88GHz (16 Cores / 32 Threads) Supermicro AS-3015A-I H13SAE-MF v1.00 (2.1 BIOS) AMD Device 14d8 2 x 32GB DRAM-4800MT/s Micron MTC20C2085S1EC48BA1 BC 3201GB Micron_7450_MTFDKCC3T2TFS + 960GB SAMSUNG MZ1L2960HCJR-00A07 ASPEED AMD Rembrandt Radeon HD Audio VA2431 2 x Intel I210 Ubuntu 24.04 6.8.0-11-generic (x86_64) GNOME Shell 45.3 X Server 1.21.1.11 GCC 13.2.0 ext4 1024x768 AMD EPYC 4484PX 12-Core @ 5.66GHz (12 Cores / 24 Threads) 6.12.2-061202-generic (x86_64) AMD EPYC 4464P 12-Core @ 5.48GHz (12 Cores / 24 Threads) AMD EPYC 4364P 8-Core @ 5.57GHz (8 Cores / 16 Threads) AMD EPYC 4584PX 16-Core @ 5.76GHz (16 Cores / 32 Threads) AMD EPYC 4124P 4-Core @ 5.17GHz (4 Cores / 8 Threads) 960GB SAMSUNG MZ1L2960HCJR-00A07 + 3201GB Micron_7450_MTFDKCC3T2TFS OpenBenchmarking.org Kernel Details - Transparent Huge Pages: madvise Compiler 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-backtrace --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-13-fxIygj/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-fxIygj/gcc-13-13.2.0/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 - a: Scaling Governor: amd-pstate-epp performance (EPP: performance) - CPU Microcode: 0xa601209 - 4484PX: Scaling Governor: amd-pstate-epp performance (Boost: Enabled EPP: performance) - CPU Microcode: 0xa601209 - px: Scaling Governor: amd-pstate-epp performance (Boost: Enabled EPP: performance) - CPU Microcode: 0xa601209 - 4464p: Scaling Governor: amd-pstate-epp performance (Boost: Enabled EPP: performance) - CPU Microcode: 0xa601209 - 4464p epyc: Scaling Governor: amd-pstate-epp performance (Boost: Enabled EPP: performance) - CPU Microcode: 0xa601209 - 4364P: Scaling Governor: amd-pstate-epp performance (Boost: Enabled EPP: performance) - CPU Microcode: 0xa601209 - 4584PX: Scaling Governor: amd-pstate-epp performance (Boost: Enabled EPP: performance) - CPU Microcode: 0xa601209 - EPYC 4584PX amd: Scaling Governor: amd-pstate-epp performance (Boost: Enabled EPP: performance) - CPU Microcode: 0xa601209 - 45: Scaling Governor: amd-pstate-epp performance (Boost: Enabled EPP: performance) - CPU Microcode: 0xa601209 - 41: Scaling Governor: amd-pstate-epp performance (Boost: Enabled EPP: performance) - CPU Microcode: 0xa601209 - 41 b: Scaling Governor: amd-pstate-epp performance (Boost: Enabled EPP: performance) - CPU Microcode: 0xa601209 Java Details - OpenJDK Runtime Environment (build 21.0.2+13-Ubuntu-2) Python Details - Python 3.12.3 Security Details - a: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + 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 / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - 4484PX: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + 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 / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected - px: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + 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 / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected - 4464p: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + 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 / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected - 4464p epyc: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + 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 / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected - 4364P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + 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 / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected - 4584PX: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + 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 / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 4584PX amd: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + 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 / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected - 45: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + 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 / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected - 41: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + 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 / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected - 41 b: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + 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 / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected
eoy2024 quantlib: S svt-av1: Preset 3 - Beauty 4K 10-bit relion: Basic - CPU blender: Barbershop - CPU-Only whisper-cpp: ggml-medium.en - 2016 State of the Union cp2k: H20-256 couchdb: 500 - 3000 - 30 quantlib: XXS whisperfile: Medium llamafile: wizardcoder-python-34b-v1.0.Q6_K - Prompt Processing 2048 couchdb: 300 - 3000 - 30 svt-av1: Preset 3 - Bosphorus 4K byte: Whetstone Double whisper-cpp: ggml-small.en - 2016 State of the Union llamafile: wizardcoder-python-34b-v1.0.Q6_K - Text Generation 128 blender: Pabellon Barcelona - CPU-Only llamafile: mistral-7b-instruct-v0.2.Q5_K_M - Prompt Processing 2048 svt-av1: Preset 5 - Beauty 4K 10-bit couchdb: 100 - 3000 - 30 byte: Pipe byte: Dhrystone 2 byte: System Call blender: Classroom - CPU-Only rustls: handshake-resume - TLS13_CHACHA20_POLY1305_SHA256 rustls: handshake-ticket - TLS13_CHACHA20_POLY1305_SHA256 llamafile: wizardcoder-python-34b-v1.0.Q6_K - Prompt Processing 1024 llama-cpp: CPU BLAS - Llama-3.1-Tulu-3-8B-Q8_0 - Prompt Processing 2048 llama-cpp: CPU BLAS - Mistral-7B-Instruct-v0.3-Q8_0 - Prompt Processing 2048 xnnpack: QS8MobileNetV2 xnnpack: FP16MobileNetV3Small xnnpack: FP16MobileNetV3Large xnnpack: FP16MobileNetV2 xnnpack: FP32MobileNetV3Small xnnpack: FP32MobileNetV3Large xnnpack: FP32MobileNetV2 xnnpack: FP32MobileNetV1 xnnpack: FP16MobileNetV1 rustls: handshake-ticket - TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384 whisperfile: Small rustls: handshake-resume - TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384 openssl: ChaCha20 openssl: ChaCha20-Poly1305 openssl: AES-256-GCM openssl: AES-128-GCM llamafile: mistral-7b-instruct-v0.2.Q5_K_M - Text Generation 128 svt-av1: Preset 3 - Bosphorus 1080p couchdb: 500 - 1000 - 30 gcrypt: astcenc: Very Thorough astcenc: Exhaustive ospray: particle_volume/scivis/real_time ospray: particle_volume/pathtracer/real_time svt-av1: Preset 8 - Beauty 4K 10-bit gromacs: water_GMX50_bare build2: Time To Compile primesieve: 1e13 llamafile: mistral-7b-instruct-v0.2.Q5_K_M - Prompt Processing 1024 blender: Junkshop - CPU-Only cassandra: Writes blender: Fishy Cat - CPU-Only couchdb: 300 - 1000 - 30 pyperformance: async_tree_io openvino-genai: Gemma-7b-int4-ov - CPU - Time Per Output Token openvino-genai: Gemma-7b-int4-ov - CPU - Time To First Token openvino-genai: Gemma-7b-int4-ov - CPU llamafile: Llama-3.2-3B-Instruct.Q6_K - Prompt Processing 2048 rustls: handshake - TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384 llamafile: wizardcoder-python-34b-v1.0.Q6_K - Prompt Processing 512 svt-av1: Preset 5 - Bosphorus 4K ospray: particle_volume/ao/real_time cp2k: Fayalite-FIST whisper-cpp: ggml-base.en - 2016 State of the Union pyperformance: asyncio_tcp_ssl pyperformance: xml_etree numpy: simdjson: Kostya llama-cpp: CPU BLAS - Llama-3.1-Tulu-3-8B-Q8_0 - Prompt Processing 1024 llama-cpp: CPU BLAS - Mistral-7B-Instruct-v0.3-Q8_0 - Prompt Processing 1024 blender: BMW27 - CPU-Only llama-cpp: CPU BLAS - Llama-3.1-Tulu-3-8B-Q8_0 - Text Generation 128 llamafile: Llama-3.2-3B-Instruct.Q6_K - Text Generation 128 namd: STMV with 1,066,628 Atoms llama-cpp: CPU BLAS - Mistral-7B-Instruct-v0.3-Q8_0 - Text Generation 128 openvino-genai: Falcon-7b-instruct-int4-ov - CPU - Time Per Output Token openvino-genai: Falcon-7b-instruct-int4-ov - CPU - Time To First Token openvino-genai: Falcon-7b-instruct-int4-ov - CPU svt-av1: Preset 13 - Beauty 4K 10-bit stockfish: Chess Benchmark pyperformance: python_startup simdjson: LargeRand mt-dgemm: Sustained Floating-Point Rate build-eigen: Time To Compile couchdb: 100 - 1000 - 30 onednn: Recurrent Neural Network Training - CPU renaissance: ALS Movie Lens onednn: Recurrent Neural Network Inference - CPU simdjson: DistinctUserID stockfish: Chess Benchmark simdjson: TopTweet simdjson: PartialTweets llamafile: TinyLlama-1.1B-Chat-v1.0.BF16 - Text Generation 128 renaissance: In-Memory Database Shootout renaissance: Akka Unbalanced Cobwebbed Tree renaissance: Apache Spark PageRank llamafile: mistral-7b-instruct-v0.2.Q5_K_M - Prompt Processing 512 cp2k: H20-64 renaissance: Savina Reactors.IO renaissance: Gaussian Mixture Model pyperformance: gc_collect renaissance: Apache Spark Bayes renaissance: Finagle HTTP Requests renaissance: Rand Forest ospray: gravity_spheres_volume/dim_512/scivis/real_time ospray: gravity_spheres_volume/dim_512/ao/real_time onnx: ResNet101_DUC_HDC-12 - CPU - Standard onnx: ResNet101_DUC_HDC-12 - CPU - Standard renaissance: Scala Dotty renaissance: Genetic Algorithm Using Jenetics + Futures onnx: fcn-resnet101-11 - CPU - Standard onnx: fcn-resnet101-11 - CPU - Standard onnx: GPT-2 - CPU - Standard onnx: GPT-2 - CPU - Standard ospray: gravity_spheres_volume/dim_512/pathtracer/real_time onnx: ZFNet-512 - CPU - Standard onnx: ZFNet-512 - CPU - Standard onnx: bertsquad-12 - CPU - Standard onnx: bertsquad-12 - CPU - Standard onnx: T5 Encoder - CPU - Standard onnx: T5 Encoder - CPU - Standard onnx: yolov4 - CPU - Standard onnx: yolov4 - CPU - Standard onnx: ArcFace ResNet-100 - CPU - Standard onnx: ArcFace ResNet-100 - CPU - Standard onnx: Faster R-CNN R-50-FPN-int8 - CPU - Standard onnx: Faster R-CNN R-50-FPN-int8 - CPU - Standard onnx: CaffeNet 12-int8 - CPU - Standard onnx: CaffeNet 12-int8 - CPU - Standard onnx: ResNet50 v1-12-int8 - CPU - Standard onnx: ResNet50 v1-12-int8 - CPU - Standard onnx: super-resolution-10 - CPU - Standard onnx: super-resolution-10 - CPU - Standard llama-cpp: CPU BLAS - granite-3.0-3b-a800m-instruct-Q8_0 - Prompt Processing 2048 pyperformance: asyncio_websockets llamafile: Llama-3.2-3B-Instruct.Q6_K - Prompt Processing 1024 llamafile: wizardcoder-python-34b-v1.0.Q6_K - Prompt Processing 256 llamafile: TinyLlama-1.1B-Chat-v1.0.BF16 - Prompt Processing 2048 svt-av1: Preset 5 - Bosphorus 1080p litert: Inception V4 litert: Inception ResNet V2 litert: NASNet Mobile litert: DeepLab V3 litert: Mobilenet Float litert: SqueezeNet litert: Quantized COCO SSD MobileNet v1 litert: Mobilenet Quant rustls: handshake-ticket - TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 llama-cpp: CPU BLAS - Llama-3.1-Tulu-3-8B-Q8_0 - Prompt Processing 512 llama-cpp: CPU BLAS - Mistral-7B-Instruct-v0.3-Q8_0 - Prompt Processing 512 rustls: handshake-resume - TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 astcenc: Thorough llamafile: wizardcoder-python-34b-v1.0.Q6_K - Text Generation 16 financebench: Bonds OpenMP etcpak: Multi-Threaded - ETC2 whisperfile: Tiny rustls: handshake - TLS13_CHACHA20_POLY1305_SHA256 svt-av1: Preset 8 - Bosphorus 4K rustls: handshake - TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 pyperformance: django_template openvino-genai: Phi-3-mini-128k-instruct-int4-ov - CPU - Time Per Output Token openvino-genai: Phi-3-mini-128k-instruct-int4-ov - CPU - Time To First Token openvino-genai: Phi-3-mini-128k-instruct-int4-ov - CPU povray: Trace Time namd: ATPase with 327,506 Atoms llamafile: mistral-7b-instruct-v0.2.Q5_K_M - Prompt Processing 256 pyperformance: raytrace financebench: Repo OpenMP pyperformance: crypto_pyaes pyperformance: chaos pyperformance: go pyperformance: float x265: Bosphorus 4K pyperformance: regex_compile llamafile: Llama-3.2-3B-Instruct.Q6_K - Prompt Processing 512 llama-cpp: CPU BLAS - granite-3.0-3b-a800m-instruct-Q8_0 - Prompt Processing 1024 y-cruncher: 1B pyperformance: pathlib pyperformance: pickle_pure_python onednn: Deconvolution Batch shapes_1d - CPU llamafile: mistral-7b-instruct-v0.2.Q5_K_M - Text Generation 16 llamafile: TinyLlama-1.1B-Chat-v1.0.BF16 - Prompt Processing 1024 pyperformance: json_loads astcenc: Fast pyperformance: nbody compress-7zip: Decompression Rating compress-7zip: Compression Rating svt-av1: Preset 13 - Bosphorus 4K svt-av1: Preset 8 - Bosphorus 1080p onednn: IP Shapes 1D - CPU astcenc: Medium llama-cpp: CPU BLAS - granite-3.0-3b-a800m-instruct-Q8_0 - Text Generation 128 llamafile: Llama-3.2-3B-Instruct.Q6_K - Prompt Processing 256 llama-cpp: CPU BLAS - granite-3.0-3b-a800m-instruct-Q8_0 - Prompt Processing 512 llamafile: Llama-3.2-3B-Instruct.Q6_K - Text Generation 16 y-cruncher: 500M primesieve: 1e12 llamafile: TinyLlama-1.1B-Chat-v1.0.BF16 - Prompt Processing 512 llamafile: TinyLlama-1.1B-Chat-v1.0.BF16 - Text Generation 16 onednn: IP Shapes 3D - CPU x265: Bosphorus 1080p onednn: Convolution Batch Shapes Auto - CPU svt-av1: Preset 13 - Bosphorus 1080p llamafile: TinyLlama-1.1B-Chat-v1.0.BF16 - Prompt Processing 256 onednn: Deconvolution Batch shapes_3d - CPU openssl: SHA256 a 4484PX px 4464p 4464p epyc 4364P 4584PX EPYC 4584PX amd 45 41 41 b 12.7476 1.422 944.27 506.2 700.91 592.857 511.775 13.432 534.919 12288 367.83 9.59 343491.9 245.07838 1.99 166.12 32768 6.504 232.188 48806257.1 1866536062.7 49140426.6 143.36 388077.69 404263.45 6144 63.09 62.97 844 920 1498 1190 979 1810 1495 1252 1143 1553632.14 195.41642 1820810.21 130588495050 92393529340 97172751700 104784522170 10.47 29.573 148.049 162.125 2.741 1.6844 8.98486 236.245 12.468 1.692 92.053 78.498 16384 73.56 271333 71.35 106.13 755 101.72 106.62 9.83 32768 423535.68 3072 34.538 9.00917 94.032 87.48973 645 35.8 775.75 5.97 70.85 69.26 53.55 6.88 20.13 0.75656 7.24 77.34 86.06 12.93 18.588 54752796 5.77 1.83 1141.194104 58.655 69.929 1372.03 9805.7 700.859 10.46 46507038 10.46 9.76 26.28 3256.1 4403.8 2412.2 8192 58.191 3506.4 3399.5 677 490.0 2319.4 414.4 7.58789 7.63944 648.522 1.54196 477.0 732.8 310.875 3.2167 7.42776 134.596 8.82093 9.76985 102.331 64.141 15.5899 6.39112 156.453 90.4523 11.0552 23.553 42.4537 21.2429 47.0691 1.57084 636.318 2.55898 390.597 7.08601 141.117 279.04 315 16384 1536 32768 101.971 21477.8 19530.2 16936 3579.67 1211.48 1794.11 2129.52 823.17 2620332 70.76 68.4 3563852.57 20.3025 1.78 33061.21875 577.817 41.70935 76454.45 102.005 80462.6 20.7 51.86 55.93 19.28 18.542 2.79632 4096 175 21418.445312 41.7 38.2 77.8 50.7 32.57 69.8 8192 355.09 18.485 14.2 165 2.97612 10.22 16384 12.1 396.6495 59 165916 163859 212.52 339.023 1.12573 156.2217 47.72 4096 327.3 19.03 8.772 6.347 8192 24.59 4.058 114.45 6.67287 842.558 4096 2.41294 11.8647 1.188 729.4 679.34 809.78969 628.104 559.346 12.1169 473.55091 12288 406.12 7.684 244075.3 268.23891 2.05 226.34 32768 5.602 253.99 33443359.2 1346521770.3 30761218.9 197.2 333882.92 344296.24 6144 63.8 63.61 717 779 1467 1217 809 1515 1365 1257 1383 1329363.1 173.38197 1586292.42 97105235690 68816544020 71160291870 76496336760 10.91 25.446 164.468 171.023 1.9412 1.1887 6.44913 199.023 10.967 1.577 111.651 110.608 16384 97.01 174960 96.67 117.566 666 97.79 121.48 10.23 32768 306153.2 3072 29.094 6.52776 92.21 92.70933 590 36.8 745.59 6.11 66.57 66.85 74.08 7.11 20.39 0.65119 7.41 74.65 93.01 13.4 17.406 45267546 6.08 1.84 842.730642 67.364 75.901 1898.36 9378.8 965.015 10.76 33702298 10.82 10.1 27.59 3241.5 4038.4 2138.1 8192 53.005 3655.8 3860.6 699 513.2 2492.2 422.0 5.54888 5.63122 850.141 1.17627 428.6 904.0 355.751 2.81093 6.25815 159.71 6.41198 9.01322 110.94 68.9051 14.5122 4.80287 208.174 93.1605 10.7338 26.7478 37.3832 24.9402 40.0935 1.06188 941.401 2.80544 356.409 7.98873 125.172 222.75 321 16384 1536 32768 88.415 22083.3 19477.8 8057.56 2343.38 1244.7 1809.18 1420.15 848.943 2282729.64 69.11 68.2 3035330.21 14.17 1.83 34600.773438 410.726 37.13462 57716.64 85.201 59308.75 21 49.31 58.91 20.28 25.264 2.38124 4096 182 22320.332031 43.1 39.7 78.6 51.3 27.16 71.7 8192 232.26 18.379 14.4 169 3.40293 10.45 16384 12.4 278.2445 59.5 125698 141263 198.112 287.047 1.93806 109.0265 52.3 4096 243.14 19.49 8.688 9.116 8192 25.86 2.73072 101.37 4.11551 776.115 4096 3.5084 11.839 1.184 733.02 678.4 809.489 631.31 560.7 12.1057 475.51084 12288 408.483 7.646 244131 266.81425 2.05 224.64 32768 5.551 254.733 33381363.1 1340340196.6 30701622.8 197.53 333574.3 342775.29 6144 63.79 63.41 723 798 1527 1248 837 1574 1368 1272 1386 1340712.85 167.89219 1572010.68 97019897450 68678955550 70902656480 76184405610 10.93 25.447 164.812 163.839 1.9391 1.1862 6.52304 197.2 10.855 1.575 113.78 110.709 16384 97.1 173946 97.09 119.349 656 97.61 122.3 10.24 32768 304060.28 3072 28.824 6.52206 94.896 93.45463 590 36.5 831.42 5.45 66.35 66.52 73.16 7.12 20.51 0.65448 7.44 74.54 93 13.41 17.355 42973396 6.09 1.84 842.012831 67.076 76.389 1895.68 9275.7 966.013 8.97 33871595 10.51 8.35 27.8 3175.6 4002.3 2229.7 8192 52.724 3676.0 3815.2 706 474.9 2483.1 453.2 5.6147 5.71084 854.334 1.1705 436.2 920.7 357.602 2.79638 6.33034 157.893 6.4074 9.01687 110.892 68.6104 14.5747 4.85142 206.091 93.3441 10.7127 26.9485 37.1048 23.0604 43.362 1.066 937.778 2.80695 356.194 7.99486 125.076 208.99 322 16384 1536 32768 88.27 22752.4 19490.7 7931.64 2359.99 1244.51 1821.35 1417.35 849.209 2292879.44 67.95 68.81 3038723.48 14.1464 1.84 34896.835938 409.875 38.71828 57688.08 84.998 59206.34 21.2 49.28 58.86 20.29 25.328 2.35379 4096 182 22318.738281 43.3 39.4 79.4 50.8 26.94 72.5 8192 244.77 18.365 14.4 168 3.40628 10.45 16384 12.5 277.2994 59.2 125605 142213 194.024 286.962 1.93913 108.8588 52.37 4096 232.86 19.5 8.623 9.147 8192 25.94 2.72942 101.25 4.13321 769.818 4096 3.51243 10.2873 1.129 937.33 739.91 838.18425 676.742 551.346 10.5788 491.98412 12288 401.989 7.252 242679 280.03419 2.05 243.35 32768 5.445 259.196 32471450.7 1272938788 29969810.9 212.3 328877.92 336267.09 6144 60.68 60.47 703 759 1429 1208 799 1473 1357 1267 1418 1297646.72 178.32917 1543373.12 93956743710 66548747470 67431535960 72709954220 10.93 24.854 164.358 156.343 1.8546 1.1365 6.29615 199.512 10.914 1.398 121.737 118.664 16384 104.7 173633 105.45 118.161 738 98.33 124.38 10.17 32768 301017.23 3072 27.36 6.28807 100.486 97.94408 626 35.3 781.66 5.78 63.71 64.15 79.19 7.09 20.48 0.61452 7.42 74.94 93 13.34 18.002 43158853 5.8 1.81 818.678141 69.379 76.696 1911.44 9243.0 1009.58 10.54 30315972 10.66 9.8 27.36 3288.5 4084.6 2366.1 8192 58.397 3523.1 3647.2 671 515.3 2650.9 415.3 5.35756 5.47043 872.742 1.14581 424.6 924.5 386.563 2.58688 7.33331 136.299 6.17399 9.85896 101.413 68.0775 14.6885 5.95035 168.036 95.2507 10.4982 27.6828 36.1205 21.5708 46.3543 1.3797 724.533 2.77422 360.354 7.41037 134.941 213.02 315 16384 1536 32768 86.96 23141.4 20123.5 7759.69 2272.28 1248.12 1820.34 1431.19 859.757 2283083.54 64.96 65.24 3061522.26 13.523 1.83 33551.523438 388.404 40.847 56047.38 82.525 57954.19 20.6 50.45 59.87 19.82 26.609 2.21645 4096 177 21588.90625 41.8 39 78.3 49.7 26.09 69.5 8192 235.01 19.364 14.1 164 3.43287 10.48 16384 12 263.5475 57.2 118569 135033 192.171 282.008 1.88731 103.8089 50.5 4096 235.94 19.47 9.041 9.701 8192 25.77 3.75003 101.23 6.36475 752.155 4096 3.38958 10.2531 1.132 946.43 731.39 839.98575 678.676 551.6 10.5616 492.69519 12288 405.448 7.232 242340.5 280.47206 2.06 245.12 32768 5.422 259.055 32439611.6 1263644317.8 29923094.4 211.4 326342.42 335411.96 6144 60.8 60.3 703 766 1427 1210 791 1542 1327 1246 1414 1300937.3 177.87 1532788.04 93907572870 66497739430 67396254380 72667527300 10.91 24.805 163.363 156.712 1.8559 1.1368 6.27188 199.546 10.885 1.393 117.067 118.848 16384 104.01 174025 105.24 116.417 742 98.03 124.52 10.2 32768 301217.61 3072 27.564 6.29211 97.446 97.54562 628 35.6 783.43 6.02 63.5 62.84 79.24 7.1 20.5 0.61093 7.44 75.06 94.77 13.32 17.944 42042951 5.81 1.73 812.147089 69.578 76.577 1913.86 8980.8 1012.19 10.28 29940074 10.69 9.63 27.35 3275.5 4098.5 2373.0 8192 58.864 3688.6 3703.2 671 513.3 2708.1 412.9 5.36189 5.47057 864.81 1.15632 442.9 884.0 386.061 2.59025 7.36544 135.718 6.17624 9.83893 101.618 68.0934 14.685 6.0047 166.512 95.3576 10.4865 27.4729 36.3968 21.6933 46.0921 1.33516 748.66 2.76714 361.297 7.43372 134.517 212.3 316 16384 1536 32768 86.386 22928.3 20503.5 7640.81 2289.58 1248.6 1822.7 1415.11 857.489 2304145.79 64.43 64.68 3082657.5 13.5139 1.83 33538.269531 388.128 38.57843 56176.9 82.164 58056.01 20.6 50.46 59.86 19.82 26.595 2.20728 4096 176 21612.042969 42 38.6 76 49.7 25.85 70 8192 227.71 19.278 14.1 164 3.42547 10.45 16384 12 263.3662 58.6 118587 133804 190.343 281.892 1.88919 103.6744 50.43 4096 242.1 19.53 9.067 9.705 8192 25.78 3.74578 101.33 6.3715 748.147 4096 3.38823 8.02696 0.872 1369.225 978.03 966.00638 740.089 545.967 8.19153 528.33262 12288 398.386 5.373 174760.6 314.47722 1.93 320.59 32768 3.802 239.078 23448988.5 966576789.4 21559197 276.44 272035.68 265949.2 6144 55.95 56.05 400 455 1067 1088 335 753 708 1056 1578 1187839.34 193.93664 1298373.66 68048329660 48278109930 49930945600 53755663700 10.32 18.744 163.915 154.317 1.4052 0.8612 4.62423 177.118 6.458 1.104 154.429 150.668 16384 141.99 139110 136.51 118.317 726 105.09 154.83 9.52 32768 225379.66 3072 19.85 4.63965 108.436 103.71549 610 34.7 813.31 6.06 59.27 59.54 103.97 6.64 19.6 0.49771 7.02 79.41 115.73 12.59 9.188 31514937 6.34 1.81 629.16512 84.477 73.237 2422.83 5993.4 1270.76 10.85 24017024 10.86 10.04 25.52 3244.8 3720.1 2506.8 8192 68.171 3349.5 3397.0 649 483.4 1705.5 381.6 3.83152 3.91749 1091.32 0.916317 464.5 749.9 487.301 2.05211 7.96069 125.52 4.52318 10.2835 97.2268 50.4698 19.8127 7.06075 141.568 75.5414 13.2372 25.5845 39.0828 22.3192 44.8002 1.31121 762.248 2.65482 376.531 6.36564 157.078 171.68 308 16384 1536 32768 65.318 24502.1 21153.5 3808.8 1569.75 1044.53 1654.65 1494.51 935.994 2353677.96 60.87 61.09 2874011.56 10.3392 1.76 32715.808594 304.244 41.29562 43571.86 65.194 45095.35 20.1 52.64 71.58 19 34.917 1.74038 4096 174 21121.613281 40.7 38.3 75.3 48.2 18.85 68.1 8192 185.97 23.886 13.5 162 3.78629 9.92 16384 11.8 205.0144 56.8 88229 101872 156.598 215.213 1.90686 80.3686 51.86 4096 204.06 18.83 10.954 12.273 8192 24.27 4.45096 88.71 7.87367 652.965 4096 3.78008 14.0171 1.426 551.722 514.7 686.87944 517.664 553.805 14.624 425.50969 12288 397.426 9.329 329601.4 233.68441 2.03 171.82 32768 6.327 252.825 44169926.7 1765175324.6 40531084 149.37 379874.32 393408.85 6144 86.17 86.56 809 877 1527 1224 933 1760 1484 1194 1200 1559859.89 155.13923 1863395.07 127793294810 90711447140 93168087590 100450008810 10.81 29.067 163.805 150.639 2.5634 1.5697 8.49097 218.181 12.086 1.817 91.724 84.489 16384 73.87 235893 73.96 116.241 653 98.91 105.36 10.11 32768 435194.2 3072 34.164 8.51573 84.802 84.68398 582 36.1 806.83 6.07 90.01 91.12 55.26 7 20.23 0.75987 7.32 75.06 86.04 13.32 17.845 57247763 6.01 1.82 1091.011509 58.818 75.893 1320.99 9891.4 672.506 11.05 44446458 11.25 9.85 27.2 3232.8 4384.0 2182.5 8192 43.167 3582.4 3806.5 682 477.0 2627.3 420.7 7.32353 7.3876 650.96 1.53619 441.9 1050.3 295.281 3.38658 6.02069 166.047 8.43752 8.87444 112.671 66.6172 15.0106 4.90285 203.947 88.0632 11.3551 22.8062 43.8436 20.7183 48.2636 0.994193 1005.5 2.60974 383.134 8.18519 122.168 289.41 315 16384 1536 32768 98.789 20102.8 18224 9311.77 2185.78 1165.5 1769.56 1337.78 734.352 2528614.52 92.89 93.24 3397871.03 18.7075 1.82 34053.0625 556.043 35.91499 72793.26 101.332 74468.49 20.8 50.84 56.06 19.67 19.55 2.83913 4096 176 21948.935547 42.7 39.2 78.4 51.1 32.59 71.4 8192 290.1 17.343 14.1 166 3.07219 10.3 16384 12.5 368.5451 59 164581 170361 217.908 320.184 1.12465 144.4782 49.9 4096 316.8 19.27 8.173 6.926 8192 25.32 2.74699 113.37 4.03808 842.344 4096 2.67219 14.0472 1.433 553.782 513.66 684.38662 517.918 554.397 14.6973 424.94666 12288 400.399 9.4 331202 233.05316 2.03 171.25 32768 6.391 253.493 44323920 1771958344.7 40718581.6 148.97 382691.63 394570 6144 86.47 86.44 807 871 1542 1222 931 1756 1477 1192 1196 1569308.65 155.9745 1874050.7 128265416980 90927418030 93555991350 100666591640 10.85 29.205 161.089 159.247 2.5694 1.5743 8.51233 220.799 11.897 1.808 92.021 84.263 16384 73.55 238227 73.86 116.617 651 98.64 104.38 10.14 32768 436101.34 3072 34.414 8.57146 84.211 84.02813 586 36.2 754.41 6.04 91.31 90.48 55.74 7 20.26 0.75645 7.37 75.51 86.2 13.24 17.863 58676636 6.01 1.83 1093.097867 58.692 76.669 1320.37 9969.8 670.234 11.14 45724794 9.19 9.98 27.19 3442.9 4344.8 2227.9 8192 42.51 3641.9 3773.1 686 515.6 2571.2 402.2 7.32284 7.44575 647.814 1.54365 411.9 959.3 294.927 3.39064 6.03024 165.769 8.47727 8.82581 113.294 66.9202 14.9427 4.91381 203.491 88.3622 11.3166 22.6984 44.051 20.8816 47.885 0.999701 999.973 2.61221 382.771 8.18265 122.205 289.04 317 16384 1536 32768 99.445 19859.4 18412.3 9231.97 2200.93 1175.5 1765.39 1339.69 742.029 2558621.38 93.91 93.49 3431395.73 18.7499 1.83 33913.464844 556.412 35.4262 72508.13 101.537 74683.74 21 50.68 55.47 19.73 20.325 2.85485 4096 176 21848.070312 42.7 39.1 79.6 51 32.22 69.8 8192 324.04 17.567 14.1 165 3.06341 10.37 16384 12.2 368.5734 58.8 165934 169750 218.529 327.264 1.12801 144.6262 49.76 4096 326.16 19.24 8.279 6.903 8192 25.3 2.77727 113.16 4.04877 845.584 4096 2.66429 14.0431 1.427 549.8 513.54 684.13056 517.502 552.209 14.6226 424.57459 12288 399.261 9.345 331193.7 232.58353 2.03 170.09 32768 6.322 253.279 44309550.9 1766450267.1 40711986.9 149.37 382067.3 395112.39 6144 86.68 86.91 801 898 1547 1223 929 1754 1478 1189 1195 1569479.7 154.43658 1862938 128278766110 91025092880 93614594180 100742301290 10.82 29.179 161.473 167.631 2.5696 1.5767 8.54395 220.728 12.112 1.811 91.017 84.315 16384 73.69 236826 73.48 116.222 633 98.31 104.87 10.17 32768 435926.4 3072 34.66 8.54878 84.072 84.11276 581 36.1 833.64 5.54 90.96 91.17 55.18 7 20.29 0.76227 7.37 74.86 87.34 13.36 17.836 55762702 6.03 1.83 1092.088495 57.18 76.662 1320.23 9972.0 670.7 9.15 48797079 9.4 10.32 27.21 3408.4 4359.3 2179.0 8192 42.469 3589.4 3829.9 689 500.0 2570.5 407.7 7.31594 7.46399 645.208 1.54988 455.6 891.2 293.961 3.40178 6.01631 166.168 8.49646 8.83524 113.171 67.1849 14.8837 4.90313 203.934 88.4104 11.3105 22.9923 43.4886 20.9753 47.6721 0.986643 1013.19 2.62573 380.801 8.18175 122.219 273.67 314 16384 1536 32768 98.905 19686 17824.6 9245.76 2152.48 1178.94 1763.22 1343.22 730.167 2559245.2 91.71 94.08 3379343.72 18.7661 1.86 33906.6875 556.865 35.11036 72140.75 101.522 74672.57 20.9 50.74 55.67 19.71 19.521 2.83798 4096 179 22060.152344 42.4 38.8 78.1 50.7 32.23 70.4 8192 291.02 17.364 14.2 163 3.05918 10.34 16384 12.3 369.0274 59.6 166180 169373 217.957 325.945 1.13012 144.6131 50.33 4096 337.06 19.16 8.119 6.937 8192 25.38 2.77835 113.31 4.0355 835.45 4096 2.66296 4.43021 0.443 2771.778 2077.69 1281.082 695.352 4.47643 515.495 2.808 84017.4 581.27325 667.71 2.178 314.485 11669425.7 478992410.7 10707715 560.89 135362.96 132265.65 623 714 2018 2131 412 1342 1425 2367 3512 596141.19 732992.32 33280835860 23701092760 23845653160 25376123290 10.578 212.829 165.547 0.6923 0.4244 2.29544 112.72 4.334 0.664 300.72 301.596 292.14 66243 280.26 153.029 855 106833.78 10.869 2.30223 177.144 187.28645 723 36.7 705.99 5.7 215.53 0.27352 8.007 12630264 6.93 1.7 259.146491 145.21 96.744 4726.08 7465.1 2415.25 9.92 12024682 10.02 9.32 2951.1 4474.8 2267.1 123.656 3745.8 3087.0 702 566.2 1664.4 439.9 1.90804 1.84407 2151.82 0.464723 506.5 806.4 926.756 1.07903 8.04835 124.166 2.23851 13.5323 73.8845 95.0664 10.5186 7.04122 141.955 139.659 7.16015 41.9418 23.8412 27.6891 36.107 2.01609 495.836 4.8153 207.605 13.6457 73.2707 330 37.171 49048.8 41844 5153.49 2950.8 2193.88 3173.55 2736.98 1781.6 1656876.39 2105346.65 5.0655 47198.941406 155.441 21499.73 35.849 22255.23 21.2 70.064 0.88895 182 32271.666016 43 39.9 78.4 52.8 10.91 72.6 40.562 14.6 171 7.26738 12.4 100.3659 60.4 43781 54079 85.948 123.792 3.83117 39.3205 18.165 24.723 4.94689 56.61 9.18658 380.633 7.53655 4.43091 0.441 3046.805 2095.98 1797.8625 1282.66 698.074 4.46411 845.17 12288 513.583 2.793 83540.7 583.2045 1.93 671.42 32768 2.172 316.459 11518844.2 475486444.1 10629944.8 563.14 135016.23 132025.36 6144 30.93 31.04 623 720 1998 2097 414 1359 1443 2407 3508 602002.02 310.38181 690282.02 33289542860 23701138360 23760535010 25582374640 10.46 10.554 212.528 169.141 0.6906 0.4223 2.27718 111.483 4.332 0.661 294.951 302.737 16384 294.04 66519 282.19 154.041 858 124.24 288.64 8.05 32768 105324.92 3072 10.805 2.26612 179.122 186.22906 737 36.6 700.76 5.68 31.91 32.59 214.89 6.55 20.08 0.27320 6.9 87.9 217.85 11.38 7.984 14943737 6.97 1.71 263.204392 146.655 96.554 4752.85 7147.4 2441.27 10.16 11607502 9.97 9.3 25.09 3681.9 4457.1 2292.2 8192 130.813 3818.5 3130.4 704 600.1 1669.9 485.4 1.89469 1.90481 2152.92 0.464485 515.8 801.4 927.388 1.07829 8.10767 123.269 2.22318 13.6955 73.0047 94.1216 10.6241 7.13857 140.034 139.718 7.15712 41.8199 23.9108 25.9499 38.5294 2.01913 495.078 4.94636 202.118 13.5417 73.8364 105.1 331 16384 1536 32768 37.07 49101.9 41757.9 5034.38 2994.59 2190.79 3179.55 2741.68 1785.4 1571309.62 32.97 32.94 2103587.5 5.0448 1.73 46851.421875 154.248 66.11323 21444.09 35.641 22201.03 21.3 60.07 130.47 16.65 70.774 0.89241 4096 180 32477.349609 43.1 40.2 78.7 52.9 10.89 72.7 8192 115.67 39.245 14.6 171 7.28139 9.99 16384 12.4 99.3999 60 43210 53425 85.603 122.921 3.82885 39.0377 51.5 4096 122.8 19.17 18.29 24.821 8192 23.8 4.95867 56.03 9.16956 377.791 4096 7.55315 OpenBenchmarking.org
QuantLib Size: S OpenBenchmarking.org tasks/s, More Is Better QuantLib 1.35-dev Size: S 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 4 8 12 16 20 4.43021 4.43091 8.02696 10.28730 10.25310 11.86470 14.04310 14.01710 14.04720 12.74760 11.83900 1. (CXX) g++ options: -O3 -march=native -fPIE -pie
SVT-AV1 Encoder Mode: Preset 3 - Input: Beauty 4K 10-bit OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 3 - Input: Beauty 4K 10-bit 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.3224 0.6448 0.9672 1.2896 1.612 0.443 0.441 0.872 1.129 1.132 1.188 1.427 1.426 1.433 1.422 1.184 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
RELION Test: Basic - Device: CPU OpenBenchmarking.org Seconds, Fewer Is Better RELION 5.0 Test: Basic - Device: CPU 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 700 1400 2100 2800 3500 2771.78 3046.81 1369.23 937.33 946.43 729.40 549.80 551.72 553.78 944.27 733.02 1. (CXX) g++ options: -fPIC -std=c++14 -fopenmp -O3 -rdynamic -lfftw3f -lfftw3 -ldl -ltiff -lpng -ljpeg -lmpi_cxx -lmpi
Blender Blend File: Barbershop - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.3 Blend File: Barbershop - Compute: CPU-Only 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 400 800 1200 1600 2000 2077.69 2095.98 978.03 739.91 731.39 679.34 513.54 514.70 513.66 506.20 678.40
Whisper.cpp Model: ggml-medium.en - Input: 2016 State of the Union OpenBenchmarking.org Seconds, Fewer Is Better Whisper.cpp 1.6.2 Model: ggml-medium.en - Input: 2016 State of the Union 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 400 800 1200 1600 2000 1797.86 966.01 838.18 839.99 809.79 684.13 686.88 684.39 700.91 809.49 1. (CXX) g++ options: -O3 -std=c++11 -fPIC -pthread -msse3 -mssse3 -mavx -mf16c -mfma -mavx2 -mavx512f -mavx512cd -mavx512vl -mavx512dq -mavx512bw -mavx512vbmi -mavx512vnni
CP2K Molecular Dynamics Input: H20-256 OpenBenchmarking.org Seconds, Fewer Is Better CP2K Molecular Dynamics 2024.3 Input: H20-256 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 300 600 900 1200 1500 1281.08 1282.66 740.09 676.74 678.68 628.10 517.50 517.66 517.92 592.86 631.31 1. (F9X) gfortran options: -fopenmp -march=native -mtune=native -O3 -funroll-loops -fbacktrace -ffree-form -fimplicit-none -std=f2008 -lcp2kstart -lcp2kmc -lcp2kswarm -lcp2kmotion -lcp2kthermostat -lcp2kemd -lcp2ktmc -lcp2kmain -lcp2kdbt -lcp2ktas -lcp2kgrid -lcp2kgriddgemm -lcp2kgridcpu -lcp2kgridref -lcp2kgridcommon -ldbcsrarnoldi -ldbcsrx -lcp2kdbx -lcp2kdbm -lcp2kshg_int -lcp2keri_mme -lcp2kminimax -lcp2khfxbase -lcp2ksubsys -lcp2kxc -lcp2kao -lcp2kpw_env -lcp2kinput -lcp2kpw -lcp2kgpu -lcp2kfft -lcp2kfpga -lcp2kfm -lcp2kcommon -lcp2koffload -lcp2kmpiwrap -lcp2kbase -ldbcsr -lsirius -lspla -lspfft -lsymspg -lvdwxc -l:libhdf5_fortran.a -l:libhdf5.a -lz -lgsl -lelpa_openmp -lcosma -lcosta -lscalapack -lxsmmf -lxsmm -ldl -lpthread -llibgrpp -lxcf03 -lxc -lint2 -lfftw3_mpi -lfftw3 -lfftw3_omp -lmpi_cxx -lmpi -l:libopenblas.a -lvori -lstdc++ -lmpi_usempif08 -lmpi_mpifh -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm
Apache CouchDB Bulk Size: 500 - Inserts: 3000 - Rounds: 30 OpenBenchmarking.org Seconds, Fewer Is Better Apache CouchDB 3.4.1 Bulk Size: 500 - Inserts: 3000 - Rounds: 30 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 150 300 450 600 750 695.35 698.07 545.97 551.35 551.60 559.35 552.21 553.81 554.40 511.78 560.70 1. (CXX) g++ options: -flto -lstdc++ -shared -lei
QuantLib Size: XXS OpenBenchmarking.org tasks/s, More Is Better QuantLib 1.35-dev Size: XXS 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 4 8 12 16 20 4.47643 4.46411 8.19153 10.57880 10.56160 12.11690 14.62260 14.62400 14.69730 13.43200 12.10570 1. (CXX) g++ options: -O3 -march=native -fPIE -pie
Whisperfile Model Size: Medium OpenBenchmarking.org Seconds, Fewer Is Better Whisperfile 20Aug24 Model Size: Medium 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 200 400 600 800 1000 845.17 528.33 491.98 492.70 473.55 424.57 425.51 424.95 534.92 475.51
Llamafile Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Prompt Processing 2048 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Prompt Processing 2048 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3K 6K 9K 12K 15K 12288 12288 12288 12288 12288 12288 12288 12288 12288 12288
Apache CouchDB Bulk Size: 300 - Inserts: 3000 - Rounds: 30 OpenBenchmarking.org Seconds, Fewer Is Better Apache CouchDB 3.4.1 Bulk Size: 300 - Inserts: 3000 - Rounds: 30 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 110 220 330 440 550 515.50 513.58 398.39 401.99 405.45 406.12 399.26 397.43 400.40 367.83 408.48 1. (CXX) g++ options: -flto -lstdc++ -shared -lei
SVT-AV1 Encoder Mode: Preset 3 - Input: Bosphorus 4K OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 3 - Input: Bosphorus 4K 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 2.808 2.793 5.373 7.252 7.232 7.684 9.345 9.329 9.400 9.590 7.646 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
BYTE Unix Benchmark Computational Test: Whetstone Double OpenBenchmarking.org MWIPS, More Is Better BYTE Unix Benchmark 5.1.3-git Computational Test: Whetstone Double 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 70K 140K 210K 280K 350K 84017.4 83540.7 174760.6 242679.0 242340.5 244075.3 331193.7 329601.4 331202.0 343491.9 244131.0 1. (CC) gcc options: -pedantic -O3 -ffast-math -march=native -mtune=native -lm
Whisper.cpp Model: ggml-small.en - Input: 2016 State of the Union OpenBenchmarking.org Seconds, Fewer Is Better Whisper.cpp 1.6.2 Model: ggml-small.en - Input: 2016 State of the Union 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 130 260 390 520 650 581.27 583.20 314.48 280.03 280.47 268.24 232.58 233.68 233.05 245.08 266.81 1. (CXX) g++ options: -O3 -std=c++11 -fPIC -pthread -msse3 -mssse3 -mavx -mf16c -mfma -mavx2 -mavx512f -mavx512cd -mavx512vl -mavx512dq -mavx512bw -mavx512vbmi -mavx512vnni
Llamafile Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Text Generation 128 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Text Generation 128 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.4635 0.927 1.3905 1.854 2.3175 1.93 1.93 2.05 2.06 2.05 2.03 2.03 2.03 1.99 2.05
Blender Blend File: Pabellon Barcelona - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.3 Blend File: Pabellon Barcelona - Compute: CPU-Only 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 140 280 420 560 700 667.71 671.42 320.59 243.35 245.12 226.34 170.09 171.82 171.25 166.12 224.64
Llamafile Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Prompt Processing 2048 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Prompt Processing 2048 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 7K 14K 21K 28K 35K 32768 32768 32768 32768 32768 32768 32768 32768 32768 32768
SVT-AV1 Encoder Mode: Preset 5 - Input: Beauty 4K 10-bit OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 5 - Input: Beauty 4K 10-bit 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2 4 6 8 10 2.178 2.172 3.802 5.445 5.422 5.602 6.322 6.327 6.391 6.504 5.551 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
Apache CouchDB Bulk Size: 100 - Inserts: 3000 - Rounds: 30 OpenBenchmarking.org Seconds, Fewer Is Better Apache CouchDB 3.4.1 Bulk Size: 100 - Inserts: 3000 - Rounds: 30 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 70 140 210 280 350 314.49 316.46 239.08 259.20 259.06 253.99 253.28 252.83 253.49 232.19 254.73 1. (CXX) g++ options: -flto -lstdc++ -shared -lei
BYTE Unix Benchmark Computational Test: Pipe OpenBenchmarking.org LPS, More Is Better BYTE Unix Benchmark 5.1.3-git Computational Test: Pipe 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 10M 20M 30M 40M 50M 11669425.7 11518844.2 23448988.5 32471450.7 32439611.6 33443359.2 44309550.9 44169926.7 44323920.0 48806257.1 33381363.1 1. (CC) gcc options: -pedantic -O3 -ffast-math -march=native -mtune=native -lm
BYTE Unix Benchmark Computational Test: Dhrystone 2 OpenBenchmarking.org LPS, More Is Better BYTE Unix Benchmark 5.1.3-git Computational Test: Dhrystone 2 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 400M 800M 1200M 1600M 2000M 478992410.7 475486444.1 966576789.4 1272938788.0 1263644317.8 1346521770.3 1766450267.1 1765175324.6 1771958344.7 1866536062.7 1340340196.6 1. (CC) gcc options: -pedantic -O3 -ffast-math -march=native -mtune=native -lm
BYTE Unix Benchmark Computational Test: System Call OpenBenchmarking.org LPS, More Is Better BYTE Unix Benchmark 5.1.3-git Computational Test: System Call 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 11M 22M 33M 44M 55M 10707715.0 10629944.8 21559197.0 29969810.9 29923094.4 30761218.9 40711986.9 40531084.0 40718581.6 49140426.6 30701622.8 1. (CC) gcc options: -pedantic -O3 -ffast-math -march=native -mtune=native -lm
Blender Blend File: Classroom - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.3 Blend File: Classroom - Compute: CPU-Only 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 120 240 360 480 600 560.89 563.14 276.44 212.30 211.40 197.20 149.37 149.37 148.97 143.36 197.53
Rustls Benchmark: handshake-resume - Suite: TLS13_CHACHA20_POLY1305_SHA256 OpenBenchmarking.org handshakes/s, More Is Better Rustls 0.23.17 Benchmark: handshake-resume - Suite: TLS13_CHACHA20_POLY1305_SHA256 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 80K 160K 240K 320K 400K 135362.96 135016.23 272035.68 328877.92 326342.42 333882.92 382067.30 379874.32 382691.63 388077.69 333574.30 1. (CC) gcc options: -m64 -lgcc_s -lutil -lrt -lpthread -lm -ldl -lc -pie -nodefaultlibs
Rustls Benchmark: handshake-ticket - Suite: TLS13_CHACHA20_POLY1305_SHA256 OpenBenchmarking.org handshakes/s, More Is Better Rustls 0.23.17 Benchmark: handshake-ticket - Suite: TLS13_CHACHA20_POLY1305_SHA256 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 90K 180K 270K 360K 450K 132265.65 132025.36 265949.20 336267.09 335411.96 344296.24 395112.39 393408.85 394570.00 404263.45 342775.29 1. (CC) gcc options: -m64 -lgcc_s -lutil -lrt -lpthread -lm -ldl -lc -pie -nodefaultlibs
Llamafile Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Prompt Processing 1024 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Prompt Processing 1024 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 1300 2600 3900 5200 6500 6144 6144 6144 6144 6144 6144 6144 6144 6144 6144
Llama.cpp Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048 OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4154 Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20 40 60 80 100 30.93 55.95 60.68 60.80 63.80 86.68 86.17 86.47 63.09 63.79 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas
Llama.cpp Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 2048 OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4154 Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 2048 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20 40 60 80 100 31.04 56.05 60.47 60.30 63.61 86.91 86.56 86.44 62.97 63.41 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas
XNNPACK Model: QS8MobileNetV2 OpenBenchmarking.org us, Fewer Is Better XNNPACK b7b048 Model: QS8MobileNetV2 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 200 400 600 800 1000 623 623 400 703 703 717 801 809 807 844 723 1. (CXX) g++ options: -O3 -lrt -lm
XNNPACK Model: FP16MobileNetV3Small OpenBenchmarking.org us, Fewer Is Better XNNPACK b7b048 Model: FP16MobileNetV3Small 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 200 400 600 800 1000 714 720 455 759 766 779 898 877 871 920 798 1. (CXX) g++ options: -O3 -lrt -lm
XNNPACK Model: FP16MobileNetV3Large OpenBenchmarking.org us, Fewer Is Better XNNPACK b7b048 Model: FP16MobileNetV3Large 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 400 800 1200 1600 2000 2018 1998 1067 1429 1427 1467 1547 1527 1542 1498 1527 1. (CXX) g++ options: -O3 -lrt -lm
XNNPACK Model: FP16MobileNetV2 OpenBenchmarking.org us, Fewer Is Better XNNPACK b7b048 Model: FP16MobileNetV2 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 500 1000 1500 2000 2500 2131 2097 1088 1208 1210 1217 1223 1212 1222 1190 1248 1. (CXX) g++ options: -O3 -lrt -lm
XNNPACK Model: FP32MobileNetV3Small OpenBenchmarking.org us, Fewer Is Better XNNPACK b7b048 Model: FP32MobileNetV3Small 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 200 400 600 800 1000 412 414 335 799 791 809 929 933 931 979 837 1. (CXX) g++ options: -O3 -lrt -lm
XNNPACK Model: FP32MobileNetV3Large OpenBenchmarking.org us, Fewer Is Better XNNPACK b7b048 Model: FP32MobileNetV3Large 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 400 800 1200 1600 2000 1342 1359 753 1473 1542 1515 1754 1749 1756 1810 1574 1. (CXX) g++ options: -O3 -lrt -lm
XNNPACK Model: FP32MobileNetV2 OpenBenchmarking.org us, Fewer Is Better XNNPACK b7b048 Model: FP32MobileNetV2 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 300 600 900 1200 1500 1425 1443 708 1357 1327 1365 1478 1481 1477 1495 1368 1. (CXX) g++ options: -O3 -lrt -lm
XNNPACK Model: FP32MobileNetV1 OpenBenchmarking.org us, Fewer Is Better XNNPACK b7b048 Model: FP32MobileNetV1 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 500 1000 1500 2000 2500 2367 2407 1056 1267 1246 1257 1189 1191 1192 1252 1272 1. (CXX) g++ options: -O3 -lrt -lm
XNNPACK Model: FP16MobileNetV1 OpenBenchmarking.org us, Fewer Is Better XNNPACK b7b048 Model: FP16MobileNetV1 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 800 1600 2400 3200 4000 3512 3508 1578 1418 1414 1383 1195 1200 1196 1143 1386 1. (CXX) g++ options: -O3 -lrt -lm
Rustls Benchmark: handshake-ticket - Suite: TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384 OpenBenchmarking.org handshakes/s, More Is Better Rustls 0.23.17 Benchmark: handshake-ticket - Suite: TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 300K 600K 900K 1200K 1500K 596141.19 602002.02 1187839.34 1297646.72 1300937.30 1329363.10 1569479.70 1559859.89 1569308.65 1553632.14 1340712.85 1. (CC) gcc options: -m64 -lgcc_s -lutil -lrt -lpthread -lm -ldl -lc -pie -nodefaultlibs
Whisperfile Model Size: Small OpenBenchmarking.org Seconds, Fewer Is Better Whisperfile 20Aug24 Model Size: Small 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 70 140 210 280 350 310.38 193.94 178.33 177.87 173.38 154.44 155.14 155.97 195.42 167.89
Rustls Benchmark: handshake-resume - Suite: TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384 OpenBenchmarking.org handshakes/s, More Is Better Rustls 0.23.17 Benchmark: handshake-resume - Suite: TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 400K 800K 1200K 1600K 2000K 732992.32 690282.02 1298373.66 1543373.12 1532788.04 1586292.42 1862938.00 1863395.07 1874050.70 1820810.21 1572010.68 1. (CC) gcc options: -m64 -lgcc_s -lutil -lrt -lpthread -lm -ldl -lc -pie -nodefaultlibs
OpenSSL Algorithm: ChaCha20 OpenBenchmarking.org byte/s, More Is Better OpenSSL Algorithm: ChaCha20 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 30000M 60000M 90000M 120000M 150000M 33280835860 33289542860 68048329660 93956743710 93907572870 97105235690 128278766110 127793294810 128265416980 130588495050 97019897450 1. OpenSSL 3.0.13 30 Jan 2024 (Library: OpenSSL 3.0.13 30 Jan 2024) - Additional Parameters: -engine qatengine -async_jobs 8
OpenSSL Algorithm: ChaCha20-Poly1305 OpenBenchmarking.org byte/s, More Is Better OpenSSL Algorithm: ChaCha20-Poly1305 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20000M 40000M 60000M 80000M 100000M 23701092760 23701138360 48278109930 66548747470 66497739430 68816544020 91025092880 90711447140 90927418030 92393529340 68678955550 1. OpenSSL 3.0.13 30 Jan 2024 (Library: OpenSSL 3.0.13 30 Jan 2024) - Additional Parameters: -engine qatengine -async_jobs 8
OpenSSL Algorithm: AES-256-GCM OpenBenchmarking.org byte/s, More Is Better OpenSSL Algorithm: AES-256-GCM 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20000M 40000M 60000M 80000M 100000M 23845653160 23760535010 49930945600 67431535960 67396254380 71160291870 93614594180 93168087590 93555991350 97172751700 70902656480 1. OpenSSL 3.0.13 30 Jan 2024 (Library: OpenSSL 3.0.13 30 Jan 2024) - Additional Parameters: -engine qatengine -async_jobs 8
OpenSSL Algorithm: AES-128-GCM OpenBenchmarking.org byte/s, More Is Better OpenSSL Algorithm: AES-128-GCM 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20000M 40000M 60000M 80000M 100000M 25376123290 25582374640 53755663700 72709954220 72667527300 76496336760 100742301290 100450008810 100666591640 104784522170 76184405610 1. OpenSSL 3.0.13 30 Jan 2024 (Library: OpenSSL 3.0.13 30 Jan 2024) - Additional Parameters: -engine qatengine -async_jobs 8
Llamafile Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Text Generation 128 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Text Generation 128 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 10.46 10.32 10.93 10.91 10.91 10.82 10.81 10.85 10.47 10.93
SVT-AV1 Encoder Mode: Preset 3 - Input: Bosphorus 1080p OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 3 - Input: Bosphorus 1080p 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 7 14 21 28 35 10.58 10.55 18.74 24.85 24.81 25.45 29.18 29.07 29.21 29.57 25.45 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
Apache CouchDB Bulk Size: 500 - Inserts: 1000 - Rounds: 30 OpenBenchmarking.org Seconds, Fewer Is Better Apache CouchDB 3.4.1 Bulk Size: 500 - Inserts: 1000 - Rounds: 30 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 50 100 150 200 250 212.83 212.53 163.92 164.36 163.36 164.47 161.47 163.81 161.09 148.05 164.81 1. (CXX) g++ options: -flto -lstdc++ -shared -lei
Gcrypt Library OpenBenchmarking.org Seconds, Fewer Is Better Gcrypt Library 1.10.3 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 40 80 120 160 200 165.55 169.14 154.32 156.34 156.71 171.02 167.63 150.64 159.25 162.13 163.84 1. (CC) gcc options: -O2 -fvisibility=hidden
ASTC Encoder Preset: Very Thorough OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 5.0 Preset: Very Thorough 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.6167 1.2334 1.8501 2.4668 3.0835 0.6923 0.6906 1.4052 1.8546 1.8559 1.9412 2.5696 2.5634 2.5694 2.7410 1.9391 1. (CXX) g++ options: -O3 -flto -pthread
ASTC Encoder Preset: Exhaustive OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 5.0 Preset: Exhaustive 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.379 0.758 1.137 1.516 1.895 0.4244 0.4223 0.8612 1.1365 1.1368 1.1887 1.5767 1.5697 1.5743 1.6844 1.1862 1. (CXX) g++ options: -O3 -flto -pthread
OSPRay Benchmark: particle_volume/scivis/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPRay 3.2 Benchmark: particle_volume/scivis/real_time 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 2.29544 2.27718 4.62423 6.29615 6.27188 6.44913 8.54395 8.49097 8.51233 8.98486 6.52304
OSPRay Benchmark: particle_volume/pathtracer/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPRay 3.2 Benchmark: particle_volume/pathtracer/real_time 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 50 100 150 200 250 112.72 111.48 177.12 199.51 199.55 199.02 220.73 218.18 220.80 236.25 197.20
SVT-AV1 Encoder Mode: Preset 8 - Input: Beauty 4K 10-bit OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 8 - Input: Beauty 4K 10-bit 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 4.334 4.332 6.458 10.914 10.885 10.967 12.112 12.086 11.897 12.468 10.855 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
GROMACS Input: water_GMX50_bare OpenBenchmarking.org Ns Per Day, More Is Better GROMACS Input: water_GMX50_bare 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.4088 0.8176 1.2264 1.6352 2.044 0.664 0.661 1.104 1.398 1.393 1.577 1.811 1.817 1.808 1.692 1.575 1. GROMACS version: 2023.3-Ubuntu_2023.3_1ubuntu3
Build2 Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Build2 0.17 Time To Compile 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 70 140 210 280 350 300.72 294.95 154.43 121.74 117.07 111.65 91.02 91.72 92.02 92.05 113.78
Primesieve Length: 1e13 OpenBenchmarking.org Seconds, Fewer Is Better Primesieve 12.6 Length: 1e13 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 70 140 210 280 350 301.60 302.74 150.67 118.66 118.85 110.61 84.32 84.49 84.26 78.50 110.71 1. (CXX) g++ options: -O3
Llamafile Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Prompt Processing 1024 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Prompt Processing 1024 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 4K 8K 12K 16K 20K 16384 16384 16384 16384 16384 16384 16384 16384 16384 16384
Blender Blend File: Junkshop - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.3 Blend File: Junkshop - Compute: CPU-Only 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 60 120 180 240 300 292.14 294.04 141.99 104.70 104.01 97.01 73.69 73.87 73.55 73.56 97.10
Apache Cassandra Test: Writes OpenBenchmarking.org Op/s, More Is Better Apache Cassandra 5.0 Test: Writes 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 60K 120K 180K 240K 300K 66243 66519 139110 173633 174025 174960 236826 235893 238227 271333 173946
Blender Blend File: Fishy Cat - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.3 Blend File: Fishy Cat - Compute: CPU-Only 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 60 120 180 240 300 280.26 282.19 136.51 105.45 105.24 96.67 73.48 73.96 73.86 71.35 97.09
Apache CouchDB Bulk Size: 300 - Inserts: 1000 - Rounds: 30 OpenBenchmarking.org Seconds, Fewer Is Better Apache CouchDB 3.4.1 Bulk Size: 300 - Inserts: 1000 - Rounds: 30 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 30 60 90 120 150 153.03 154.04 118.32 118.16 116.42 117.57 116.22 116.24 116.62 106.13 119.35 1. (CXX) g++ options: -flto -lstdc++ -shared -lei
PyPerformance Benchmark: async_tree_io OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: async_tree_io 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 200 400 600 800 1000 855 858 726 738 742 666 633 653 651 755 656
OpenVINO GenAI Model: Gemma-7b-int4-ov - Device: CPU - Time Per Output Token OpenBenchmarking.org ms, Fewer Is Better OpenVINO GenAI 2024.5 Model: Gemma-7b-int4-ov - Device: CPU - Time Per Output Token 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 30 60 90 120 150 124.24 105.09 98.33 98.03 97.79 98.31 98.91 98.64 101.72 97.61
OpenVINO GenAI Model: Gemma-7b-int4-ov - Device: CPU - Time To First Token OpenBenchmarking.org ms, Fewer Is Better OpenVINO GenAI 2024.5 Model: Gemma-7b-int4-ov - Device: CPU - Time To First Token 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 60 120 180 240 300 288.64 154.83 124.38 124.52 121.48 104.87 105.36 104.38 106.62 122.30
OpenVINO GenAI Model: Gemma-7b-int4-ov - Device: CPU OpenBenchmarking.org tokens/s, More Is Better OpenVINO GenAI 2024.5 Model: Gemma-7b-int4-ov - Device: CPU 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 8.05 9.52 10.17 10.20 10.23 10.17 10.11 10.14 9.83 10.24
Llamafile Model: Llama-3.2-3B-Instruct.Q6_K - Test: Prompt Processing 2048 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: Llama-3.2-3B-Instruct.Q6_K - Test: Prompt Processing 2048 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 7K 14K 21K 28K 35K 32768 32768 32768 32768 32768 32768 32768 32768 32768 32768
Rustls Benchmark: handshake - Suite: TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384 OpenBenchmarking.org handshakes/s, More Is Better Rustls 0.23.17 Benchmark: handshake - Suite: TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 90K 180K 270K 360K 450K 106833.78 105324.92 225379.66 301017.23 301217.61 306153.20 435926.40 435194.20 436101.34 423535.68 304060.28 1. (CC) gcc options: -m64 -lgcc_s -lutil -lrt -lpthread -lm -ldl -lc -pie -nodefaultlibs
Llamafile Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Prompt Processing 512 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Prompt Processing 512 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 700 1400 2100 2800 3500 3072 3072 3072 3072 3072 3072 3072 3072 3072 3072
SVT-AV1 Encoder Mode: Preset 5 - Input: Bosphorus 4K OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 5 - Input: Bosphorus 4K 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 8 16 24 32 40 10.87 10.81 19.85 27.36 27.56 29.09 34.66 34.16 34.41 34.54 28.82 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OSPRay Benchmark: particle_volume/ao/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPRay 3.2 Benchmark: particle_volume/ao/real_time 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 2.30223 2.26612 4.63965 6.28807 6.29211 6.52776 8.54878 8.51573 8.57146 9.00917 6.52206
CP2K Molecular Dynamics Input: Fayalite-FIST OpenBenchmarking.org Seconds, Fewer Is Better CP2K Molecular Dynamics 2024.3 Input: Fayalite-FIST 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 40 80 120 160 200 177.14 179.12 108.44 100.49 97.45 92.21 84.07 84.80 84.21 94.03 94.90 1. (F9X) gfortran options: -fopenmp -march=native -mtune=native -O3 -funroll-loops -fbacktrace -ffree-form -fimplicit-none -std=f2008 -lcp2kstart -lcp2kmc -lcp2kswarm -lcp2kmotion -lcp2kthermostat -lcp2kemd -lcp2ktmc -lcp2kmain -lcp2kdbt -lcp2ktas -lcp2kgrid -lcp2kgriddgemm -lcp2kgridcpu -lcp2kgridref -lcp2kgridcommon -ldbcsrarnoldi -ldbcsrx -lcp2kdbx -lcp2kdbm -lcp2kshg_int -lcp2keri_mme -lcp2kminimax -lcp2khfxbase -lcp2ksubsys -lcp2kxc -lcp2kao -lcp2kpw_env -lcp2kinput -lcp2kpw -lcp2kgpu -lcp2kfft -lcp2kfpga -lcp2kfm -lcp2kcommon -lcp2koffload -lcp2kmpiwrap -lcp2kbase -ldbcsr -lsirius -lspla -lspfft -lsymspg -lvdwxc -l:libhdf5_fortran.a -l:libhdf5.a -lz -lgsl -lelpa_openmp -lcosma -lcosta -lscalapack -lxsmmf -lxsmm -ldl -lpthread -llibgrpp -lxcf03 -lxc -lint2 -lfftw3_mpi -lfftw3 -lfftw3_omp -lmpi_cxx -lmpi -l:libopenblas.a -lvori -lstdc++ -lmpi_usempif08 -lmpi_mpifh -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm
Whisper.cpp Model: ggml-base.en - Input: 2016 State of the Union OpenBenchmarking.org Seconds, Fewer Is Better Whisper.cpp 1.6.2 Model: ggml-base.en - Input: 2016 State of the Union 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 40 80 120 160 200 187.29 186.23 103.72 97.94 97.55 92.71 84.11 84.68 84.03 87.49 93.45 1. (CXX) g++ options: -O3 -std=c++11 -fPIC -pthread -msse3 -mssse3 -mavx -mf16c -mfma -mavx2 -mavx512f -mavx512cd -mavx512vl -mavx512dq -mavx512bw -mavx512vbmi -mavx512vnni
PyPerformance Benchmark: asyncio_tcp_ssl OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: asyncio_tcp_ssl 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 160 320 480 640 800 723 737 610 626 628 590 581 582 586 645 590
PyPerformance Benchmark: xml_etree OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: xml_etree 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 8 16 24 32 40 36.7 36.6 34.7 35.3 35.6 36.8 36.1 36.1 36.2 35.8 36.5
Numpy Benchmark OpenBenchmarking.org Score, More Is Better Numpy Benchmark 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 200 400 600 800 1000 705.99 700.76 813.31 781.66 783.43 745.59 833.64 806.83 754.41 775.75 831.42
simdjson Throughput Test: Kostya OpenBenchmarking.org GB/s, More Is Better simdjson 3.10 Throughput Test: Kostya 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2 4 6 8 10 5.70 5.68 6.06 5.78 6.02 6.11 5.54 6.07 6.04 5.97 5.45 1. (CXX) g++ options: -O3 -lrt
Llama.cpp Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024 OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4154 Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20 40 60 80 100 31.91 59.27 63.71 63.50 66.57 90.96 90.01 91.31 70.85 66.35 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas
Llama.cpp Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 1024 OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4154 Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 1024 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20 40 60 80 100 32.59 59.54 64.15 62.84 66.85 91.17 91.12 90.48 69.26 66.52 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas
Blender Blend File: BMW27 - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.3 Blend File: BMW27 - Compute: CPU-Only 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 50 100 150 200 250 215.53 214.89 103.97 79.19 79.24 74.08 55.18 55.26 55.74 53.55 73.16
Llama.cpp Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128 OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4154 Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2 4 6 8 10 6.55 6.64 7.09 7.10 7.11 7.00 7.00 7.00 6.88 7.12 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas
Llamafile Model: Llama-3.2-3B-Instruct.Q6_K - Test: Text Generation 128 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: Llama-3.2-3B-Instruct.Q6_K - Test: Text Generation 128 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 5 10 15 20 25 20.08 19.60 20.48 20.50 20.39 20.29 20.23 20.26 20.13 20.51
NAMD Input: STMV with 1,066,628 Atoms OpenBenchmarking.org ns/day, More Is Better NAMD 3.0 Input: STMV with 1,066,628 Atoms 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.1715 0.343 0.5145 0.686 0.8575 0.27352 0.27320 0.49771 0.61452 0.61093 0.65119 0.76227 0.75987 0.75645 0.75656 0.65448
Llama.cpp Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128 OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4154 Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2 4 6 8 10 6.90 7.02 7.42 7.44 7.41 7.37 7.32 7.37 7.24 7.44 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas
OpenVINO GenAI Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time Per Output Token OpenBenchmarking.org ms, Fewer Is Better OpenVINO GenAI 2024.5 Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time Per Output Token 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20 40 60 80 100 87.90 79.41 74.94 75.06 74.65 74.86 75.06 75.51 77.34 74.54
OpenVINO GenAI Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time To First Token OpenBenchmarking.org ms, Fewer Is Better OpenVINO GenAI 2024.5 Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time To First Token 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 50 100 150 200 250 217.85 115.73 93.00 94.77 93.01 87.34 86.04 86.20 86.06 93.00
OpenVINO GenAI Model: Falcon-7b-instruct-int4-ov - Device: CPU OpenBenchmarking.org tokens/s, More Is Better OpenVINO GenAI 2024.5 Model: Falcon-7b-instruct-int4-ov - Device: CPU 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 11.38 12.59 13.34 13.32 13.40 13.36 13.32 13.24 12.93 13.41
SVT-AV1 Encoder Mode: Preset 13 - Input: Beauty 4K 10-bit OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 13 - Input: Beauty 4K 10-bit 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 5 10 15 20 25 8.007 7.984 9.188 18.002 17.944 17.406 17.836 17.845 17.863 18.588 17.355 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
Stockfish Chess Benchmark OpenBenchmarking.org Nodes Per Second, More Is Better Stockfish 17 Chess Benchmark 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 13M 26M 39M 52M 65M 12630264 14943737 31514937 43158853 42042951 45267546 55762702 57247763 58676636 54752796 42973396 1. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fno-peel-loops -fno-tracer -pedantic -O3 -funroll-loops -msse -msse3 -mpopcnt -mavx2 -mbmi -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto-partition=one -flto=jobserver
PyPerformance Benchmark: python_startup OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: python_startup 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2 4 6 8 10 6.93 6.97 6.34 5.80 5.81 6.08 6.03 6.01 6.01 5.77 6.09
simdjson Throughput Test: LargeRandom OpenBenchmarking.org GB/s, More Is Better simdjson 3.10 Throughput Test: LargeRandom 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.414 0.828 1.242 1.656 2.07 1.70 1.71 1.81 1.81 1.73 1.84 1.83 1.82 1.83 1.83 1.84 1. (CXX) g++ options: -O3 -lrt
ACES DGEMM Sustained Floating-Point Rate OpenBenchmarking.org GFLOP/s, More Is Better ACES DGEMM 1.0 Sustained Floating-Point Rate 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 200 400 600 800 1000 259.15 263.20 629.17 818.68 812.15 842.73 1092.09 1091.01 1093.10 1141.19 842.01 1. (CC) gcc options: -ffast-math -mavx2 -O3 -fopenmp -lopenblas
Timed Eigen Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed Eigen Compilation 3.4.0 Time To Compile 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 30 60 90 120 150 145.21 146.66 84.48 69.38 69.58 67.36 57.18 58.82 58.69 58.66 67.08
Apache CouchDB Bulk Size: 100 - Inserts: 1000 - Rounds: 30 OpenBenchmarking.org Seconds, Fewer Is Better Apache CouchDB 3.4.1 Bulk Size: 100 - Inserts: 1000 - Rounds: 30 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20 40 60 80 100 96.74 96.55 73.24 76.70 76.58 75.90 76.66 75.89 76.67 69.93 76.39 1. (CXX) g++ options: -flto -lstdc++ -shared -lei
oneDNN Harness: Recurrent Neural Network Training - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.6 Harness: Recurrent Neural Network Training - Engine: CPU 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 1000 2000 3000 4000 5000 4726.08 4752.85 2422.83 1911.44 1913.86 1898.36 1320.23 1320.99 1320.37 1372.03 1895.68 MIN: 4716.28 MIN: 4746.78 MIN: 2416.25 MIN: 1895 MIN: 1899.63 MIN: 1894.26 MIN: 1302.81 MIN: 1308.56 MIN: 1301.96 MIN: 1342.06 MIN: 1892.59 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl
Renaissance Test: ALS Movie Lens OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.16 Test: ALS Movie Lens 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2K 4K 6K 8K 10K 7465.1 7147.4 5993.4 9243.0 8980.8 9378.8 9972.0 9891.4 9969.8 9805.7 9275.7 MIN: 6716.72 / MAX: 7815.35 MIN: 6653.82 / MAX: 7863.42 MIN: 5836.57 / MAX: 6535.35 MIN: 8920.42 / MAX: 9406.46 MIN: 8480.95 / MAX: 9113.57 MIN: 8718.36 / MAX: 9413.7 MIN: 9479.38 / MAX: 10040.33 MIN: 9364.27 / MAX: 10037.94 MIN: 9680.91 / MAX: 9983.16 MIN: 9253.4 / MAX: 10057.61 MIN: 8821.09 / MAX: 9495.91
oneDNN Harness: Recurrent Neural Network Inference - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.6 Harness: Recurrent Neural Network Inference - Engine: CPU 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 500 1000 1500 2000 2500 2415.25 2441.27 1270.76 1009.58 1012.19 965.02 670.70 672.51 670.23 700.86 966.01 MIN: 2406.21 MIN: 2435.83 MIN: 1266.25 MIN: 994.85 MIN: 999.17 MIN: 963.27 MIN: 662.66 MIN: 665.03 MIN: 663.39 MIN: 679.89 MIN: 963.43 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl
simdjson Throughput Test: DistinctUserID OpenBenchmarking.org GB/s, More Is Better simdjson 3.10 Throughput Test: DistinctUserID 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 9.92 10.16 10.85 10.54 10.28 10.76 9.15 11.05 11.14 10.46 8.97 1. (CXX) g++ options: -O3 -lrt
Stockfish Chess Benchmark OpenBenchmarking.org Nodes Per Second, More Is Better Stockfish Chess Benchmark 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 10M 20M 30M 40M 50M 12024682 11607502 24017024 30315972 29940074 33702298 48797079 44446458 45724794 46507038 33871595 1. Stockfish 16 by the Stockfish developers (see AUTHORS file)
simdjson Throughput Test: TopTweet OpenBenchmarking.org GB/s, More Is Better simdjson 3.10 Throughput Test: TopTweet 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 10.02 9.97 10.86 10.66 10.69 10.82 9.40 11.25 9.19 10.46 10.51 1. (CXX) g++ options: -O3 -lrt
simdjson Throughput Test: PartialTweets OpenBenchmarking.org GB/s, More Is Better simdjson 3.10 Throughput Test: PartialTweets 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 9.32 9.30 10.04 9.80 9.63 10.10 10.32 9.85 9.98 9.76 8.35 1. (CXX) g++ options: -O3 -lrt
Llamafile Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Text Generation 128 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Text Generation 128 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 7 14 21 28 35 25.09 25.52 27.36 27.35 27.59 27.21 27.20 27.19 26.28 27.80
Renaissance Test: In-Memory Database Shootout OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.16 Test: In-Memory Database Shootout 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 800 1600 2400 3200 4000 2951.1 3681.9 3244.8 3288.5 3275.5 3241.5 3408.4 3232.8 3442.9 3256.1 3175.6 MIN: 2787.28 / MAX: 3070.94 MIN: 2559.5 MIN: 2350.89 MIN: 2991.78 / MAX: 3586.63 MIN: 3012.8 / MAX: 3533.02 MIN: 3037.03 / MAX: 3491.91 MIN: 3187.55 / MAX: 3638.98 MIN: 3057.86 / MAX: 3585.74 MIN: 3258.62 / MAX: 3709.4 MIN: 3019.89 / MAX: 3599.5 MIN: 2896.06 / MAX: 3367.44
Renaissance Test: Akka Unbalanced Cobwebbed Tree OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.16 Test: Akka Unbalanced Cobwebbed Tree 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 1000 2000 3000 4000 5000 4474.8 4457.1 3720.1 4084.6 4098.5 4038.4 4359.3 4384.0 4344.8 4403.8 4002.3 MIN: 4474.77 / MAX: 5751.45 MIN: 4457.08 / MAX: 5796.57 MIN: 3720.09 / MAX: 4686.78 MAX: 5256.95 MIN: 4098.48 / MAX: 5163.21 MIN: 4038.36 / MAX: 5089.28 MIN: 4359.25 / MAX: 5618.71 MIN: 4383.98 / MAX: 5691.67 MAX: 5622.35 MAX: 5719.11 MIN: 4002.27 / MAX: 4983.72
Renaissance Test: Apache Spark PageRank OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.16 Test: Apache Spark PageRank 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 500 1000 1500 2000 2500 2267.1 2292.2 2506.8 2366.1 2373.0 2138.1 2179.0 2182.5 2227.9 2412.2 2229.7 MIN: 2117.39 / MAX: 2335.98 MIN: 2106.39 / MAX: 2374.31 MIN: 1771.47 MIN: 1667.92 / MAX: 2366.13 MIN: 1684.52 MIN: 1499.64 MIN: 1591.55 / MAX: 2179.02 MIN: 1564.17 MIN: 1592.13 / MAX: 2227.91 MIN: 1691.04 MIN: 1612.96 / MAX: 2229.74
Llamafile Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Prompt Processing 512 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Prompt Processing 512 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2K 4K 6K 8K 10K 8192 8192 8192 8192 8192 8192 8192 8192 8192 8192
CP2K Molecular Dynamics Input: H20-64 OpenBenchmarking.org Seconds, Fewer Is Better CP2K Molecular Dynamics 2024.3 Input: H20-64 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 30 60 90 120 150 123.66 130.81 68.17 58.40 58.86 53.01 42.47 43.17 42.51 58.19 52.72 1. (F9X) gfortran options: -fopenmp -march=native -mtune=native -O3 -funroll-loops -fbacktrace -ffree-form -fimplicit-none -std=f2008 -lcp2kstart -lcp2kmc -lcp2kswarm -lcp2kmotion -lcp2kthermostat -lcp2kemd -lcp2ktmc -lcp2kmain -lcp2kdbt -lcp2ktas -lcp2kgrid -lcp2kgriddgemm -lcp2kgridcpu -lcp2kgridref -lcp2kgridcommon -ldbcsrarnoldi -ldbcsrx -lcp2kdbx -lcp2kdbm -lcp2kshg_int -lcp2keri_mme -lcp2kminimax -lcp2khfxbase -lcp2ksubsys -lcp2kxc -lcp2kao -lcp2kpw_env -lcp2kinput -lcp2kpw -lcp2kgpu -lcp2kfft -lcp2kfpga -lcp2kfm -lcp2kcommon -lcp2koffload -lcp2kmpiwrap -lcp2kbase -ldbcsr -lsirius -lspla -lspfft -lsymspg -lvdwxc -l:libhdf5_fortran.a -l:libhdf5.a -lz -lgsl -lelpa_openmp -lcosma -lcosta -lscalapack -lxsmmf -lxsmm -ldl -lpthread -llibgrpp -lxcf03 -lxc -lint2 -lfftw3_mpi -lfftw3 -lfftw3_omp -lmpi_cxx -lmpi -l:libopenblas.a -lvori -lstdc++ -lmpi_usempif08 -lmpi_mpifh -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm
Renaissance Test: Savina Reactors.IO OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.16 Test: Savina Reactors.IO 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 800 1600 2400 3200 4000 3745.8 3818.5 3349.5 3523.1 3688.6 3655.8 3589.4 3582.4 3641.9 3506.4 3676.0 MIN: 3745.79 / MAX: 4547.1 MAX: 4702.95 MIN: 3349.49 / MAX: 4130.17 MIN: 3523.09 / MAX: 4370.41 MAX: 4840.82 MIN: 3655.76 / MAX: 4484.97 MAX: 4472.27 MIN: 3582.35 / MAX: 4689.63 MAX: 4585.77 MIN: 3506.38 / MAX: 4329.37 MAX: 4536.84
Renaissance Test: Gaussian Mixture Model OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.16 Test: Gaussian Mixture Model 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 800 1600 2400 3200 4000 3087.0 3130.4 3397.0 3647.2 3703.2 3860.6 3829.9 3806.5 3773.1 3399.5 3815.2 MIN: 2935.02 / MAX: 3439.02 MIN: 2997.45 / MAX: 3481.57 MIN: 2497.54 / MAX: 3397.03 MIN: 2576.86 / MAX: 3647.22 MIN: 2648.51 / MAX: 3703.23 MIN: 2758.89 / MAX: 3860.61 MIN: 2792.29 MIN: 2770.53 / MAX: 3806.52 MIN: 2755.26 / MAX: 3773.12 MIN: 2471.52 MIN: 2749.56 / MAX: 3815.24
PyPerformance Benchmark: gc_collect OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: gc_collect 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 150 300 450 600 750 702 704 649 671 671 699 689 682 686 677 706
Renaissance Test: Apache Spark Bayes OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.16 Test: Apache Spark Bayes 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 130 260 390 520 650 566.2 600.1 483.4 515.3 513.3 513.2 500.0 477.0 515.6 490.0 474.9 MIN: 493.35 / MAX: 1029.04 MIN: 506.71 / MAX: 741.31 MIN: 456.88 / MAX: 535.94 MIN: 456.38 / MAX: 531.31 MIN: 455.04 / MAX: 535.53 MIN: 453.66 / MAX: 554.7 MIN: 457.03 / MAX: 581.14 MIN: 460.78 / MAX: 515.59 MIN: 459.35 / MAX: 536.06 MIN: 459.29 / MAX: 580.9 MIN: 454.77 / MAX: 514.32
Renaissance Test: Finagle HTTP Requests OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.16 Test: Finagle HTTP Requests 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 600 1200 1800 2400 3000 1664.4 1669.9 1705.5 2650.9 2708.1 2492.2 2570.5 2627.3 2571.2 2319.4 2483.1 MIN: 1625.17 / MAX: 1776.68 MIN: 1612.75 / MAX: 1713.57 MIN: 1676.49 / MAX: 1730.57 MIN: 2066.79 MIN: 2074.84 / MAX: 2708.11 MIN: 1947.63 MIN: 1960.33 / MAX: 2570.51 MIN: 2034.59 / MAX: 2627.31 MIN: 1999.64 MIN: 1832.84 MIN: 1933.43
Renaissance Test: Random Forest OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.16 Test: Random Forest 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 110 220 330 440 550 439.9 485.4 381.6 415.3 412.9 422.0 407.7 420.7 402.2 414.4 453.2 MIN: 388.86 / MAX: 516.9 MIN: 403.24 / MAX: 525.59 MIN: 332.96 / MAX: 445.77 MIN: 342.21 / MAX: 486.26 MIN: 342.95 / MAX: 473.15 MIN: 357.91 / MAX: 497.55 MIN: 337.66 / MAX: 471.92 MIN: 339.2 / MAX: 478.26 MIN: 343.06 / MAX: 486.63 MIN: 322.79 / MAX: 466.1 MIN: 352.31 / MAX: 513.31
OSPRay Benchmark: gravity_spheres_volume/dim_512/scivis/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPRay 3.2 Benchmark: gravity_spheres_volume/dim_512/scivis/real_time 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2 4 6 8 10 1.90804 1.89469 3.83152 5.35756 5.36189 5.54888 7.31594 7.32353 7.32284 7.58789 5.61470
OSPRay Benchmark: gravity_spheres_volume/dim_512/ao/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPRay 3.2 Benchmark: gravity_spheres_volume/dim_512/ao/real_time 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2 4 6 8 10 1.84407 1.90481 3.91749 5.47043 5.47057 5.63122 7.46399 7.38760 7.44575 7.63944 5.71084
ONNX Runtime Model: ResNet101_DUC_HDC-12 - Device: CPU - Executor: Standard OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.19 Model: ResNet101_DUC_HDC-12 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 500 1000 1500 2000 2500 2151.82 2152.92 1091.32 872.74 864.81 850.14 645.21 650.96 647.81 648.52 854.33 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: ResNet101_DUC_HDC-12 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.19 Model: ResNet101_DUC_HDC-12 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.3487 0.6974 1.0461 1.3948 1.7435 0.464723 0.464485 0.916317 1.145810 1.156320 1.176270 1.549880 1.536190 1.543650 1.541960 1.170500 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
Renaissance Test: Scala Dotty OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.16 Test: Scala Dotty 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 110 220 330 440 550 506.5 515.8 464.5 424.6 442.9 428.6 455.6 441.9 411.9 477.0 436.2 MIN: 420.76 / MAX: 1038.72 MIN: 426.08 / MAX: 1026.55 MIN: 358.07 / MAX: 790.35 MIN: 376.26 / MAX: 659.55 MIN: 376.04 / MAX: 787.17 MIN: 378.22 / MAX: 628.77 MIN: 394.05 / MAX: 695.67 MIN: 387.7 / MAX: 646.01 MIN: 362.36 / MAX: 713.71 MIN: 371.54 / MAX: 736.5 MIN: 380.62 / MAX: 721.56
Renaissance Test: Genetic Algorithm Using Jenetics + Futures OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.16 Test: Genetic Algorithm Using Jenetics + Futures 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 200 400 600 800 1000 806.4 801.4 749.9 924.5 884.0 904.0 891.2 1050.3 959.3 732.8 920.7 MIN: 786.25 / MAX: 832.34 MIN: 786.46 / MAX: 836.72 MIN: 737.7 / MAX: 777.92 MIN: 821.03 MIN: 863.46 / MAX: 897.46 MIN: 886.83 / MAX: 919.31 MIN: 861.29 / MAX: 903.79 MIN: 1016.46 / MAX: 1068.02 MIN: 844.95 MIN: 713.67 / MAX: 813.49 MIN: 888.75 / MAX: 934.44
ONNX Runtime Model: fcn-resnet101-11 - Device: CPU - Executor: Standard OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.19 Model: fcn-resnet101-11 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 200 400 600 800 1000 926.76 927.39 487.30 386.56 386.06 355.75 293.96 295.28 294.93 310.88 357.60 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: fcn-resnet101-11 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.19 Model: fcn-resnet101-11 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.7654 1.5308 2.2962 3.0616 3.827 1.07903 1.07829 2.05211 2.58688 2.59025 2.81093 3.40178 3.38658 3.39064 3.21670 2.79638 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: GPT-2 - Device: CPU - Executor: Standard OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.19 Model: GPT-2 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2 4 6 8 10 8.04835 8.10767 7.96069 7.33331 7.36544 6.25815 6.01631 6.02069 6.03024 7.42776 6.33034 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: GPT-2 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.19 Model: GPT-2 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 40 80 120 160 200 124.17 123.27 125.52 136.30 135.72 159.71 166.17 166.05 165.77 134.60 157.89 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OSPRay Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPRay 3.2 Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_time 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2 4 6 8 10 2.23851 2.22318 4.52318 6.17399 6.17624 6.41198 8.49646 8.43752 8.47727 8.82093 6.40740
ONNX Runtime Model: ZFNet-512 - Device: CPU - Executor: Standard OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.19 Model: ZFNet-512 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 4 8 12 16 20 13.53230 13.69550 10.28350 9.85896 9.83893 9.01322 8.83524 8.87444 8.82581 9.76985 9.01687 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: ZFNet-512 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.19 Model: ZFNet-512 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 30 60 90 120 150 73.88 73.00 97.23 101.41 101.62 110.94 113.17 112.67 113.29 102.33 110.89 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: bertsquad-12 - Device: CPU - Executor: Standard OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.19 Model: bertsquad-12 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20 40 60 80 100 95.07 94.12 50.47 68.08 68.09 68.91 67.18 66.62 66.92 64.14 68.61 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: bertsquad-12 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.19 Model: bertsquad-12 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 5 10 15 20 25 10.52 10.62 19.81 14.69 14.69 14.51 14.88 15.01 14.94 15.59 14.57 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: T5 Encoder - Device: CPU - Executor: Standard OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.19 Model: T5 Encoder - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2 4 6 8 10 7.04122 7.13857 7.06075 5.95035 6.00470 4.80287 4.90313 4.90285 4.91381 6.39112 4.85142 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: T5 Encoder - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.19 Model: T5 Encoder - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 50 100 150 200 250 141.96 140.03 141.57 168.04 166.51 208.17 203.93 203.95 203.49 156.45 206.09 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: yolov4 - Device: CPU - Executor: Standard OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.19 Model: yolov4 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 30 60 90 120 150 139.66 139.72 75.54 95.25 95.36 93.16 88.41 88.06 88.36 90.45 93.34 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: yolov4 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.19 Model: yolov4 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 7.16015 7.15712 13.23720 10.49820 10.48650 10.73380 11.31050 11.35510 11.31660 11.05520 10.71270 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.19 Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 10 20 30 40 50 41.94 41.82 25.58 27.68 27.47 26.75 22.99 22.81 22.70 23.55 26.95 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.19 Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 10 20 30 40 50 23.84 23.91 39.08 36.12 36.40 37.38 43.49 43.84 44.05 42.45 37.10 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: Standard OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.19 Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 7 14 21 28 35 27.69 25.95 22.32 21.57 21.69 24.94 20.98 20.72 20.88 21.24 23.06 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.19 Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 11 22 33 44 55 36.11 38.53 44.80 46.35 46.09 40.09 47.67 48.26 47.89 47.07 43.36 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.19 Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.4543 0.9086 1.3629 1.8172 2.2715 2.016090 2.019130 1.311210 1.379700 1.335160 1.061880 0.986643 0.994193 0.999701 1.570840 1.066000 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.19 Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 200 400 600 800 1000 495.84 495.08 762.25 724.53 748.66 941.40 1013.19 1005.50 999.97 636.32 937.78 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standard OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.19 Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 1.1129 2.2258 3.3387 4.4516 5.5645 4.81530 4.94636 2.65482 2.77422 2.76714 2.80544 2.62573 2.60974 2.61221 2.55898 2.80695 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.19 Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 80 160 240 320 400 207.61 202.12 376.53 360.35 361.30 356.41 380.80 383.13 382.77 390.60 356.19 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: super-resolution-10 - Device: CPU - Executor: Standard OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.19 Model: super-resolution-10 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 4 8 12 16 20 13.64570 13.54170 6.36564 7.41037 7.43372 7.98873 8.18175 8.18519 8.18265 7.08601 7.99486 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ONNX Runtime Model: super-resolution-10 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.19 Model: super-resolution-10 - Device: CPU - Executor: Standard 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 30 60 90 120 150 73.27 73.84 157.08 134.94 134.52 125.17 122.22 122.17 122.21 141.12 125.08 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
Llama.cpp Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 2048 OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4154 Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 2048 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 60 120 180 240 300 105.10 171.68 213.02 212.30 222.75 273.67 289.41 289.04 279.04 208.99 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas
PyPerformance Benchmark: asyncio_websockets OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: asyncio_websockets 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 70 140 210 280 350 330 331 308 315 316 321 314 315 317 315 322
Llamafile Model: Llama-3.2-3B-Instruct.Q6_K - Test: Prompt Processing 1024 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: Llama-3.2-3B-Instruct.Q6_K - Test: Prompt Processing 1024 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 4K 8K 12K 16K 20K 16384 16384 16384 16384 16384 16384 16384 16384 16384 16384
Llamafile Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Prompt Processing 256 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Prompt Processing 256 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 300 600 900 1200 1500 1536 1536 1536 1536 1536 1536 1536 1536 1536 1536
Llamafile Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Prompt Processing 2048 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Prompt Processing 2048 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 7K 14K 21K 28K 35K 32768 32768 32768 32768 32768 32768 32768 32768 32768 32768
SVT-AV1 Encoder Mode: Preset 5 - Input: Bosphorus 1080p OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 5 - Input: Bosphorus 1080p 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20 40 60 80 100 37.17 37.07 65.32 86.96 86.39 88.42 98.91 98.79 99.45 101.97 88.27 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
LiteRT Model: Inception V4 OpenBenchmarking.org Microseconds, Fewer Is Better LiteRT 2024-10-15 Model: Inception V4 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 11K 22K 33K 44K 55K 49048.8 49101.9 24502.1 23141.4 22928.3 22083.3 19686.0 20102.8 19859.4 21477.8 22752.4
LiteRT Model: Inception ResNet V2 OpenBenchmarking.org Microseconds, Fewer Is Better LiteRT 2024-10-15 Model: Inception ResNet V2 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 9K 18K 27K 36K 45K 41844.0 41757.9 21153.5 20123.5 20503.5 19477.8 17824.6 18224.0 18412.3 19530.2 19490.7
LiteRT Model: NASNet Mobile OpenBenchmarking.org Microseconds, Fewer Is Better LiteRT 2024-10-15 Model: NASNet Mobile 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 4K 8K 12K 16K 20K 5153.49 5034.38 3808.80 7759.69 7640.81 8057.56 9245.76 9311.77 9231.97 16936.00 7931.64
LiteRT Model: DeepLab V3 OpenBenchmarking.org Microseconds, Fewer Is Better LiteRT 2024-10-15 Model: DeepLab V3 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 800 1600 2400 3200 4000 2950.80 2994.59 1569.75 2272.28 2289.58 2343.38 2152.48 2185.78 2200.93 3579.67 2359.99
LiteRT Model: Mobilenet Float OpenBenchmarking.org Microseconds, Fewer Is Better LiteRT 2024-10-15 Model: Mobilenet Float 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 500 1000 1500 2000 2500 2193.88 2190.79 1044.53 1248.12 1248.60 1244.70 1178.94 1165.50 1175.50 1211.48 1244.51
LiteRT Model: SqueezeNet OpenBenchmarking.org Microseconds, Fewer Is Better LiteRT 2024-10-15 Model: SqueezeNet 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 700 1400 2100 2800 3500 3173.55 3179.55 1654.65 1820.34 1822.70 1809.18 1763.22 1769.56 1765.39 1794.11 1821.35
LiteRT Model: Quantized COCO SSD MobileNet v1 OpenBenchmarking.org Microseconds, Fewer Is Better LiteRT 2024-10-15 Model: Quantized COCO SSD MobileNet v1 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 600 1200 1800 2400 3000 2736.98 2741.68 1494.51 1431.19 1415.11 1420.15 1343.22 1337.78 1339.69 2129.52 1417.35
LiteRT Model: Mobilenet Quant OpenBenchmarking.org Microseconds, Fewer Is Better LiteRT 2024-10-15 Model: Mobilenet Quant 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 400 800 1200 1600 2000 1781.60 1785.40 935.99 859.76 857.49 848.94 730.17 734.35 742.03 823.17 849.21
Rustls Benchmark: handshake-ticket - Suite: TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 OpenBenchmarking.org handshakes/s, More Is Better Rustls 0.23.17 Benchmark: handshake-ticket - Suite: TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 600K 1200K 1800K 2400K 3000K 1656876.39 1571309.62 2353677.96 2283083.54 2304145.79 2282729.64 2559245.20 2528614.52 2558621.38 2620332.00 2292879.44 1. (CC) gcc options: -m64 -lgcc_s -lutil -lrt -lpthread -lm -ldl -lc -pie -nodefaultlibs
Llama.cpp Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512 OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4154 Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20 40 60 80 100 32.97 60.87 64.96 64.43 69.11 91.71 92.89 93.91 70.76 67.95 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas
Llama.cpp Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 512 OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4154 Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 512 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20 40 60 80 100 32.94 61.09 65.24 64.68 68.20 94.08 93.24 93.49 68.40 68.81 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas
Rustls Benchmark: handshake-resume - Suite: TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 OpenBenchmarking.org handshakes/s, More Is Better Rustls 0.23.17 Benchmark: handshake-resume - Suite: TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 800K 1600K 2400K 3200K 4000K 2105346.65 2103587.50 2874011.56 3061522.26 3082657.50 3035330.21 3379343.72 3397871.03 3431395.73 3563852.57 3038723.48 1. (CC) gcc options: -m64 -lgcc_s -lutil -lrt -lpthread -lm -ldl -lc -pie -nodefaultlibs
ASTC Encoder Preset: Thorough OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 5.0 Preset: Thorough 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 5 10 15 20 25 5.0655 5.0448 10.3392 13.5230 13.5139 14.1700 18.7661 18.7075 18.7499 20.3025 14.1464 1. (CXX) g++ options: -O3 -flto -pthread
Llamafile Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Text Generation 16 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Text Generation 16 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.4185 0.837 1.2555 1.674 2.0925 1.73 1.76 1.83 1.83 1.83 1.86 1.82 1.83 1.78 1.84
FinanceBench Benchmark: Bonds OpenMP OpenBenchmarking.org ms, Fewer Is Better FinanceBench 2016-07-25 Benchmark: Bonds OpenMP 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 10K 20K 30K 40K 50K 47198.94 46851.42 32715.81 33551.52 33538.27 34600.77 33906.69 34053.06 33913.46 33061.22 34896.84 1. (CXX) g++ options: -O3 -march=native -fopenmp
Etcpak Benchmark: Multi-Threaded - Configuration: ETC2 OpenBenchmarking.org Mpx/s, More Is Better Etcpak 2.0 Benchmark: Multi-Threaded - Configuration: ETC2 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 120 240 360 480 600 155.44 154.25 304.24 388.40 388.13 410.73 556.87 556.04 556.41 577.82 409.88 1. (CXX) g++ options: -flto -pthread
Whisperfile Model Size: Tiny OpenBenchmarking.org Seconds, Fewer Is Better Whisperfile 20Aug24 Model Size: Tiny 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 15 30 45 60 75 66.11 41.30 40.85 38.58 37.13 35.11 35.91 35.43 41.71 38.72
Rustls Benchmark: handshake - Suite: TLS13_CHACHA20_POLY1305_SHA256 OpenBenchmarking.org handshakes/s, More Is Better Rustls 0.23.17 Benchmark: handshake - Suite: TLS13_CHACHA20_POLY1305_SHA256 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 16K 32K 48K 64K 80K 21499.73 21444.09 43571.86 56047.38 56176.90 57716.64 72140.75 72793.26 72508.13 76454.45 57688.08 1. (CC) gcc options: -m64 -lgcc_s -lutil -lrt -lpthread -lm -ldl -lc -pie -nodefaultlibs
SVT-AV1 Encoder Mode: Preset 8 - Input: Bosphorus 4K OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 8 - Input: Bosphorus 4K 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20 40 60 80 100 35.85 35.64 65.19 82.53 82.16 85.20 101.52 101.33 101.54 102.01 85.00 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
Rustls Benchmark: handshake - Suite: TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 OpenBenchmarking.org handshakes/s, More Is Better Rustls 0.23.17 Benchmark: handshake - Suite: TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20K 40K 60K 80K 100K 22255.23 22201.03 45095.35 57954.19 58056.01 59308.75 74672.57 74468.49 74683.74 80462.60 59206.34 1. (CC) gcc options: -m64 -lgcc_s -lutil -lrt -lpthread -lm -ldl -lc -pie -nodefaultlibs
PyPerformance Benchmark: django_template OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: django_template 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 5 10 15 20 25 21.2 21.3 20.1 20.6 20.6 21.0 20.9 20.8 21.0 20.7 21.2
OpenVINO GenAI Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time Per Output Token OpenBenchmarking.org ms, Fewer Is Better OpenVINO GenAI 2024.5 Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time Per Output Token 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 13 26 39 52 65 60.07 52.64 50.45 50.46 49.31 50.74 50.84 50.68 51.86 49.28
OpenVINO GenAI Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time To First Token OpenBenchmarking.org ms, Fewer Is Better OpenVINO GenAI 2024.5 Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time To First Token 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 30 60 90 120 150 130.47 71.58 59.87 59.86 58.91 55.67 56.06 55.47 55.93 58.86
OpenVINO GenAI Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU OpenBenchmarking.org tokens/s, More Is Better OpenVINO GenAI 2024.5 Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 5 10 15 20 25 16.65 19.00 19.82 19.82 20.28 19.71 19.67 19.73 19.28 20.29
POV-Ray Trace Time OpenBenchmarking.org Seconds, Fewer Is Better POV-Ray Trace Time 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 16 32 48 64 80 70.06 70.77 34.92 26.61 26.60 25.26 19.52 19.55 20.33 18.54 25.33 1. POV-Ray 3.7.0.10.unofficial
NAMD Input: ATPase with 327,506 Atoms OpenBenchmarking.org ns/day, More Is Better NAMD 3.0 Input: ATPase with 327,506 Atoms 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.6423 1.2846 1.9269 2.5692 3.2115 0.88895 0.89241 1.74038 2.21645 2.20728 2.38124 2.83798 2.83913 2.85485 2.79632 2.35379
Llamafile Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Prompt Processing 256 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Prompt Processing 256 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 900 1800 2700 3600 4500 4096 4096 4096 4096 4096 4096 4096 4096 4096 4096
PyPerformance Benchmark: raytrace OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: raytrace 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 40 80 120 160 200 182 180 174 177 176 182 179 176 176 175 182
FinanceBench Benchmark: Repo OpenMP OpenBenchmarking.org ms, Fewer Is Better FinanceBench 2016-07-25 Benchmark: Repo OpenMP 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 7K 14K 21K 28K 35K 32271.67 32477.35 21121.61 21588.91 21612.04 22320.33 22060.15 21948.94 21848.07 21418.45 22318.74 1. (CXX) g++ options: -O3 -march=native -fopenmp
PyPerformance Benchmark: crypto_pyaes OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: crypto_pyaes 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 10 20 30 40 50 43.0 43.1 40.7 41.8 42.0 43.1 42.4 42.7 42.7 41.7 43.3
PyPerformance Benchmark: chaos OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: chaos 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 9 18 27 36 45 39.9 40.2 38.3 39.0 38.6 39.7 38.8 39.2 39.1 38.2 39.4
PyPerformance Benchmark: go OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: go 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 20 40 60 80 100 78.4 78.7 75.3 78.3 76.0 78.6 78.1 78.4 79.6 77.8 79.4
PyPerformance Benchmark: float OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: float 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 12 24 36 48 60 52.8 52.9 48.2 49.7 49.7 51.3 50.7 51.1 51.0 50.7 50.8
x265 Video Input: Bosphorus 4K OpenBenchmarking.org Frames Per Second, More Is Better x265 Video Input: Bosphorus 4K 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 8 16 24 32 40 10.91 10.89 18.85 26.09 25.85 27.16 32.23 32.59 32.22 32.57 26.94 1. x265 [info]: HEVC encoder version 3.5+1-f0c1022b6
PyPerformance Benchmark: regex_compile OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: regex_compile 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 16 32 48 64 80 72.6 72.7 68.1 69.5 70.0 71.7 70.4 71.4 69.8 69.8 72.5
Llamafile Model: Llama-3.2-3B-Instruct.Q6_K - Test: Prompt Processing 512 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: Llama-3.2-3B-Instruct.Q6_K - Test: Prompt Processing 512 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2K 4K 6K 8K 10K 8192 8192 8192 8192 8192 8192 8192 8192 8192 8192
Llama.cpp Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 1024 OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4154 Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 1024 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 80 160 240 320 400 115.67 185.97 235.01 227.71 232.26 291.02 290.10 324.04 355.09 244.77 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas
Y-Cruncher Pi Digits To Calculate: 1B OpenBenchmarking.org Seconds, Fewer Is Better Y-Cruncher 0.8.5 Pi Digits To Calculate: 1B 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 9 18 27 36 45 40.56 39.25 23.89 19.36 19.28 18.38 17.36 17.34 17.57 18.49 18.37
PyPerformance Benchmark: pathlib OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: pathlib 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 4 8 12 16 20 14.6 14.6 13.5 14.1 14.1 14.4 14.2 14.1 14.1 14.2 14.4
PyPerformance Benchmark: pickle_pure_python OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: pickle_pure_python 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 40 80 120 160 200 171 171 162 164 164 169 163 166 165 165 168
oneDNN Harness: Deconvolution Batch shapes_1d - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.6 Harness: Deconvolution Batch shapes_1d - Engine: CPU 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2 4 6 8 10 7.26738 7.28139 3.78629 3.43287 3.42547 3.40293 3.05918 3.07219 3.06341 2.97612 3.40628 MIN: 6.92 MIN: 6.94 MIN: 3.55 MIN: 2.96 MIN: 2.9 MIN: 3.03 MIN: 2.56 MIN: 2.59 MIN: 2.58 MIN: 2.42 MIN: 3.03 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl
Llamafile Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Text Generation 16 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Text Generation 16 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 9.99 9.92 10.48 10.45 10.45 10.34 10.30 10.37 10.22 10.45
Llamafile Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Prompt Processing 1024 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Prompt Processing 1024 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 4K 8K 12K 16K 20K 16384 16384 16384 16384 16384 16384 16384 16384 16384 16384
PyPerformance Benchmark: json_loads OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: json_loads 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 12.4 12.4 11.8 12.0 12.0 12.4 12.3 12.5 12.2 12.1 12.5
ASTC Encoder Preset: Fast OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 5.0 Preset: Fast 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 90 180 270 360 450 100.37 99.40 205.01 263.55 263.37 278.24 369.03 368.55 368.57 396.65 277.30 1. (CXX) g++ options: -O3 -flto -pthread
PyPerformance Benchmark: nbody OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: nbody 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 14 28 42 56 70 60.4 60.0 56.8 57.2 58.6 59.5 59.6 59.0 58.8 59.0 59.2
7-Zip Compression Test: Decompression Rating OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression Test: Decompression Rating 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 40K 80K 120K 160K 200K 43781 43210 88229 118569 118587 125698 166180 164581 165934 165916 125605 1. 7-Zip 23.01 (x64) : Copyright (c) 1999-2023 Igor Pavlov : 2023-06-20
7-Zip Compression Test: Compression Rating OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression Test: Compression Rating 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 40K 80K 120K 160K 200K 54079 53425 101872 135033 133804 141263 169373 170361 169750 163859 142213 1. 7-Zip 23.01 (x64) : Copyright (c) 1999-2023 Igor Pavlov : 2023-06-20
SVT-AV1 Encoder Mode: Preset 13 - Input: Bosphorus 4K OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 13 - Input: Bosphorus 4K 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 50 100 150 200 250 85.95 85.60 156.60 192.17 190.34 198.11 217.96 217.91 218.53 212.52 194.02 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
SVT-AV1 Encoder Mode: Preset 8 - Input: Bosphorus 1080p OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 8 - Input: Bosphorus 1080p 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 70 140 210 280 350 123.79 122.92 215.21 282.01 281.89 287.05 325.95 320.18 327.26 339.02 286.96 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
oneDNN Harness: IP Shapes 1D - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.6 Harness: IP Shapes 1D - Engine: CPU 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 0.862 1.724 2.586 3.448 4.31 3.83117 3.82885 1.90686 1.88731 1.88919 1.93806 1.13012 1.12465 1.12801 1.12573 1.93913 MIN: 3.77 MIN: 3.77 MIN: 1.86 MIN: 1.83 MIN: 1.83 MIN: 1.92 MIN: 1.1 MIN: 1.09 MIN: 1.1 MIN: 1.03 MIN: 1.91 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl
ASTC Encoder Preset: Medium OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 5.0 Preset: Medium 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 30 60 90 120 150 39.32 39.04 80.37 103.81 103.67 109.03 144.61 144.48 144.63 156.22 108.86 1. (CXX) g++ options: -O3 -flto -pthread
Llama.cpp Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Text Generation 128 OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4154 Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Text Generation 128 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 12 24 36 48 60 51.50 51.86 50.50 50.43 52.30 50.33 49.90 49.76 47.72 52.37 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas
Llamafile Model: Llama-3.2-3B-Instruct.Q6_K - Test: Prompt Processing 256 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: Llama-3.2-3B-Instruct.Q6_K - Test: Prompt Processing 256 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 900 1800 2700 3600 4500 4096 4096 4096 4096 4096 4096 4096 4096 4096 4096
Llama.cpp Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 512 OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4154 Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 512 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 70 140 210 280 350 122.80 204.06 235.94 242.10 243.14 337.06 316.80 326.16 327.30 232.86 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas
Llamafile Model: Llama-3.2-3B-Instruct.Q6_K - Test: Text Generation 16 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: Llama-3.2-3B-Instruct.Q6_K - Test: Text Generation 16 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 5 10 15 20 25 19.17 18.83 19.47 19.53 19.49 19.16 19.27 19.24 19.03 19.50
Y-Cruncher Pi Digits To Calculate: 500M OpenBenchmarking.org Seconds, Fewer Is Better Y-Cruncher 0.8.5 Pi Digits To Calculate: 500M 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 5 10 15 20 25 18.165 18.290 10.954 9.041 9.067 8.688 8.119 8.173 8.279 8.772 8.623
Primesieve Length: 1e12 OpenBenchmarking.org Seconds, Fewer Is Better Primesieve 12.6 Length: 1e12 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 6 12 18 24 30 24.723 24.821 12.273 9.701 9.705 9.116 6.937 6.926 6.903 6.347 9.147 1. (CXX) g++ options: -O3
Llamafile Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Prompt Processing 512 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Prompt Processing 512 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2K 4K 6K 8K 10K 8192 8192 8192 8192 8192 8192 8192 8192 8192 8192
Llamafile Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Text Generation 16 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Text Generation 16 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 6 12 18 24 30 23.80 24.27 25.77 25.78 25.86 25.38 25.32 25.30 24.59 25.94
oneDNN Harness: IP Shapes 3D - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.6 Harness: IP Shapes 3D - Engine: CPU 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 1.1157 2.2314 3.3471 4.4628 5.5785 4.94689 4.95867 4.45096 3.75003 3.74578 2.73072 2.77835 2.74699 2.77727 4.05800 2.72942 MIN: 4.85 MIN: 4.87 MIN: 4.38 MIN: 3.71 MIN: 3.71 MIN: 2.7 MIN: 2.74 MIN: 2.72 MIN: 2.75 MIN: 3.75 MIN: 2.7 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl
x265 Video Input: Bosphorus 1080p OpenBenchmarking.org Frames Per Second, More Is Better x265 Video Input: Bosphorus 1080p 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 30 60 90 120 150 56.61 56.03 88.71 101.23 101.33 101.37 113.31 113.37 113.16 114.45 101.25 1. x265 [info]: HEVC encoder version 3.5+1-f0c1022b6
oneDNN Harness: Convolution Batch Shapes Auto - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.6 Harness: Convolution Batch Shapes Auto - Engine: CPU 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 3 6 9 12 15 9.18658 9.16956 7.87367 6.36475 6.37150 4.11551 4.03550 4.03808 4.04877 6.67287 4.13321 MIN: 9.06 MIN: 9.03 MIN: 7.77 MIN: 6.28 MIN: 6.28 MIN: 4.05 MIN: 3.98 MIN: 3.98 MIN: 3.99 MIN: 6.2 MIN: 4.07 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl
SVT-AV1 Encoder Mode: Preset 13 - Input: Bosphorus 1080p OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 13 - Input: Bosphorus 1080p 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 200 400 600 800 1000 380.63 377.79 652.97 752.16 748.15 776.12 835.45 842.34 845.58 842.56 769.82 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
Llamafile Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Prompt Processing 256 OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.8.16 Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Prompt Processing 256 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 900 1800 2700 3600 4500 4096 4096 4096 4096 4096 4096 4096 4096 4096 4096
oneDNN Harness: Deconvolution Batch shapes_3d - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.6 Harness: Deconvolution Batch shapes_3d - Engine: CPU 41 41 b 4364P 4464p 4464p epyc 4484PX 45 4584PX EPYC 4584PX amd a px 2 4 6 8 10 7.53655 7.55315 3.78008 3.38958 3.38823 3.50840 2.66296 2.67219 2.66429 2.41294 3.51243 MIN: 7.53 MIN: 7.52 MIN: 3.64 MIN: 3.28 MIN: 3.28 MIN: 3.46 MIN: 2.62 MIN: 2.62 MIN: 2.62 MIN: 2.34 MIN: 3.47 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -fcf-protection=full -pie -ldl
Phoronix Test Suite v10.8.5