GCC 13 Compiler Benchmarks AMD EPYC Genoa AMD EPYC 9654 GCC 13 development compiler benchmarks by Michael Larabel for a future article. Znver4: Processor: 2 x AMD EPYC 9654 96-Core @ 3.71GHz (192 Cores / 384 Threads), Motherboard: AMD Titanite_4G (RTI1002E BIOS), Chipset: AMD Device 14a4, Memory: 1520GB, Disk: 800GB INTEL SSDPF21Q800GB, Graphics: ASPEED, Monitor: VGA HDMI, Network: Broadcom NetXtreme BCM5720 PCIe OS: Ubuntu 23.04, Kernel: 5.19.0-21-generic (x86_64), Desktop: GNOME Shell 43.1, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 13.0.0 20230103, File-System: ext4, Screen Resolution: 1920x1080 Znver4 + Prefer AVX-512: Processor: 2 x AMD EPYC 9654 96-Core @ 3.71GHz (192 Cores / 384 Threads), Motherboard: AMD Titanite_4G (RTI1002E BIOS), Chipset: AMD Device 14a4, Memory: 1520GB, Disk: 800GB INTEL SSDPF21Q800GB, Graphics: ASPEED, Monitor: VGA HDMI, Network: Broadcom NetXtreme BCM5720 PCIe OS: Ubuntu 23.04, Kernel: 5.19.0-21-generic (x86_64), Desktop: GNOME Shell 43.1, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 13.0.0 20230103, File-System: ext4, Screen Resolution: 1920x1080 Znver3: Processor: 2 x AMD EPYC 9654 96-Core @ 3.71GHz (192 Cores / 384 Threads), Motherboard: AMD Titanite_4G (RTI1002E BIOS), Chipset: AMD Device 14a4, Memory: 1520GB, Disk: 800GB INTEL SSDPF21Q800GB, Graphics: ASPEED, Monitor: VGA HDMI, Network: Broadcom NetXtreme BCM5720 PCIe OS: Ubuntu 23.04, Kernel: 5.19.0-21-generic (x86_64), Desktop: GNOME Shell 43.1, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 13.0.0 20230103, File-System: ext4, Screen Resolution: 1920x1080 Znver3 + AVX-512: Processor: 2 x AMD EPYC 9654 96-Core @ 3.71GHz (192 Cores / 384 Threads), Motherboard: AMD Titanite_4G (RTI1002E BIOS), Chipset: AMD Device 14a4, Memory: 1520GB, Disk: 800GB INTEL SSDPF21Q800GB, Graphics: ASPEED, Monitor: VGA HDMI, Network: Broadcom NetXtreme BCM5720 PCIe OS: Ubuntu 23.04, Kernel: 5.19.0-21-generic (x86_64), Desktop: GNOME Shell 43.1, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 13.0.0 20230103, File-System: ext4, Screen Resolution: 1920x1080 Cpuminer-Opt 3.20.3 Algorithm: LBC, LBRY Credits kH/s > Higher Is Better Znver4 .................. 1065487 |=========================================== Znver4 + Prefer AVX-512 . 1085827 |============================================ Znver3 .................. 497020 |==================== Znver3 + AVX-512 ........ 1067743 |=========================================== Cpuminer-Opt 3.20.3 Algorithm: Quad SHA-256, Pyrite kH/s > Higher Is Better Znver4 .................. 2251067 |=========================================== Znver4 + Prefer AVX-512 . 2323995 |============================================ Znver3 .................. 1378987 |========================== Znver3 + AVX-512 ........ 2264747 |=========================================== Cpuminer-Opt 3.20.3 Algorithm: scrypt kH/s > Higher Is Better Znver4 .................. 4790.11 |============================================ Znver4 + Prefer AVX-512 . 4782.74 |============================================ Znver3 .................. 2959.15 |=========================== Znver3 + AVX-512 ........ 4763.91 |============================================ Cpuminer-Opt 3.20.3 Algorithm: Garlicoin kH/s > Higher Is Better Znver4 .................. 72413 |============================================== Znver4 + Prefer AVX-512 . 72130 |============================================== Znver3 .................. 49523 |=============================== Znver3 + AVX-512 ........ 72837 |============================================== Cpuminer-Opt 3.20.3 Algorithm: Skeincoin kH/s > Higher Is Better Znver4 .................. 2014770 |============================================ Znver4 + Prefer AVX-512 . 2009367 |============================================ Znver3 .................. 1414047 |=============================== Znver3 + AVX-512 ........ 2004990 |============================================ Cpuminer-Opt 3.20.3 Algorithm: x25x kH/s > Higher Is Better Znver4 .................. 8042.88 |=========================================== Znver4 + Prefer AVX-512 . 8217.70 |============================================ Znver3 .................. 6116.97 |================================= Znver3 + AVX-512 ........ 7941.38 |=========================================== GraphicsMagick 1.3.38 Operation: Enhanced Iterations Per Minute > Higher Is Better Znver4 .................. 2234 |=============================================== Znver4 + Prefer AVX-512 . 2150 |============================================= Znver3 .................. 1837 |======================================= Znver3 + AVX-512 ........ 2208 |============================================== ASTC Encoder 4.0 Preset: Medium MT/s > Higher Is Better Znver4 .................. 493.22 |=========================================== Znver4 + Prefer AVX-512 . 420.70 |===================================== Znver3 .................. 459.69 |======================================== Znver3 + AVX-512 ........ 511.52 |============================================= CP2K Molecular Dynamics 8.2 Input: Fayalite-FIST Seconds < Lower Is Better Znver4 .................. 1174.67 |===================================== Znver4 + Prefer AVX-512 . 1263.71 |======================================== Znver3 .................. 1211.13 |====================================== Znver3 + AVX-512 ........ 1407.13 |============================================ WebP Image Encode 1.2.4 Encode Settings: Quality 100, Highest Compression MP/s > Higher Is Better Znver4 .................. 3.25 |========================================== Znver4 + Prefer AVX-512 . 3.23 |========================================== Znver3 .................. 3.64 |=============================================== Znver3 + AVX-512 ........ 3.11 |======================================== GraphicsMagick 1.3.38 Operation: Swirl Iterations Per Minute > Higher Is Better Znver4 .................. 2862 |=============================================== Znver4 + Prefer AVX-512 . 2681 |============================================ Znver3 .................. 2563 |========================================== Znver3 + AVX-512 ........ 2826 |============================================== GraphicsMagick 1.3.38 Operation: Rotate Iterations Per Minute > Higher Is Better Znver4 .................. 673 |================================================ Znver4 + Prefer AVX-512 . 656 |=============================================== Znver3 .................. 645 |============================================== Znver3 + AVX-512 ........ 605 |=========================================== GraphicsMagick 1.3.38 Operation: HWB Color Space Iterations Per Minute > Higher Is Better Znver4 .................. 1180 |=============================================== Znver4 + Prefer AVX-512 . 1167 |============================================== Znver3 .................. 1134 |============================================= Znver3 + AVX-512 ........ 1062 |========================================== ASTC Encoder 4.0 Preset: Exhaustive MT/s > Higher Is Better Znver4 .................. 12.94 |============================================= Znver4 + Prefer AVX-512 . 13.09 |============================================== Znver3 .................. 12.22 |=========================================== Znver3 + AVX-512 ........ 13.03 |============================================== GROMACS 2022.1 Implementation: MPI CPU - Input: water_GMX50_bare Ns Per Day > Higher Is Better Znver4 .................. 19.49 |============================================== Znver4 + Prefer AVX-512 . 19.44 |============================================== Znver3 .................. 18.23 |=========================================== Znver3 + AVX-512 ........ 19.09 |============================================= Kripke 1.2.4 Throughput FoM > Higher Is Better Znver4 .................. 261562280 |======================================== Znver4 + Prefer AVX-512 . 254648533 |======================================= Znver3 .................. 263812847 |========================================= Znver3 + AVX-512 ........ 271735708 |========================================== simdjson 2.0 Throughput Test: DistinctUserID GB/s > Higher Is Better Znver4 .................. 6.53 |============================================= Znver4 + Prefer AVX-512 . 6.86 |=============================================== Znver3 .................. 6.46 |============================================ Znver3 + AVX-512 ........ 6.61 |============================================= oneDNN 3.0 Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better Znver4 .................. 4.26600 |========================================= Znver4 + Prefer AVX-512 . 4.33366 |========================================== Znver3 .................. 4.52440 |============================================ Znver3 + AVX-512 ........ 4.33658 |========================================== GraphicsMagick 1.3.38 Operation: Sharpen Iterations Per Minute > Higher Is Better Znver4 .................. 1359 |=============================================== Znver4 + Prefer AVX-512 . 1314 |============================================= Znver3 .................. 1285 |============================================ Znver3 + AVX-512 ........ 1321 |============================================== JPEG XL libjxl 0.7 Input: JPEG - Quality: 100 MP/s > Higher Is Better Znver4 .................. 0.73 |============================================== Znver4 + Prefer AVX-512 . 0.75 |=============================================== Znver3 .................. 0.74 |============================================== Znver3 + AVX-512 ........ 0.71 |============================================ oneDNN 3.0 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better Znver4 .................. 0.500579 |========================================== Znver4 + Prefer AVX-512 . 0.483979 |========================================= Znver3 .................. 0.510939 |=========================================== Znver3 + AVX-512 ........ 0.504842 |========================================== simdjson 2.0 Throughput Test: TopTweet GB/s > Higher Is Better Znver4 .................. 6.96 |=============================================== Znver4 + Prefer AVX-512 . 6.91 |=============================================== Znver3 .................. 6.61 |============================================= Znver3 + AVX-512 ........ 6.73 |============================================= oneDNN 3.0 Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better Znver4 .................. 2020.24 |========================================== Znver4 + Prefer AVX-512 . 2123.35 |============================================ Znver3 .................. 2108.46 |============================================ Znver3 + AVX-512 ........ 2099.43 |============================================ GPAW 22.1 Input: Carbon Nanotube Seconds < Lower Is Better Znver4 .................. 22.35 |============================================= Znver4 + Prefer AVX-512 . 22.82 |============================================== Znver3 .................. 21.72 |============================================ Znver3 + AVX-512 ........ 22.82 |============================================== GraphicsMagick 1.3.38 Operation: Noise-Gaussian Iterations Per Minute > Higher Is Better Znver4 .................. 1024 |=============================================== Znver4 + Prefer AVX-512 . 1013 |============================================== Znver3 .................. 1018 |=============================================== Znver3 + AVX-512 ........ 975 |============================================= Kvazaar 2.1 Video Input: Bosphorus 4K - Video Preset: Ultra Fast Frames Per Second > Higher Is Better Znver4 .................. 78.69 |============================================== Znver4 + Prefer AVX-512 . 74.98 |============================================ Znver3 .................. 78.01 |============================================== Znver3 + AVX-512 ........ 76.76 |============================================= simdjson 2.0 Throughput Test: PartialTweets GB/s > Higher Is Better Znver4 .................. 6.63 |============================================== Znver4 + Prefer AVX-512 . 6.43 |============================================= Znver3 .................. 6.74 |=============================================== Znver3 + AVX-512 ........ 6.68 |=============================================== oneDNN 3.0 Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better Znver4 .................. 2479.72 |=========================================== Znver4 + Prefer AVX-512 . 2444.52 |========================================== Znver3 .................. 2418.28 |========================================== Znver3 + AVX-512 ........ 2531.42 |============================================ JPEG XL Decoding libjxl 0.7 CPU Threads: All MP/s > Higher Is Better Znver4 .................. 277.20 |============================================= Znver4 + Prefer AVX-512 . 272.31 |============================================ Znver3 .................. 269.59 |============================================ Znver3 + AVX-512 ........ 266.53 |=========================================== libavif avifenc 0.11 Encoder Speed: 6 Seconds < Lower Is Better Znver4 .................. 2.347 |============================================= Znver4 + Prefer AVX-512 . 2.317 |============================================ Znver3 .................. 2.331 |============================================= Znver3 + AVX-512 ........ 2.406 |============================================== Cpuminer-Opt 3.20.3 Algorithm: Deepcoin kH/s > Higher Is Better Znver4 .................. 159147 |=========================================== Znver4 + Prefer AVX-512 . 162242 |============================================ Znver3 .................. 164993 |============================================= Znver3 + AVX-512 ........ 160157 |============================================ oneDNN 3.0 Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better Znver4 .................. 2405.63 |=========================================== Znver4 + Prefer AVX-512 . 2359.29 |=========================================== Znver3 .................. 2442.35 |============================================ Znver3 + AVX-512 ........ 2377.17 |=========================================== GraphicsMagick 1.3.38 Operation: Resizing Iterations Per Minute > Higher Is Better Znver4 .................. 88 |================================================ Znver4 + Prefer AVX-512 . 87 |================================================ Znver3 .................. 89 |================================================= Znver3 + AVX-512 ........ 86 |=============================================== JPEG XL Decoding libjxl 0.7 CPU Threads: 1 MP/s > Higher Is Better Znver4 .................. 48.64 |============================================== Znver4 + Prefer AVX-512 . 48.21 |============================================== Znver3 .................. 48.22 |============================================== Znver3 + AVX-512 ........ 47.02 |============================================ oneDNN 3.0 Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better Znver4 .................. 0.443891 |=========================================== Znver4 + Prefer AVX-512 . 0.445532 |=========================================== Znver3 .................. 0.442748 |=========================================== Znver3 + AVX-512 ........ 0.431299 |========================================== JPEG XL libjxl 0.7 Input: JPEG - Quality: 90 MP/s > Higher Is Better Znver4 .................. 9.48 |=============================================== Znver4 + Prefer AVX-512 . 9.43 |=============================================== Znver3 .................. 9.18 |============================================== Znver3 + AVX-512 ........ 9.27 |============================================== oneDNN 3.0 Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better Znver4 .................. 0.380486 |========================================== Znver4 + Prefer AVX-512 . 0.388483 |=========================================== Znver3 .................. 0.392765 |=========================================== Znver3 + AVX-512 ........ 0.392875 |=========================================== Zstd Compression 1.5.0 Compression Level: 19 - Compression Speed MB/s > Higher Is Better Znver4 .................. 102.1 |============================================= Znver4 + Prefer AVX-512 . 105.3 |============================================== Znver3 .................. 104.4 |============================================== Znver3 + AVX-512 ........ 102.9 |============================================= oneDNN 3.0 Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better Znver4 .................. 2134.85 |============================================ Znver4 + Prefer AVX-512 . 2070.13 |=========================================== Znver3 .................. 2093.75 |=========================================== Znver3 + AVX-512 ........ 2103.41 |=========================================== Stargate Digital Audio Workstation 22.11.5 Sample Rate: 192000 - Buffer Size: 1024 Render Ratio > Higher Is Better Znver4 .................. 2.783717 |========================================== Znver4 + Prefer AVX-512 . 2.867233 |=========================================== Znver3 .................. 2.820079 |========================================== Znver3 + AVX-512 ........ 2.868373 |=========================================== oneDNN 3.0 Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better Znver4 .................. 0.886642 |========================================== Znver4 + Prefer AVX-512 . 0.881303 |========================================== Znver3 .................. 0.907941 |=========================================== Znver3 + AVX-512 ........ 0.902303 |=========================================== Coremark 1.0 CoreMark Size 666 - Iterations Per Second Iterations/Sec > Higher Is Better Znver4 .................. 7694653.90 |======================================== Znver4 + Prefer AVX-512 . 7861097.86 |========================================= Znver3 .................. 7640546.27 |======================================== Znver3 + AVX-512 ........ 7871273.93 |========================================= libavif avifenc 0.11 Encoder Speed: 10, Lossless Seconds < Lower Is Better Znver4 .................. 3.541 |============================================= Znver4 + Prefer AVX-512 . 3.572 |============================================= Znver3 .................. 3.581 |============================================= Znver3 + AVX-512 ........ 3.647 |============================================== oneDNN 3.0 Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better Znver4 .................. 2.33363 |=========================================== Znver4 + Prefer AVX-512 . 2.29880 |=========================================== Znver3 .................. 2.31858 |=========================================== Znver3 + AVX-512 ........ 2.36166 |============================================ libavif avifenc 0.11 Encoder Speed: 0 Seconds < Lower Is Better Znver4 .................. 61.12 |============================================= Znver4 + Prefer AVX-512 . 61.07 |============================================= Znver3 .................. 61.66 |============================================= Znver3 + AVX-512 ........ 62.68 |============================================== libavif avifenc 0.11 Encoder Speed: 6, Lossless Seconds < Lower Is Better Znver4 .................. 4.398 |============================================= Znver4 + Prefer AVX-512 . 4.462 |============================================= Znver3 .................. 4.437 |============================================= Znver3 + AVX-512 ........ 4.514 |============================================== JPEG XL libjxl 0.7 Input: PNG - Quality: 90 MP/s > Higher Is Better Znver4 .................. 9.83 |=============================================== Znver4 + Prefer AVX-512 . 9.81 |=============================================== Znver3 .................. 9.61 |============================================== Znver3 + AVX-512 ........ 9.86 |=============================================== ASTC Encoder 4.0 Preset: Thorough MT/s > Higher Is Better Znver4 .................. 118.74 |============================================= Znver4 + Prefer AVX-512 . 118.99 |============================================= Znver3 .................. 116.09 |============================================ Znver3 + AVX-512 ........ 117.69 |============================================= JPEG XL libjxl 0.7 Input: PNG - Quality: 100 MP/s > Higher Is Better Znver4 .................. 0.83 |=============================================== Znver4 + Prefer AVX-512 . 0.82 |============================================== Znver3 .................. 0.82 |============================================== Znver3 + AVX-512 ........ 0.81 |============================================== simdjson 2.0 Throughput Test: LargeRandom GB/s > Higher Is Better Znver4 .................. 1.27 |=============================================== Znver4 + Prefer AVX-512 . 1.24 |============================================== Znver3 .................. 1.25 |============================================== Znver3 + AVX-512 ........ 1.25 |============================================== SVT-AV1 1.4 Encoder Mode: Preset 8 - Input: Bosphorus 4K Frames Per Second > Higher Is Better Znver4 .................. 93.76 |============================================= Znver4 + Prefer AVX-512 . 94.34 |============================================== Znver3 .................. 95.12 |============================================== Znver3 + AVX-512 ........ 92.96 |============================================= PJSIP 2.11 Method: INVITE Responses Per Second > Higher Is Better Znver4 .................. 5200 |=============================================== Znver4 + Prefer AVX-512 . 5084 |============================================== Znver3 .................. 5149 |=============================================== Znver3 + AVX-512 ........ 5132 |============================================== Cpuminer-Opt 3.20.3 Algorithm: Triple SHA-256, Onecoin kH/s > Higher Is Better Znver4 .................. 3306643 |============================================ Znver4 + Prefer AVX-512 . 3301217 |============================================ Znver3 .................. 3255253 |=========================================== Znver3 + AVX-512 ........ 3323680 |============================================ oneDNN 3.0 Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better Znver4 .................. 0.937164 |========================================== Znver4 + Prefer AVX-512 . 0.954140 |=========================================== Znver3 .................. 0.947013 |=========================================== Znver3 + AVX-512 ........ 0.936374 |========================================== Stargate Digital Audio Workstation 22.11.5 Sample Rate: 96000 - Buffer Size: 1024 Render Ratio > Higher Is Better Znver4 .................. 4.331230 |========================================== Znver4 + Prefer AVX-512 . 4.408290 |=========================================== Znver3 .................. 4.356446 |========================================== Znver3 + AVX-512 ........ 4.413379 |=========================================== Kvazaar 2.1 Video Input: Bosphorus 4K - Video Preset: Very Fast Frames Per Second > Higher Is Better Znver4 .................. 70.71 |============================================= Znver4 + Prefer AVX-512 . 70.81 |============================================= Znver3 .................. 72.01 |============================================== Znver3 + AVX-512 ........ 70.71 |============================================= oneDNN 3.0 Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better Znver4 .................. 0.274769 |========================================== Znver4 + Prefer AVX-512 . 0.274662 |========================================== Znver3 .................. 0.279655 |=========================================== Znver3 + AVX-512 ........ 0.276965 |=========================================== WebP Image Encode 1.2.4 Encode Settings: Quality 100, Lossless, Highest Compression MP/s > Higher Is Better Znver4 .................. 0.57 |============================================== Znver4 + Prefer AVX-512 . 0.58 |=============================================== Znver3 .................. 0.58 |=============================================== Znver3 + AVX-512 ........ 0.58 |=============================================== miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM1 Billion Interactions/s > Higher Is Better Znver4 .................. 214.50 |============================================= Znver4 + Prefer AVX-512 . 211.02 |============================================ Znver3 .................. 214.08 |============================================= Znver3 + AVX-512 ........ 214.63 |============================================= miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM1 GFInst/s > Higher Is Better Znver4 .................. 5362.54 |============================================ Znver4 + Prefer AVX-512 . 5275.43 |=========================================== Znver3 .................. 5351.94 |============================================ Znver3 + AVX-512 ........ 5365.61 |============================================ oneDNN 3.0 Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better Znver4 .................. 0.658101 |=========================================== Znver4 + Prefer AVX-512 . 0.650067 |========================================== Znver3 .................. 0.656342 |=========================================== Znver3 + AVX-512 ........ 0.647087 |========================================== LAMMPS Molecular Dynamics Simulator 23Jun2022 Model: Rhodopsin Protein ns/day > Higher Is Better Znver4 .................. 51.59 |============================================= Znver4 + Prefer AVX-512 . 52.29 |============================================== Znver3 .................. 51.46 |============================================= Znver3 + AVX-512 ........ 51.82 |============================================== Cpuminer-Opt 3.20.3 Algorithm: Magi kH/s > Higher Is Better Znver4 .................. 8440.79 |============================================ Znver4 + Prefer AVX-512 . 8467.24 |============================================ Znver3 .................. 8490.73 |============================================ Znver3 + AVX-512 ........ 8355.75 |=========================================== Ngspice 34 Circuit: C7552 Seconds < Lower Is Better Znver4 .................. 92.10 |============================================= Znver4 + Prefer AVX-512 . 93.45 |============================================== Znver3 .................. 92.94 |============================================== Znver3 + AVX-512 ........ 93.52 |============================================== WebP Image Encode 1.2.4 Encode Settings: Quality 100 MP/s > Higher Is Better Znver4 .................. 11.54 |============================================== Znver4 + Prefer AVX-512 . 11.54 |============================================== Znver3 .................. 11.48 |============================================== Znver3 + AVX-512 ........ 11.38 |============================================= WebP Image Encode 1.2.4 Encode Settings: Quality 100, Lossless MP/s > Higher Is Better Znver4 .................. 1.45 |============================================== Znver4 + Prefer AVX-512 . 1.47 |=============================================== Znver3 .................. 1.47 |=============================================== Znver3 + AVX-512 ........ 1.45 |============================================== SVT-AV1 1.4 Encoder Mode: Preset 4 - Input: Bosphorus 4K Frames Per Second > Higher Is Better Znver4 .................. 5.392 |============================================== Znver4 + Prefer AVX-512 . 5.423 |============================================== Znver3 .................. 5.360 |============================================= Znver3 + AVX-512 ........ 5.374 |============================================== Liquid-DSP 2021.01.31 Threads: 384 - Buffer Length: 256 - Filter Length: 57 samples/s > Higher Is Better Znver4 .................. 11249333333 |======================================== Znver4 + Prefer AVX-512 . 11270666667 |======================================== Znver3 .................. 11176000000 |======================================== Znver3 + AVX-512 ........ 11301666667 |======================================== libavif avifenc 0.11 Encoder Speed: 2 Seconds < Lower Is Better Znver4 .................. 34.10 |============================================== Znver4 + Prefer AVX-512 . 33.90 |============================================== Znver3 .................. 33.91 |============================================== Znver3 + AVX-512 ........ 34.23 |============================================== simdjson 2.0 Throughput Test: Kostya GB/s > Higher Is Better Znver4 .................. 4.17 |=============================================== Znver4 + Prefer AVX-512 . 4.19 |=============================================== Znver3 .................. 4.20 |=============================================== Znver3 + AVX-512 ........ 4.16 |=============================================== Kvazaar 2.1 Video Input: Bosphorus 4K - Video Preset: Medium Frames Per Second > Higher Is Better Znver4 .................. 64.35 |============================================== Znver4 + Prefer AVX-512 . 64.05 |============================================== Znver3 .................. 64.27 |============================================== Znver3 + AVX-512 ........ 63.80 |============================================== SecureMark 1.0.4 Benchmark: SecureMark-TLS marks > Higher Is Better Znver4 .................. 296548 |============================================= Znver4 + Prefer AVX-512 . 294122 |============================================= Znver3 .................. 294057 |============================================= Znver3 + AVX-512 ........ 296575 |============================================= Liquid-DSP 2021.01.31 Threads: 128 - Buffer Length: 256 - Filter Length: 57 samples/s > Higher Is Better Znver4 .................. 6990866667 |========================================= Znver4 + Prefer AVX-512 . 6977633333 |========================================= Znver3 .................. 6940833333 |========================================= Znver3 + AVX-512 ........ 6999233333 |========================================= Liquid-DSP 2021.01.31 Threads: 256 - Buffer Length: 256 - Filter Length: 57 samples/s > Higher Is Better Znver4 .................. 9789500000 |========================================= Znver4 + Prefer AVX-512 . 9809800000 |========================================= Znver3 .................. 9735666667 |========================================= Znver3 + AVX-512 ........ 9813700000 |========================================= QuantLib 1.21 MFLOPS > Higher Is Better Znver4 .................. 3096.9 |============================================= Znver4 + Prefer AVX-512 . 3112.6 |============================================= Znver3 .................. 3120.5 |============================================= Znver3 + AVX-512 ........ 3114.9 |============================================= miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM2 Billion Interactions/s > Higher Is Better Znver4 .................. 264.15 |============================================= Znver4 + Prefer AVX-512 . 265.52 |============================================= Znver3 .................. 264.28 |============================================= Znver3 + AVX-512 ........ 266.16 |============================================= miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM2 GFInst/s > Higher Is Better Znver4 .................. 6603.68 |============================================ Znver4 + Prefer AVX-512 . 6638.01 |============================================ Znver3 .................. 6607.09 |============================================ Znver3 + AVX-512 ........ 6653.86 |============================================ LAMMPS Molecular Dynamics Simulator 23Jun2022 Model: 20k Atoms ns/day > Higher Is Better Znver4 .................. 55.62 |============================================== Znver4 + Prefer AVX-512 . 55.74 |============================================== Znver3 .................. 55.72 |============================================== Znver3 + AVX-512 ........ 56.04 |============================================== WebP Image Encode 1.2.4 Encode Settings: Default MP/s > Higher Is Better Znver4 .................. 18.95 |============================================== Znver4 + Prefer AVX-512 . 18.97 |============================================== Znver3 .................. 18.99 |============================================== Znver3 + AVX-512 ........ 18.85 |============================================== PJSIP 2.11 Method: OPTIONS, Stateful Responses Per Second > Higher Is Better Znver4 .................. 9237 |=============================================== Znver4 + Prefer AVX-512 . 9288 |=============================================== Znver3 .................. 9226 |=============================================== Znver3 + AVX-512 ........ 9236 |=============================================== Zstd Compression 1.5.0 Compression Level: 19, Long Mode - Decompression Speed MB/s > Higher Is Better Znver4 .................. 3708.8 |============================================= Znver4 + Prefer AVX-512 . 3685.7 |============================================= Znver3 .................. 3695.5 |============================================= Znver3 + AVX-512 ........ 3684.7 |============================================= OpenSSL 3.0 Algorithm: SHA256 byte/s > Higher Is Better Znver4 .................. 265326713587 |======================================= Znver4 + Prefer AVX-512 . 266230124070 |======================================= Znver3 .................. 266464361193 |======================================= Znver3 + AVX-512 ........ 266899089453 |======================================= Ngspice 34 Circuit: C2670 Seconds < Lower Is Better Znver4 .................. 95.07 |============================================== Znver4 + Prefer AVX-512 . 95.60 |============================================== Znver3 .................. 95.18 |============================================== Znver3 + AVX-512 ........ 95.19 |============================================== Zstd Compression 1.5.0 Compression Level: 19 - Decompression Speed MB/s > Higher Is Better Znver4 .................. 3584.1 |============================================= Znver4 + Prefer AVX-512 . 3581.6 |============================================= Znver3 .................. 3574.9 |============================================= Znver3 + AVX-512 ........ 3594.8 |============================================= OpenSSL 3.0 Algorithm: RSA4096 verify/s > Higher Is Better Znver4 .................. 2939503.5 |========================================== Znver4 + Prefer AVX-512 . 2924586.4 |========================================== Znver3 .................. 2938488.5 |========================================== Znver3 + AVX-512 ........ 2935372.1 |========================================== ACES DGEMM 1.0 Sustained Floating-Point Rate GFLOP/s > Higher Is Better Znver4 .................. 70.05 |============================================== Znver4 + Prefer AVX-512 . 70.30 |============================================== Znver3 .................. 70.38 |============================================== Znver3 + AVX-512 ........ 70.19 |============================================== OpenSSL 3.0 Algorithm: RSA4096 sign/s > Higher Is Better Znver4 .................. 44490.1 |============================================ Znver4 + Prefer AVX-512 . 44301.8 |============================================ Znver3 .................. 44435.3 |============================================ Znver3 + AVX-512 ........ 44499.3 |============================================ SMHasher 2022-08-22 Hash: t1ha0_aes_avx2 x86_64 MiB/sec > Higher Is Better Znver4 .................. 102354.57 |========================================== Znver4 + Prefer AVX-512 . 102399.74 |========================================== Znver3 .................. 102351.87 |========================================== Znver3 + AVX-512 ........ 102403.92 |========================================== SMHasher 2022-08-22 Hash: MeowHash x86_64 AES-NI MiB/sec > Higher Is Better Znver4 .................. 54272.14 |=========================================== Znver4 + Prefer AVX-512 . 54297.04 |=========================================== Znver3 .................. 54281.94 |=========================================== Znver3 + AVX-512 ........ 54284.65 |=========================================== SMHasher 2022-08-22 Hash: FarmHash32 x86_64 AVX MiB/sec > Higher Is Better Znver4 .................. 40565.07 |=========================================== Znver4 + Prefer AVX-512 . 40563.96 |=========================================== Znver3 .................. 40565.72 |=========================================== Znver3 + AVX-512 ........ 40559.33 |=========================================== oneDNN 3.0 Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better Znver4 .................. 21.14 |========================================= Znver4 + Prefer AVX-512 . 23.53 |============================================== Znver3 .................. 19.59 |====================================== Znver3 + AVX-512 ........ 22.94 |============================================= oneDNN 3.0 Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better Znver4 .................. 14.82 |============================================== Znver4 + Prefer AVX-512 . 14.29 |============================================ Znver3 .................. 14.62 |============================================= Znver3 + AVX-512 ........ 14.25 |============================================ PJSIP 2.11 Method: OPTIONS, Stateless Responses Per Second > Higher Is Better Znver4 .................. 335767 |============================================= Znver4 + Prefer AVX-512 . 336791 |============================================= Znver3 .................. 336885 |============================================= Znver3 + AVX-512 ........ 336615 |============================================= SVT-AV1 1.4 Encoder Mode: Preset 13 - Input: Bosphorus 4K Frames Per Second > Higher Is Better Znver4 .................. 196.12 |======================================== Znver4 + Prefer AVX-512 . 208.64 |=========================================== Znver3 .................. 219.57 |============================================= Znver3 + AVX-512 ........ 210.43 |=========================================== SVT-AV1 1.4 Encoder Mode: Preset 12 - Input: Bosphorus 4K Frames Per Second > Higher Is Better Znver4 .................. 210.39 |========================================== Znver4 + Prefer AVX-512 . 196.80 |======================================== Znver3 .................. 222.90 |============================================= Znver3 + AVX-512 ........ 206.68 |========================================== Zstd Compression 1.5.0 Compression Level: 19, Long Mode - Compression Speed MB/s > Higher Is Better Znver4 .................. 40.8 |============================================ Znver4 + Prefer AVX-512 . 42.5 |============================================== Znver3 .................. 39.8 |=========================================== Znver3 + AVX-512 ........ 43.9 |=============================================== SMHasher 2022-08-22 Hash: MeowHash x86_64 AES-NI cycles/hash < Lower Is Better Znver4 .................. 44.98 |============================================== Znver4 + Prefer AVX-512 . 44.95 |============================================== Znver3 .................. 44.96 |============================================== Znver3 + AVX-512 ........ 44.97 |============================================== SMHasher 2022-08-22 Hash: t1ha0_aes_avx2 x86_64 cycles/hash < Lower Is Better Znver4 .................. 20.52 |============================================= Znver4 + Prefer AVX-512 . 20.81 |============================================== Znver3 .................. 20.53 |============================================= Znver3 + AVX-512 ........ 20.53 |============================================= SMHasher 2022-08-22 Hash: FarmHash32 x86_64 AVX cycles/hash < Lower Is Better Znver4 .................. 26.49 |============================================== Znver4 + Prefer AVX-512 . 26.50 |============================================== Znver3 .................. 26.49 |============================================== Znver3 + AVX-512 ........ 26.49 |==============================================