3970X april AMD Ryzen Threadripper 3970X 32-Core testing with a ASUS ROG ZENITH II EXTREME (1502 BIOS) and AMD Radeon RX 5700 8GB on Ubuntu 21.10 via the Phoronix Test Suite. A: Processor: AMD Ryzen Threadripper 3970X 32-Core @ 3.70GHz (32 Cores / 64 Threads), Motherboard: ASUS ROG ZENITH II EXTREME (1502 BIOS), Chipset: AMD Starship/Matisse, Memory: 64GB, Disk: Samsung SSD 980 PRO 500GB, Graphics: AMD Radeon RX 5700 8GB (1750/875MHz), Audio: AMD Navi 10 HDMI Audio, Monitor: ASUS VP28U, Network: Aquantia AQC107 NBase-T/IEEE + Intel I211 + Intel Wi-Fi 6 AX200 OS: Ubuntu 21.10, Kernel: 5.15.5-051505-generic (x86_64), Desktop: GNOME Shell 40.5, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 21.2.2 (LLVM 12.0.1), Vulkan: 1.2.182, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 3840x2160 B: Processor: AMD Ryzen Threadripper 3970X 32-Core @ 3.70GHz (32 Cores / 64 Threads), Motherboard: ASUS ROG ZENITH II EXTREME (1502 BIOS), Chipset: AMD Starship/Matisse, Memory: 64GB, Disk: Samsung SSD 980 PRO 500GB, Graphics: AMD Radeon RX 5700 8GB (1750/875MHz), Audio: AMD Navi 10 HDMI Audio, Monitor: ASUS VP28U, Network: Aquantia AQC107 NBase-T/IEEE + Intel I211 + Intel Wi-Fi 6 AX200 OS: Ubuntu 21.10, Kernel: 5.15.5-051505-generic (x86_64), Desktop: GNOME Shell 40.5, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 21.2.2 (LLVM 12.0.1), Vulkan: 1.2.182, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 3840x2160 C: Processor: AMD Ryzen Threadripper 3970X 32-Core @ 3.70GHz (32 Cores / 64 Threads), Motherboard: ASUS ROG ZENITH II EXTREME (1502 BIOS), Chipset: AMD Starship/Matisse, Memory: 64GB, Disk: Samsung SSD 980 PRO 500GB, Graphics: AMD Radeon RX 5700 8GB (1750/875MHz), Audio: AMD Navi 10 HDMI Audio, Monitor: ASUS VP28U, Network: Aquantia AQC107 NBase-T/IEEE + Intel I211 + Intel Wi-Fi 6 AX200 OS: Ubuntu 21.10, Kernel: 5.15.5-051505-generic (x86_64), Desktop: GNOME Shell 40.5, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 21.2.2 (LLVM 12.0.1), Vulkan: 1.2.182, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 3840x2160 perf-bench Benchmark: Epoll Wait ops/sec > Higher Is Better A . 13411 |==================================================================== B . 13003 |================================================================== C . 12947 |================================================================== perf-bench Benchmark: Futex Hash ops/sec > Higher Is Better A . 4555052 |================================================================== B . 4567109 |================================================================== C . 4574779 |================================================================== perf-bench Benchmark: Memcpy 1MB GB/sec > Higher Is Better A . 14.650909 |================================================================ B . 10.658625 |=============================================== C . 9.500946 |========================================== perf-bench Benchmark: Memset 1MB GB/sec > Higher Is Better A . 74.07 |============================================================== B . 80.80 |==================================================================== C . 73.18 |============================================================== perf-bench Benchmark: Sched Pipe ops/sec > Higher Is Better A . 240248 |=========================================== B . 377962 |=================================================================== C . 326836 |========================================================== perf-bench Benchmark: Futex Lock-Pi ops/sec > Higher Is Better A . 189 |==================================================================== B . 187 |=================================================================== C . 196 |====================================================================== perf-bench Benchmark: Syscall Basic ops/sec > Higher Is Better A . 18727173 |================================================================= B . 18705119 |================================================================= C . 18699792 |================================================================= oneDNN 2.6 Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 1.34126 |================================================================== B . 1.25925 |============================================================== C . 1.26163 |============================================================== oneDNN 2.6 Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 4.23007 |===================================================== B . 5.13321 |================================================================ C . 5.28602 |================================================================== oneDNN 2.6 Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 1.25689 |================================================================ B . 1.17561 |============================================================ C . 1.29343 |================================================================== oneDNN 2.6 Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 0.846616 |============================================================ B . 0.891938 |================================================================ C . 0.910685 |================================================================= oneDNN 2.6 Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better oneDNN 2.6 Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better oneDNN 2.6 Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 5.37336 |============================================================ B . 5.70658 |================================================================ C . 5.89321 |================================================================== oneDNN 2.6 Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 4.04769 |================================================================== B . 3.77460 |============================================================== C . 3.86946 |=============================================================== oneDNN 2.6 Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 2.69958 |================================================================== B . 2.66137 |================================================================= C . 2.66465 |================================================================= oneDNN 2.6 Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 5.93698 |============================================================ B . 6.43645 |================================================================= C . 6.50507 |================================================================== oneDNN 2.6 Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 1.44982 |================================================================== B . 1.44523 |================================================================== C . 1.44296 |================================================================== oneDNN 2.6 Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 1.53131 |================================================================== B . 1.54297 |================================================================== C . 1.54292 |================================================================== oneDNN 2.6 Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 4142.38 |================================================================== B . 4136.25 |================================================================== C . 4115.60 |================================================================== oneDNN 2.6 Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 1169.78 |============================================================= B . 1256.25 |================================================================== C . 1252.01 |================================================================== oneDNN 2.6 Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 4143.75 |================================================================== B . 4112.62 |================================================================== C . 4116.99 |================================================================== oneDNN 2.6 Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better oneDNN 2.6 Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better oneDNN 2.6 Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better oneDNN 2.6 Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 1165.53 |============================================================= B . 1256.61 |================================================================== C . 1251.49 |================================================================== oneDNN 2.6 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU ms < Lower Is Better A . 4.16374 |=========================================================== B . 4.65854 |================================================================== C . 3.31743 |=============================================== oneDNN 2.6 Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better A . 4152.24 |================================================================== B . 4036.93 |================================================================ C . 4119.52 |================================================================= oneDNN 2.6 Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better A . 1250.46 |================================================================= B . 1273.28 |================================================================== C . 1210.42 |=============================================================== oneDNN 2.6 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better A . 10.24 |============================================================== B . 10.64 |================================================================ C . 11.25 |==================================================================== oneDNN 2.6 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better Java JMH Throughput Ops/s > Higher Is Better A . 63748499441.07 |=========================================================== B . 63696734759.63 |=========================================================== C . 63765483746.80 |===========================================================