sysbench oneDNN Ryzen 9 5950X AMD Ryzen 9 5950X 16-Core testing with a ASUS ROG CROSSHAIR VIII HERO (WI-FI) (3204 BIOS) and llvmpipe on Ubuntu 20.10 via the Phoronix Test Suite. 1: Processor: AMD Ryzen 9 5950X 16-Core @ 3.40GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (WI-FI) (3204 BIOS), Chipset: AMD Starship/Matisse, Memory: 32GB, Disk: 2000GB Corsair Force MP600 + 2000GB, Graphics: llvmpipe, Audio: AMD Device ab28, Network: Realtek RTL8125 2.5GbE + Intel I211 + Intel Wi-Fi 6 AX200 OS: Ubuntu 20.10, Kernel: 5.10.23-051023-generic (x86_64), Desktop: GNOME Shell 3.38.2, Display Server: X Server 1.20.9, OpenGL: 4.5 Mesa 21.1.0-devel (git-684f97d 2021-03-12 groovy-oibaf-ppa) (LLVM 11.0.1 256 bits), Vulkan: 1.0.168, Compiler: GCC 10.2.0, File-System: ext4, Screen Resolution: 3840x2160 2: Processor: AMD Ryzen 9 5950X 16-Core @ 3.40GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (WI-FI) (3204 BIOS), Chipset: AMD Starship/Matisse, Memory: 32GB, Disk: 2000GB Corsair Force MP600 + 2000GB, Graphics: llvmpipe, Audio: AMD Device ab28, Network: Realtek RTL8125 2.5GbE + Intel I211 + Intel Wi-Fi 6 AX200 OS: Ubuntu 20.10, Kernel: 5.10.23-051023-generic (x86_64), Desktop: GNOME Shell 3.38.2, Display Server: X Server 1.20.9, OpenGL: 4.5 Mesa 21.1.0-devel (git-684f97d 2021-03-12 groovy-oibaf-ppa) (LLVM 11.0.1 256 bits), Vulkan: 1.0.168, Compiler: GCC 10.2.0, File-System: ext4, Screen Resolution: 3840x2160 3: Processor: AMD Ryzen 9 5950X 16-Core @ 3.40GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (WI-FI) (3204 BIOS), Chipset: AMD Starship/Matisse, Memory: 32GB, Disk: 2000GB Corsair Force MP600 + 2000GB, Graphics: llvmpipe, Audio: AMD Device ab28, Network: Realtek RTL8125 2.5GbE + Intel I211 + Intel Wi-Fi 6 AX200 OS: Ubuntu 20.10, Kernel: 5.10.23-051023-generic (x86_64), Desktop: GNOME Shell 3.38.2, Display Server: X Server 1.20.9, OpenGL: 4.5 Mesa 21.1.0-devel (git-684f97d 2021-03-12 groovy-oibaf-ppa) (LLVM 11.0.1 256 bits), Vulkan: 1.0.168, Compiler: GCC 10.2.0, File-System: ext4, Screen Resolution: 3840x2160 4: Processor: AMD Ryzen 9 5950X 16-Core @ 3.40GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (WI-FI) (3204 BIOS), Chipset: AMD Starship/Matisse, Memory: 32GB, Disk: 2000GB Corsair Force MP600 + 2000GB, Graphics: llvmpipe, Audio: AMD Device ab28, Network: Realtek RTL8125 2.5GbE + Intel I211 + Intel Wi-Fi 6 AX200 OS: Ubuntu 20.10, Kernel: 5.10.23-051023-generic (x86_64), Desktop: GNOME Shell 3.38.2, Display Server: X Server 1.20.9, OpenGL: 4.5 Mesa 21.1.0-devel (git-684f97d 2021-03-12 groovy-oibaf-ppa) (LLVM 11.0.1 256 bits), Vulkan: 1.0.168, Compiler: GCC 10.2.0, File-System: ext4, Screen Resolution: 3840x2160 5: Processor: AMD Ryzen 9 5950X 16-Core @ 3.40GHz (16 Cores / 32 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (WI-FI) (3204 BIOS), Chipset: AMD Starship/Matisse, Memory: 32GB, Disk: 2000GB Corsair Force MP600 + 2000GB, Graphics: llvmpipe, Audio: AMD Device ab28, Network: Realtek RTL8125 2.5GbE + Intel I211 + Intel Wi-Fi 6 AX200 OS: Ubuntu 20.10, Kernel: 5.10.23-051023-generic (x86_64), Desktop: GNOME Shell 3.38.2, Display Server: X Server 1.20.9, OpenGL: 4.5 Mesa 21.1.0-devel (git-684f97d 2021-03-12 groovy-oibaf-ppa) (LLVM 11.0.1 256 bits), Vulkan: 1.0.168, Compiler: GCC 10.2.0, File-System: ext4, Screen Resolution: 3840x2160 oneDNN 2.1.2 Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 8.68256 |================================================================ 2 . 8.61122 |================================================================ 3 . 8.61285 |================================================================ 4 . 8.81145 |================================================================= 5 . 8.89636 |================================================================== oneDNN 2.1.2 Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 1755.70 |================================================================== 2 . 1722.24 |================================================================= 3 . 1727.29 |================================================================= 4 . 1721.03 |================================================================= 5 . 1714.95 |================================================================ oneDNN 2.1.2 Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 2713.47 |================================================================== 2 . 2692.66 |================================================================= 3 . 2676.92 |================================================================= 4 . 2729.23 |================================================================== 5 . 2691.73 |================================================================= oneDNN 2.1.2 Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 0.461770 |================================================================= 2 . 0.456507 |================================================================ 3 . 0.454507 |================================================================ 4 . 0.458384 |================================================================= 5 . 0.459646 |================================================================= oneDNN 2.1.2 Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better 1 . 1740.78 |================================================================== 2 . 1715.76 |================================================================= 3 . 1713.75 |================================================================= 4 . 1716.20 |================================================================= 5 . 1736.71 |================================================================== oneDNN 2.1.2 Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 18.57 |=================================================================== 2 . 18.65 |=================================================================== 3 . 18.53 |=================================================================== 4 . 18.72 |==================================================================== 5 . 18.80 |==================================================================== oneDNN 2.1.2 Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 2725.57 |================================================================== 2 . 2705.25 |================================================================== 3 . 2695.28 |================================================================= 4 . 2695.62 |================================================================= 5 . 2708.38 |================================================================== oneDNN 2.1.2 Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 1.06852 |================================================================== 2 . 1.06278 |================================================================== 3 . 1.05853 |================================================================= 4 . 1.05688 |================================================================= 5 . 1.05775 |================================================================= oneDNN 2.1.2 Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 1746.00 |================================================================== 2 . 1727.54 |================================================================= 3 . 1739.58 |================================================================== 4 . 1736.05 |================================================================== 5 . 1733.88 |================================================================== oneDNN 2.1.2 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 0.639642 |================================================================= 2 . 0.638707 |================================================================= 3 . 0.633926 |================================================================ 4 . 0.636019 |================================================================= 5 . 0.636844 |================================================================= oneDNN 2.1.2 Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 17.04 |=================================================================== 2 . 17.04 |=================================================================== 3 . 17.19 |==================================================================== 4 . 17.10 |==================================================================== 5 . 17.10 |==================================================================== oneDNN 2.1.2 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 1.41824 |================================================================== 2 . 1.41025 |================================================================== 3 . 1.40724 |================================================================= 4 . 1.40654 |================================================================= 5 . 1.41415 |================================================================== oneDNN 2.1.2 Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 3.95846 |================================================================== 2 . 3.95023 |================================================================== 3 . 3.93489 |================================================================== 4 . 3.94331 |================================================================== 5 . 3.92950 |================================================================== oneDNN 2.1.2 Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 1.67774 |================================================================== 2 . 1.67465 |================================================================== 3 . 1.67638 |================================================================== 4 . 1.67959 |================================================================== 5 . 1.66973 |================================================================== oneDNN 2.1.2 Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better 1 . 0.834899 |================================================================= 2 . 0.831855 |================================================================= 3 . 0.833777 |================================================================= 4 . 0.830406 |================================================================= 5 . 0.832645 |================================================================= Sysbench 1.0.20 Test: CPU Events Per Second > Higher Is Better 1 . 90905.95 |================================================================= 2 . 91133.12 |================================================================= 3 . 91295.96 |================================================================= 4 . 91225.86 |================================================================= 5 . 91042.94 |================================================================= oneDNN 2.1.2 Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 3.57674 |================================================================== 2 . 3.56735 |================================================================== 3 . 3.56892 |================================================================== 4 . 3.56211 |================================================================== 5 . 3.57154 |================================================================== Sysbench 1.0.20 Test: RAM / Memory MiB/sec > Higher Is Better 1 . 14807.78 |================================================================= 2 . 14841.97 |================================================================= 3 . 14820.94 |================================================================= 4 . 14845.30 |================================================================= 5 . 14868.16 |================================================================= oneDNN 2.1.2 Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better 1 . 2708.58 |================================================================== 2 . 2700.71 |================================================================== 3 . 2706.44 |================================================================== 4 . 2702.76 |================================================================== 5 . 2699.18 |================================================================== oneDNN 2.1.2 Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU ms < Lower Is Better 1 . 4.94662 |================================================================ 2 . 5.14019 |================================================================== 3 . 4.81443 |============================================================== 4 . 4.58698 |=========================================================== 5 . 4.70312 |============================================================