blender onednn 2 x AMD EPYC 7513 32-Core testing with a Supermicro H12DSi-N6 v1.02 (2.3 BIOS) and astdrmfb on AlmaLinux 9.1 via the Phoronix Test Suite. a: Processor: 2 x AMD EPYC 7513 32-Core @ 2.60GHz (64 Cores / 128 Threads), Motherboard: Supermicro H12DSi-N6 v1.02 (2.3 BIOS), Memory: 512GB, Disk: 2 x 1920GB SAMSUNG MZQL21T9HCJR-00A07, Graphics: astdrmfb OS: AlmaLinux 9.1, Kernel: 5.14.0-162.18.1.el9_1.x86_64 (x86_64), Compiler: GCC 11.3.1 20220421, File-System: ext4, Screen Resolution: 1024x768 b: Processor: 2 x AMD EPYC 7513 32-Core @ 2.60GHz (64 Cores / 128 Threads), Motherboard: Supermicro H12DSi-N6 v1.02 (2.3 BIOS), Memory: 512GB, Disk: 2 x 1920GB SAMSUNG MZQL21T9HCJR-00A07, Graphics: astdrmfb OS: AlmaLinux 9.1, Kernel: 5.14.0-162.18.1.el9_1.x86_64 (x86_64), Compiler: GCC 11.3.1 20220421, File-System: ext4, Screen Resolution: 1024x768 c: Processor: 2 x AMD EPYC 7513 32-Core @ 2.60GHz (64 Cores / 128 Threads), Motherboard: Supermicro H12DSi-N6 v1.02 (2.3 BIOS), Memory: 512GB, Disk: 2 x 1920GB SAMSUNG MZQL21T9HCJR-00A07, Graphics: astdrmfb OS: AlmaLinux 9.1, Kernel: 5.14.0-162.18.1.el9_1.x86_64 (x86_64), Compiler: GCC 11.3.1 20220421, File-System: ext4, Screen Resolution: 1024x768 d: Processor: 2 x AMD EPYC 7513 32-Core @ 2.60GHz (64 Cores / 128 Threads), Motherboard: Supermicro H12DSi-N6 v1.02 (2.3 BIOS), Memory: 512GB, Disk: 2 x 1920GB SAMSUNG MZQL21T9HCJR-00A07, Graphics: astdrmfb OS: AlmaLinux 9.1, Kernel: 5.14.0-162.18.1.el9_1.x86_64 (x86_64), Compiler: GCC 11.3.1 20220421, File-System: ext4, Screen Resolution: 1024x768 oneDNN 3.1 Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU ms < Lower Is Better a . 3.06650 |=========================================================== b . 3.40413 |================================================================== c . 3.15140 |============================================================= d . 3.04330 |=========================================================== oneDNN 3.1 Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU ms < Lower Is Better a . 4.22668 |================================================================= b . 4.27313 |================================================================== c . 4.18548 |================================================================= d . 4.21100 |================================================================= oneDNN 3.1 Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better a . 3.67571 |================================================================ b . 3.76839 |================================================================== c . 3.51317 |============================================================== d . 3.43236 |============================================================ oneDNN 3.1 Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better a . 0.734539 |================================================================ b . 0.731117 |================================================================ c . 0.735732 |================================================================ d . 0.741873 |================================================================= oneDNN 3.1 Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better oneDNN 3.1 Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better oneDNN 3.1 Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU ms < Lower Is Better a . 0.768929 |================================================================ b . 0.774695 |================================================================= c . 0.772413 |================================================================ d . 0.779659 |================================================================= oneDNN 3.1 Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU ms < Lower Is Better a . 9.95882 |================================================================== b . 9.89322 |================================================================== c . 9.95733 |================================================================== d . 9.71121 |================================================================ oneDNN 3.1 Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU ms < Lower Is Better a . 2.03753 |================================================================== b . 2.01538 |================================================================= c . 1.95940 |=============================================================== d . 2.01117 |================================================================= oneDNN 3.1 Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better a . 2.66260 |================================================================== b . 2.62474 |================================================================= c . 2.65341 |================================================================== d . 2.60496 |================================================================= oneDNN 3.1 Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better a . 1.96931 |================================================================== b . 1.97396 |================================================================== c . 1.95326 |================================================================= d . 1.97916 |================================================================== oneDNN 3.1 Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better a . 0.653792 |================================================================ b . 0.661217 |================================================================= c . 0.656191 |================================================================= d . 0.661196 |================================================================= oneDNN 3.1 Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU ms < Lower Is Better a . 2608.02 |================================================================== b . 2585.07 |================================================================= c . 2601.27 |================================================================== d . 2612.25 |================================================================== oneDNN 3.1 Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU ms < Lower Is Better a . 1004.40 |================================================================== b . 989.99 |================================================================= c . 1009.93 |================================================================== d . 942.78 |============================================================== oneDNN 3.1 Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better a . 2599.05 |================================================================= b . 2626.18 |================================================================== c . 2562.25 |================================================================ d . 2564.27 |================================================================ oneDNN 3.1 Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better oneDNN 3.1 Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better oneDNN 3.1 Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better oneDNN 3.1 Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better a . 1000.71 |=============================================================== b . 1010.55 |================================================================ c . 1050.02 |================================================================== d . 962.87 |============================================================= oneDNN 3.1 Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better a . 2622.67 |================================================================= b . 2671.08 |================================================================== c . 2572.50 |================================================================ d . 2626.09 |================================================================= oneDNN 3.1 Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better a . 1008.26 |================================================================== b . 982.42 |================================================================ c . 959.01 |=============================================================== Blender 3.5 Blend File: BMW27 - Compute: CPU-Only Seconds < Lower Is Better a . 25.44 |==================================================================== b . 25.49 |==================================================================== c . 25.51 |==================================================================== Blender 3.5 Blend File: Classroom - Compute: CPU-Only Seconds < Lower Is Better a . 63.70 |==================================================================== b . 63.70 |==================================================================== c . 63.74 |==================================================================== Blender 3.5 Blend File: Fishy Cat - Compute: CPU-Only Seconds < Lower Is Better a . 32.14 |==================================================================== b . 32.08 |==================================================================== c . 32.25 |==================================================================== Blender 3.5 Blend File: Barbershop - Compute: CPU-Only Seconds < Lower Is Better a . 254.66 |=================================================================== b . 254.75 |=================================================================== c . 254.58 |=================================================================== Blender 3.5 Blend File: Pabellon Barcelona - Compute: CPU-Only Seconds < Lower Is Better a . 80.97 |==================================================================== b . 81.07 |==================================================================== c . 80.79 |====================================================================