AMD EPYC 9755 / 9575F / 9965 Benchmarks Show Dominating Performance

Written by Michael Larabel in Processors on 10 October 2024 at 02:00 PM EDT. Page 12 of 14. 173 Comments.

👁 TensorFlow benchmark with settings of Device: CPU, Batch Size: 256, Model: ResNet-50. EPYC 9575F was the fastest.

The EPYC Turin processors have lept over the Granite Rapids flagship processor with the TensorFlow ResNet-50 performance on the CPU.

👁 TensorFlow benchmark with settings of Device: CPU, Batch Size: 256, Model: ResNet-50. EPYC 9575F was the fastest.

The EPYC 9005 series performance with TensorFlow was also leading for power efficiency and value.

👁 TensorFlow benchmark with settings of Device: CPU, Batch Size: 256, Model: AlexNet. EPYC 9575F 2P was the fastest.

👁 TensorFlow benchmark with settings of Device: CPU, Batch Size: 512, Model: ResNet-50. EPYC 9575F 2P was the fastest.

The AMD EPYC 9005 series with AVX-512 now on a 512-bit data path makes for great CPU-based AI inferencing.

👁 OpenVINO benchmark with settings of Model: Face Detection FP16-INT8, Device: CPU. Xeon 6980P 2P @ MRDIMM-8800 was the fastest.

The main strength of the Xeon 6 Granite Rapids besides memory intensive workloads where MRDIMMs can help is for workloads able to leverage Advanced Matrix Extensions (AMX). With Intel's OpenVINO AI toolkit being a leading example for demonstrating AVX usage, the Xeon 6980P processors were leading with some AI models. At the extremes a single Xeon 6980P could outperform two EPYC Turin processors.

👁 OpenVINO benchmark with settings of Model: Face Detection FP16-INT8, Device: CPU. EPYC 9575F 2P was the fastest.

The AMD EPYC Turin processors were enjoying lower latency than the Intel CPUs with some models.

👁 OpenVINO benchmark with settings of Model: Age Gender Recognition Retail 0013 FP16-INT8, Device: CPU. EPYC 9755 2P was the fastest.

👁 OpenVINO benchmark with settings of Model: Person Detection FP16, Device: CPU. EPYC 9755 2P was the fastest.

And with some OpenVINO AI models the AMD EPYC 9005 series did manage to outperform the Xeon 6980P Granite Rapids processors.

👁 OpenVINO benchmark with settings of Model: Weld Porosity Detection FP16-INT8, Device: CPU. Xeon 6980P 2P @ DDR5-6400 was the fastest.

👁 OpenVINO benchmark with settings of Model: Weld Porosity Detection FP16-INT8, Device: CPU. Xeon Platinum 8592+ was the fastest.

👁 OpenVINO benchmark with settings of Model: Person Vehicle Bike Detection FP16, Device: CPU. Xeon 6980P 2P @ MRDIMM-8800 was the fastest.

👁 OpenVINO benchmark with settings of Model: Machine Translation EN To DE FP16, Device: CPU. EPYC 9965 2P was the fastest.

👁 OpenVINO benchmark with settings of Model: Noise Suppression Poconet-Like FP16, Device: CPU. EPYC 9575F 2P was the fastest.

With OpenVINO and other CPU-based AI workloads is one of the areas where Granite Rapids is most competitive if the software can leverage AMX.

👁 OpenVINO benchmark with settings of Model: Noise Suppression Poconet-Like FP16, Device: CPU. EPYC 9575F 2P was the fastest.

The lower pricing on the EPYC 9005 series does help it maintain a lead in performance-per-dollar even for OpenVINO.

173 Comments - Next Page

Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Next Page

URL: https://www.phoronix.com/review/amd-epyc-9965-9755-benchmarks/12

⇱ AMD EPYC 9755 / 9575F / 9965 Benchmarks Show Dominating Performance Review - Phoronix

AMD EPYC 9755 / 9575F / 9965 Benchmarks Show Dominating Performance