AMD EPYC 9755 Delivers Decisive Performance Leadership Over Xeon 6 Granite Rapids With Nearly 500 Benchmarks

Written by Michael Larabel in Processors on 30 January 2026 at 10:30 AM EST. Page 3 of 8. 8 Comments.

For the OpenVINO AI toolkit that is led by Intel software engineers, the Xeon 6980P does pick up a few wins here thanks to OpenVINO supporting Advanced Matrix Extensions (AMX) use. AMX and MRDIMM capabilities remain the most compelling advantage for Granite Rapids processors. In some OpenVINO tests the AMD EPYC Zen 5 processors did still manage to secure the top spot, but at least it's some healthy competition here.

👁 Llama.cpp benchmark with settings of Backend: CPU BLAS, Model: Llama-3.1-Tulu-3-8B-Q8_0, Test: Text Generation 128. EPYC 9755 2P was the fastest.

👁 Llama.cpp benchmark with settings of Backend: CPU BLAS, Model: Llama-3.1-Tulu-3-8B-Q8_0, Test: Prompt Processing 2048. Xeon 6980P 2P was the fastest.

👁 Llama.cpp benchmark with settings of Backend: CPU BLAS, Model: Mistral-7B-Instruct-v0.3-Q8_0, Test: Text Generation 128. EPYC 9755 2P was the fastest.

👁 Llama.cpp benchmark with settings of Backend: CPU BLAS, Model: Mistral-7B-Instruct-v0.3-Q8_0, Test: Prompt Processing 2048. Xeon 6980P 2P was the fastest.

👁 Llama.cpp benchmark with settings of Backend: CPU BLAS, Model: gpt-oss-20b-Q8_0, Test: Text Generation 128. EPYC 9755 2P was the fastest.

👁 Llama.cpp benchmark with settings of Backend: CPU BLAS, Model: gpt-oss-20b-Q8_0, Test: Prompt Processing 2048. EPYC 9755 2P was the fastest.

👁 Llama.cpp benchmark with settings of Backend: CPU BLAS, Model: DeepSeek-R1-Distill-Llama-8B-Q8_0, Test: Text Generation 128. EPYC 9755 2P was the fastest.

👁 Llama.cpp benchmark with settings of Backend: CPU BLAS, Model: DeepSeek-R1-Distill-Llama-8B-Q8_0, Test: Prompt Processing 2048. Xeon 6980P 2P was the fastest.

👁 Llama.cpp benchmark with settings of Backend: CPU BLAS, Model: Qwen3-8B-Q8_0, Test: Prompt Processing 2048. Xeon 6980P 2P was the fastest.

Similarly, with Llama.cpp for CPU-based AI inferencing in some cases the Xeon 6980P was able to take the lead thanks to AMX and MRDIMMs but in other LLMs the AMD EPYC 9755 processors achieved the best performance.

👁 PyTorch benchmark with settings of Device: CPU, Batch Size: 256, Model: ResNet-50. EPYC 9755 2P was the fastest.

👁 PyTorch benchmark with settings of Device: CPU, Batch Size: 512, Model: ResNet-50. EPYC 9755 2P was the fastest.

👁 TensorFlow benchmark with settings of Device: CPU, Batch Size: 256, Model: ResNet-50. EPYC 9755 2P was the fastest.

👁 TensorFlow benchmark with settings of Device: CPU, Batch Size: 512, Model: ResNet-50. EPYC 9755 2P was the fastest.

When just looking at the basic PyTorch and TensorFlow performance, the AMD EPYC 9755 server was far outpacing the Xeon 6980P Granite Rapids setup.

8 Comments - Next Page

Page: 1 2 3 4 5 6 7 8 Next Page

URL: https://www.phoronix.com/review/amd-epyc-intel-xeon-500/3

⇱ AMD EPYC 9755 Delivers Decisive Performance Leadership Over Xeon 6 Granite Rapids With Nearly 500 Benchmarks Review - Phoronix

AMD EPYC 9755 Delivers Decisive Performance Leadership Over Xeon 6 Granite Rapids With Nearly 500 Benchmarks