2 x Intel Xeon Max 9480 vs. 2 x AMD EPYC 9684X Benchmarks

Add Processor To Compare:

* Note: Due to differences in Linux drivers/interfaces, the reported CPU clock may be displayed different depending upon the CPU - i.e. base vs. boost frequency.
* Disclosure: OpenBenchmarking.org may earn a commission/referral on affiliate shopping links should you decide to make any purchases from the third-party shopping links.

Of the components selected for comparison, there were 457 benchmarks in common (matching test profile, test version and any configurable test options) and where that data was statistically significant and at least three independent data points for each component of the public test results on OpenBenchmarking.org.

2 x AMD EPYC 9684X 96-Core had the most wins, coming in first place for 82% of the tests.

Based on the geometric mean of all complete results, the fastest (2 x AMD EPYC 9684X 96-Core) was 1.68x the speed of the slowest (2 x Intel Xeon Max 9480).

The results with the greatest spread from best to worst included: Stress-NG (Test: x86_64 RdRand) at 108.329x, oneDNN (Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU) at 23.994x, Llama.cpp (Model: llama-2-13b.Q4_0.gguf) at 11.087x, oneDNN (Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU) at 10.282x, PostgreSQL (Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latency) at 5.436x, PostgreSQL (Scaling Factor: 100 - Clients: 1000 - Mode: Read Only) at 5.364x, PostgreSQL (Scaling Factor: 100 - Clients: 800 - Mode: Read Only - Average Latency) at 5.332x, PostgreSQL (Scaling Factor: 100 - Clients: 800 - Mode: Read Only) at 5.298x, OpenVINO (Model: Vehicle Detection FP16-INT8 - Device: CPU) at 5.13x, Xmrig (Variant: Monero - Hash Count: 1M) at 3.775x.

457 Benchmarks Found In Common - Sort By: Greatest Spread