MLPerf Inference v4.1 closed results: AMD 8xMI300X system definition

Fields

id

sources/bench-amd-mlperf-v4-1-mi300x-system

URL

https://raw.githubusercontent.com/mlcommons/inference_results_v4.1/main/closed/AMD/systems/8xMI300X_2xEPYC-9374F.json

Publisher

MLCommons

Kind

benchmark

Accessed

2026-10-03

notes

| AUDIT 2026-10-03 RE-VERIFIED: fetched the raw JSON directly and parsed it. It is live and contains exactly the hardware description the MLPerf result needs - accelerator_model_name 'AMD Instinct MI300X-NPS1-SPX-192GB-750W', accelerators_per_node 8, accelerator_memory_capacity '192 GB', accelerator_interconnect 'XGMI', accelerator_host_interconnect 'PCIe Gen5 x16', system_name 'Supermicro AS-8125GS-TNMR2'. The 000 status in a plain HTTP probe is GitHub raw's bot behaviour under concurrency, not a dead link; notes were empty and are now filled with the verified fields.

Notes

| AUDIT 2026-10-03 RE-VERIFIED: fetched the raw JSON directly and parsed it. It is live and contains exactly the hardware description the MLPerf result needs - accelerator_model_name 'AMD Instinct MI300X-NPS1-SPX-192GB-750W', accelerators_per_node 8, accelerator_memory_capacity '192 GB', accelerator_interconnect 'XGMI', accelerator_host_interconnect 'PCIe Gen5 x16', system_name 'Supermicro AS-8125GS-TNMR2'. The 000 status in a plain HTTP probe is GitHub raw's bot behaviour under concurrency, not a dead link; notes were empty and are now filled with the verified fields.

Cross-references

This record references no other records.