DeepSeek R1 — MI355X vs Vera Rubin NVL72
Head-to-head AI inference benchmark comparison of MI355X (AMD CDNA 4) and Vera Rubin NVL72 (NVIDIA Vera Rubin) on DeepSeek R1. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
Throughput at 127 tok/s/user on DeepSeek R1: MI355X hits 4407 tok/s/chip, Vera Rubin NVL72 hits 0. Per-million costs land at $0.09 and $0.00 respectively. Comparable data is unavailable.
MI355X / Vera Rubin NVL72 on DeepSeek R1 at 201 tok/s/user: 910 / 0 tok/s/chip, $0.46 / $0.00 per million tokens. Comparable data is unavailable.
Toward the upper edge of the 52–350 tok/s/user interactivity band, at 276 tok/s/user on DeepSeek R1: MI355X runs 517 tok/s/chip at $0.81/M tokens, Vera Rubin NVL72 runs 0 at $0.00/M. Comparable data is unavailable. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | MI355X:4407.4Vera Rubin NVL72:0.0 | MI355X:910.4Vera Rubin NVL72:0.0 | MI355X:516.6Vera Rubin NVL72:0.0 |
| Cost ($/M tok) | MI355X:$0.095Vera Rubin NVL72:$0.000 | MI355X:$0.458Vera Rubin NVL72:$0.000 | MI355X:$0.806Vera Rubin NVL72:$0.000 |
| tok/s/MW | MI355X:2108788Vera Rubin NVL72:0 | MI355X:435589Vera Rubin NVL72:0 | MI355X:247197Vera Rubin NVL72:0 |
| Concurrency | MI355X:~77Vera Rubin NVL72:~400 | MI355X:~7Vera Rubin NVL72:~400 | MI355X:~5Vera Rubin NVL72:~400 |