DeepSeek R1 — H100 vs MI355X
Head-to-head AI inference benchmark comparison of H100 (NVIDIA Hopper) and MI355X (AMD CDNA 4) 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.
Near the low end of the 17–128 tok/s/user interactivity band, at 44 tok/s/user on DeepSeek R1: H100 runs 259 tok/s/GPU at $1.40/M tokens, MI355X runs 1526 at $0.27/M. MI355X is 419% cheaper per token; MI355X delivers 490% more tok/s/GPU.
Setting 72 tok/s/user as the target on DeepSeek R1, H100 produces 70 tok/s/GPU ($5.06 per million tokens) and MI355X produces 744 ($0.54). MI355X is 842% cheaper per token; MI355X delivers 965% more tok/s/GPU.
At 100 tok/s/user interactivity on DeepSeek R1, H100 delivers 23 tok/s/GPU at $15.7 per million tokens; MI355X delivers 434 tok/s/GPU at $0.95. MI355X is 1557% cheaper per token; MI355X delivers 1765% more tok/s/GPU at this point. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | H100:258.7MI355X:1526.1 | H100:69.9MI355X:744.1 | H100:23.3MI355X:433.9 |
| Cost ($/M tok) | H100:$1.399MI355X:$0.270 | H100:$5.061MI355X:$0.537 | H100:$15.692MI355X:$0.947 |
| tok/s/MW | H100:149557MI355X:575898 | H100:40376MI355X:280795 | H100:13443MI355X:163718 |
| Concurrency | H100:~495MI355X:~503 | H100:~48MI355X:~68 | H100:~8MI355X:~17 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.