Qwen 3.5 397B-A17B — H200 vs MI300X
Head-to-head AI inference benchmark comparison of H200 (NVIDIA Hopper) and MI300X (AMD CDNA 3) on Qwen 3.5 397B-A17B. 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 43 tok/s/user on Qwen 3.5 397B-A17B: H200 hits 626 tok/s/chip, MI300X hits 346. Per-million costs land at $0.54 and $0.76 respectively. H200 is 41% cheaper per token; H200 delivers 81% more tok/s/chip.
H200 / MI300X on Qwen 3.5 397B-A17B at 53 tok/s/user: 558 / 190 tok/s/chip, $0.61 / $1.39 per million tokens. H200 is 129% cheaper per token; H200 delivers 195% more tok/s/chip.
Toward the upper edge of the 34–71 tok/s/user interactivity band, at 62 tok/s/user on Qwen 3.5 397B-A17B: H200 runs 520 tok/s/chip at $0.65/M tokens, MI300X runs 123 at $2.15/M. H200 is 230% cheaper per token; H200 delivers 324% more tok/s/chip. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | H200:626.3MI300X:345.8 | H200:558.3MI300X:189.5 | H200:520.1MI300X:122.7 |
| Cost ($/M tok) | H200:$0.541MI300X:$0.763 | H200:$0.607MI300X:$1.392 | H200:$0.652MI300X:$2.150 |
| tok/s/MW | H200:457154MI300X:248748 | H200:407549MI300X:136357 | H200:379613MI300X:88306 |
| Concurrency | H200:~58MI300X:~34 | H200:~42MI300X:~15 | H200:~34MI300X:~8 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.