Qwen 3.5 397B-A17B — H200 vs MI355X
Head-to-head AI inference benchmark comparison of H200 (NVIDIA Hopper) and MI355X (AMD CDNA 4) 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.
At 53 tok/s/user interactivity on Qwen 3.5 397B-A17B, H200 delivers 476 tok/s/GPU at $0.82 per million tokens; MI355X delivers 2454 tok/s/GPU at $0.17. MI355X is 393% cheaper per token; MI355X delivers 416% more tok/s/GPU at this point.
H200 posts 323 tok/s/GPU for $1.22 per million tokens at 80 tok/s/user on Qwen 3.5 397B-A17B; MI355X posts 1653 tok/s/GPU for $0.25. MI355X is 381% cheaper per token; MI355X delivers 412% more tok/s/GPU.
Throughput at 106 tok/s/user on Qwen 3.5 397B-A17B: H200 hits 272 tok/s/GPU, MI355X hits 1162. Per-million costs land at $1.44 and $0.35 respectively. MI355X is 317% cheaper per token; MI355X delivers 326% more tok/s/GPU. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | H200:475.6MI355X:2453.8 | H200:322.8MI355X:1653.3 | H200:272.5MI355X:1161.8 |
| Cost ($/M tok) | H200:$0.821MI355X:$0.167 | H200:$1.216MI355X:$0.253 | H200:$1.438MI355X:$0.345 |
| tok/s/MW | H200:274884MI355X:925957 | H200:186573MI355X:623872 | H200:157493MI355X:438424 |
| Concurrency | H200:~37MI355X:~90 | H200:~17MI355X:~42 | H200:~12MI355X:~22 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.