Qwen 3.5 397B-A17B — H100 vs MI325X
Head-to-head AI inference benchmark comparison of H100 (NVIDIA Hopper) and MI325X (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 46 tok/s/user on Qwen 3.5 397B-A17B: H100 hits 886 tok/s/GPU, MI325X hits 403. Per-million costs land at $0.40 and $0.86 respectively. H100 is 114% cheaper per token; H100 delivers 120% more tok/s/GPU.
H100 / MI325X on Qwen 3.5 397B-A17B at 55 tok/s/user: 680 / 219 tok/s/GPU, $0.54 / $1.63 per million tokens. H100 is 203% cheaper per token; H100 delivers 210% more tok/s/GPU.
Toward the upper edge of the 37–72 tok/s/user interactivity band, at 64 tok/s/user on Qwen 3.5 397B-A17B: H100 runs 480 tok/s/GPU at $0.77/M tokens, MI325X runs 134 at $2.64/M. H100 is 244% cheaper per token; H100 delivers 257% more tok/s/GPU. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | H100:885.8MI325X:402.9 | H100:680.4MI325X:219.4 | H100:479.6MI325X:134.3 |
| Cost ($/M tok) | H100:$0.405MI325X:$0.864 | H100:$0.538MI325X:$1.631 | H100:$0.768MI325X:$2.638 |
| tok/s/MW | H100:512039MI325X:184830 | H100:393293MI325X:100633 | H100:277201MI325X:61591 |
| Concurrency | H100:~87MI325X:~37 | H100:~56MI325X:~16 | H100:~35MI325X:~9 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.