Qwen 3.5 397B-A17B — GB200 NVL72 vs MI355X
Head-to-head AI inference benchmark comparison of GB200 NVL72 (NVIDIA Blackwell) 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 90 tok/s/user interactivity on Qwen 3.5 397B-A17B, GB200 NVL72 delivers 10277 tok/s/chip at $0.05 per million tokens; MI355X delivers 4041 tok/s/chip at $0.10. GB200 NVL72 is 107% cheaper per token; GB200 NVL72 delivers 154% more tok/s/chip at this point.
GB200 NVL72 posts 6484 tok/s/chip for $0.08 per million tokens at 134 tok/s/user on Qwen 3.5 397B-A17B; MI355X posts 2880 tok/s/chip for $0.14. GB200 NVL72 is 74% cheaper per token; GB200 NVL72 delivers 125% more tok/s/chip.
Throughput at 178 tok/s/user on Qwen 3.5 397B-A17B: GB200 NVL72 hits 3613 tok/s/chip, MI355X hits 2166. Per-million costs land at $0.14 and $0.20 respectively. GB200 NVL72 is 36% cheaper per token; GB200 NVL72 delivers 67% more tok/s/chip. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB200 NVL72:10277.2MI355X:4040.6 | GB200 NVL72:6484.1MI355X:2880.0 | GB200 NVL72:3612.9MI355X:2166.3 |
| Cost ($/M tok) | GB200 NVL72:$0.050MI355X:$0.103 | GB200 NVL72:$0.082MI355X:$0.143 | GB200 NVL72:$0.143MI355X:$0.195 |
| tok/s/MW | GB200 NVL72:5495811MI355X:1933305 | GB200 NVL72:3467444MI355X:1377992 | GB200 NVL72:1932024MI355X:1036510 |
| Concurrency | GB200 NVL72:~1070MI355X:~22 | GB200 NVL72:~279MI355X:~11 | GB200 NVL72:~45MI355X:~6 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.