Qwen 3.5 397B-A17B — B200 vs MI300X
Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) 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.
Near the low end of the 34–71 tok/s/user interactivity band, at 43 tok/s/user on Qwen 3.5 397B-A17B: B200 runs 4781 tok/s/GPU at $0.11/M tokens, MI300X runs 346 at $0.89/M. B200 is 688% cheaper per token; B200 delivers 1283% more tok/s/GPU.
Setting 53 tok/s/user as the target on Qwen 3.5 397B-A17B, B200 produces 4181 tok/s/GPU ($0.13 per million tokens) and MI300X produces 190 ($1.63). B200 is 1168% cheaper per token; B200 delivers 2106% more tok/s/GPU.
At 62 tok/s/user interactivity on Qwen 3.5 397B-A17B, B200 delivers 3647 tok/s/GPU at $0.15 per million tokens; MI300X delivers 123 tok/s/GPU at $2.52. B200 is 1595% cheaper per token; B200 delivers 2872% more tok/s/GPU at this point. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | B200:4780.6MI300X:345.8 | B200:4180.6MI300X:189.5 | B200:3647.5MI300X:122.7 |
| Cost ($/M tok) | B200:$0.113MI300X:$0.893 | B200:$0.129MI300X:$1.633 | B200:$0.148MI300X:$2.517 |
| tok/s/MW | B200:2203062MI300X:193162 | B200:1926539MI300X:105886 | B200:1680869MI300X:68573 |
| Concurrency | B200:~245MI300X:~34 | B200:~179MI300X:~15 | B200:~129MI300X:~8 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.