Qwen 3.5 397B-A17B — B200 vs MI355X Performance per Dollar
Cost per million tokens of B200 (NVIDIA Blackwell) versus MI355X (AMD CDNA 4) on Qwen 3.5 397B-A17B. Owning-hyperscaler TCO normalized by output tokens — performance per dollar across LLM workloads. Pick the more cost-efficient SKU at every target interactivity level. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
B200 edges MI355X at 63 tok/s/user on Qwen 3.5 397B-A17B — $0.05 per million tokens versus $0.08, a 50% cost-per-token gap.
Push Qwen 3.5 397B-A17B to 116 tok/s/user and B200 lands at $0.08 per million tokens against MI355X's $0.13 — B200 pulls ahead by 63%.
B200: $0.11 per million tokens. MI355X: $0.18. Both at 169 tok/s/user on Qwen 3.5 397B-A17B, with B200 73% cheaper. (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.)
Chip pricing (owning hyperscaler): B200 $1.73/chip/hr · MI355X $1.50/chip/hr. Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Dollar per Million Tokens | B200:$0.054MI355X:$0.081 | B200:$0.078MI355X:$0.127 | B200:$0.107MI355X:$0.184 |
| Concurrency | B200:~68MI355X:~39 | B200:~26MI355X:~14 | B200:~13MI355X:~7 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.