DeepSeek R1 — B200 vs MI355X Performance per Dollar
Cost per million tokens of B200 (NVIDIA Blackwell) versus MI355X (AMD CDNA 4) on DeepSeek R1. 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.
Push DeepSeek R1 to 83 tok/s/user and B200 lands at $0.06 per million tokens against MI355X's $0.05 — MI355X pulls ahead by 22%.
B200: $0.26 per million tokens. MI355X: $0.18. Both at 156 tok/s/user on DeepSeek R1, with MI355X 42% cheaper.
Toward the upper edge of the 11–300 tok/s/user interactivity band — at 228 tok/s/user — B200 runs $0.54 per million tokens on DeepSeek R1 while MI355X runs $0.69. B200 is the cheaper choice by 27%. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
GPU pricing (owning hyperscaler): B200 $1.95/GPU/hr · MI355X $1.48/GPU/hr. Source: SemiAnalysis Market August 2025 Pricing Surveys & AI Cloud TCO Model.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Dollar per Million Tokens | B200:$0.063MI355X:$0.052 | B200:$0.262MI355X:$0.185 | B200:$0.543MI355X:$0.691 |
| Concurrency | B200:~626MI355X:~283 | B200:~63MI355X:~64 | B200:~3MI355X:~8 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.