DeepSeek V4 Pro 1.6T — B200 vs MI355X Performance per Dollar
Cost per million tokens of B200 (NVIDIA Blackwell) versus MI355X (AMD CDNA 4) on DeepSeek V4 Pro 1.6T. 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: $0.39 per million tokens. MI355X: $0.45. Both at 34 tok/s/user on DeepSeek V4 Pro 1.6T, with B200 17% cheaper.
Around the middle of the 4–127 tok/s/user interactivity band — at 65 tok/s/user — B200 runs $1.45 per million tokens on DeepSeek V4 Pro 1.6T while MI355X runs $1.22. MI355X is the cheaper choice by 19%.
On DeepSeek V4 Pro 1.6T at 97 tok/s/user, the per-million math comes out to $2.90 for B200 and $2.03 for MI355X; MI355X delivers 43% more output per dollar. (Numbers reflect the default 1k/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:$0.387MI355X:$0.451 | B200:$1.454MI355X:$1.224 | B200:$2.895MI355X:$2.032 |
| Concurrency | B200:~184MI355X:~126 | B200:~24MI355X:~23 | B200:~9MI355X:~9 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.