DeepSeek V4 Pro 1.6T — B300 vs MI355X Performance per Dollar
Cost per million tokens of B300 (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.
Push DeepSeek V4 Pro 1.6T to 32 tok/s/user and B300 lands at $0.12 per million tokens against MI355X's $0.14 — B300 pulls ahead by 21%.
B300: $0.21 per million tokens. MI355X: $0.38. Both at 62 tok/s/user on DeepSeek V4 Pro 1.6T, with B300 84% cheaper.
Toward the upper edge of the 4–120 tok/s/user interactivity band — at 91 tok/s/user — B300 runs $0.35 per million tokens on DeepSeek V4 Pro 1.6T while MI355X runs $0.70. B300 is the cheaper choice by 97%. (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): B300 $2.34/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 | B300:$0.120MI355X:$0.145 | B300:$0.208MI355X:$0.384 | B300:$0.354MI355X:$0.699 |
| Concurrency | B300:~1024MI355X:~101 | B300:~76MI355X:~16 | B300:~11MI355X:~7 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.