DeepSeek R1 — B300 vs MI355X Performance per Dollar
Cost per million tokens of B300 (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.
B300: $0.44 per million tokens. MI355X: $0.76. Both at 65 tok/s/user on DeepSeek R1, with B300 74% cheaper.
Around the middle of the 16–216 tok/s/user interactivity band — at 116 tok/s/user — B300 runs $1.60 per million tokens on DeepSeek R1 while MI355X runs $2.01. B300 is the cheaper choice by 26%.
On DeepSeek R1 at 166 tok/s/user, the per-million math comes out to $3.07 for B300 and $2.95 for MI355X; MI355X delivers 4% more output per dollar. (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.)
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:$0.435MI355X:$0.756 | B300:$1.595MI355X:$2.006 | B300:$3.069MI355X:$2.949 |
| Concurrency | B300:~624MI355X:~41 | B300:~159MI355X:~8 | B300:~5MI355X:~4 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.