MiniMax M2.5/M2.7 — GB300 NVL72 vs MI355X Performance per Dollar
Cost per million tokens of GB300 NVL72 (NVIDIA Blackwell) versus MI355X (AMD CDNA 4) on MiniMax M2.5/M2.7. 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 MiniMax M2.5/M2.7 to 47 tok/s/user and GB300 NVL72 lands at $0.14 per million tokens against MI355X's $0.12 — MI355X pulls ahead by 17%.
GB300 NVL72: $0.31 per million tokens. MI355X: $0.38. Both at 71 tok/s/user on MiniMax M2.5/M2.7, with GB300 NVL72 22% cheaper.
Toward the upper edge of the 24–119 tok/s/user interactivity band — at 95 tok/s/user — GB300 NVL72 runs $0.82 per million tokens on MiniMax M2.5/M2.7 while MI355X runs $0.73. MI355X is the cheaper choice by 13%. (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): GB300 NVL72 $2.65/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 | GB300 NVL72:$0.136MI355X:$0.117 | GB300 NVL72:$0.313MI355X:$0.383 | GB300 NVL72:$0.823MI355X:$0.732 |
| Concurrency | GB300 NVL72:~540MI355X:~256 | GB300 NVL72:~128MI355X:~32 | GB300 NVL72:~32MI355X:~6 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.