MiniMax M2.5/M2.7 — GB200 NVL72 vs MI325X
Head-to-head AI inference benchmark comparison of GB200 NVL72 (NVIDIA Blackwell) and MI325X (AMD CDNA 3) on MiniMax M2.5/M2.7. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
Setting 46 tok/s/user as the target on MiniMax M2.5/M2.7, GB200 NVL72 produces 4331 tok/s/GPU ($0.14 per million tokens) and MI325X produces 1643 ($0.22). GB200 NVL72 is 53% cheaper per token; GB200 NVL72 delivers 164% more tok/s/GPU.
At 65 tok/s/user interactivity on MiniMax M2.5/M2.7, GB200 NVL72 delivers 2884 tok/s/GPU at $0.21 per million tokens; MI325X delivers 757 tok/s/GPU at $0.48. GB200 NVL72 is 124% cheaper per token; GB200 NVL72 delivers 281% more tok/s/GPU at this point.
GB200 NVL72 posts 1108 tok/s/GPU for $0.56 per million tokens at 85 tok/s/user on MiniMax M2.5/M2.7; MI325X posts 325 tok/s/GPU for $1.09. GB200 NVL72 is 94% cheaper per token; GB200 NVL72 delivers 241% more tok/s/GPU. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | GB200 NVL72:4330.9MI325X:1642.6 | GB200 NVL72:2883.9MI325X:757.2 | GB200 NVL72:1107.6MI325X:324.7 |
| Cost ($/M tok) | GB200 NVL72:$0.142MI325X:$0.217 | GB200 NVL72:$0.214MI325X:$0.479 | GB200 NVL72:$0.563MI325X:$1.094 |
| tok/s/MW | GB200 NVL72:2062316MI325X:753479 | GB200 NVL72:1373303MI325X:347355 | GB200 NVL72:527426MI325X:148934 |
| Concurrency | GB200 NVL72:~512MI325X:~147 | GB200 NVL72:~235MI325X:~38 | GB200 NVL72:~64MI325X:~16 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.