MiniMax M2.5/M2.7 — B300 vs MI300X
Head-to-head AI inference benchmark comparison of B300 (NVIDIA Blackwell) and MI300X (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.
B300 posts 5802 tok/s/chip for $0.11 per million tokens at 44 tok/s/user on MiniMax M2.5/M2.7; MI300X posts 1249 tok/s/chip for $0.21. B300 is 95% cheaper per token; B300 delivers 365% more tok/s/chip.
Throughput at 61 tok/s/user on MiniMax M2.5/M2.7: B300 hits 3387 tok/s/chip, MI300X hits 779. Per-million costs land at $0.19 and $0.34 respectively. B300 is 84% cheaper per token; B300 delivers 335% more tok/s/chip.
B300 / MI300X on MiniMax M2.5/M2.7 at 78 tok/s/user: 1568 / 412 tok/s/chip, $0.41 / $0.63 per million tokens. B300 is 56% cheaper per token; B300 delivers 281% more tok/s/chip. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B300:5802.1MI300X:1248.6 | B300:3387.5MI300X:778.8 | B300:1568.3MI300X:411.9 |
| Cost ($/M tok) | B300:$0.109MI300X:$0.212 | B300:$0.185MI300X:$0.341 | B300:$0.407MI300X:$0.633 |
| tok/s/MW | B300:3053720MI300X:898287 | B300:1782877MI300X:560298 | B300:825422MI300X:296322 |
| Concurrency | B300:~303MI300X:~63 | B300:~232MI300X:~26 | B300:~47MI300X:~11 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.