MiniMax M2.5/M2.7 — B300 vs MI355X
Head-to-head AI inference benchmark comparison of B300 (NVIDIA Blackwell) and MI355X (AMD CDNA 4) 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 / MI355X on MiniMax M2.5/M2.7 at 42 tok/s/user: 7350 / 3690 tok/s/GPU, $0.09 / $0.11 per million tokens. B300 is 24% cheaper per token; B300 delivers 99% more tok/s/GPU.
Around the middle of the 16–119 tok/s/user interactivity band, at 68 tok/s/user on MiniMax M2.5/M2.7: B300 runs 2609 tok/s/GPU at $0.26/M tokens, MI355X runs 1158 at $0.36/M. B300 is 40% cheaper per token; B300 delivers 125% more tok/s/GPU.
Setting 93 tok/s/user as the target on MiniMax M2.5/M2.7, B300 produces 1168 tok/s/GPU ($0.55 per million tokens) and MI355X produces 568 ($0.70). B300 is 29% cheaper per token; B300 delivers 106% more tok/s/GPU. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | B300:7350.2MI355X:3689.9 | B300:2609.5MI355X:1158.1 | B300:1168.3MI355X:567.7 |
| Cost ($/M tok) | B300:$0.089MI355X:$0.111 | B300:$0.255MI355X:$0.357 | B300:$0.545MI355X:$0.703 |
| tok/s/MW | B300:3387173MI355X:1392418 | B300:1202515MI355X:437012 | B300:538366MI355X:214237 |
| Concurrency | B300:~874MI355X:~256 | B300:~76MI355X:~17 | B300:~25MI355X:~6 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.