MiniMax M3.1 Flash 428B · Chip comparison
MiniMax M3.1 Flash 428B — GB300 NVL72 vs MI355X
Head-to-head AI inference benchmark comparison of GB300 NVL72 (NVIDIA Blackwell) and MI355X (AMD CDNA 4) on MiniMax M3.1 Flash 428B. 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.
AgentX replays real coding-agent sessions rather than fixed-length prompts, so context grows turn over turn and most of each request is served from cache instead of being recomputed. That turns the comparison into a systems question: KV transfer between nodes, prefix-aware routing, and cache capacity all move the curve alongside raw chip throughput. Fixed-sequence workloads stay the clean baseline for kernel and silicon performance, so the two scenarios answer different questions about the same hardware. Learn more about AgentX →
GB300 NVL72 posts 121,197 tok/s/chip for $0.005 per million tokens at 200 tok/s/user on MiniMax M3.1 Flash 428B; MI355X posts 80,133 tok/s/chip for $0.005. MI355X is 2% cheaper per token; GB300 NVL72 delivers 51% more tok/s/chip.
Throughput at 340 tok/s/user on MiniMax M3.1 Flash 428B: GB300 NVL72 hits 27,085 tok/s/chip, MI355X hits 36,632. Costs per million tokens land at $0.024 and $0.011 respectively. MI355X is 52% cheaper per token; MI355X delivers 35% more tok/s/chip.
GB300 NVL72 / MI355X on MiniMax M3.1 Flash 428B at 481 tok/s/user: 9,668 / 17,494 tok/s/chip, $0.066 / $0.024 per million tokens. MI355X is 64% cheaper per token; MI355X delivers 81% more tok/s/chip. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB300 NVL72121,197MI355X80,133 | GB300 NVL7227,085MI355X36,632 | GB300 NVL729,668MI355X17,494 |
| Cost ($/M tok) | GB300 NVL72$0.005MI355X$0.005 | GB300 NVL72$0.024MI355X$0.011 | GB300 NVL72$0.066MI355X$0.024 |
| tok/s/MW | GB300 NVL7257,168,182MI355X38,341,359 | GB300 NVL7212,776,009MI355X17,527,503 | GB300 NVL724,560,525MI355X8,370,504 |
| Concurrency | GB300 NVL72~416MI355X~42 | GB300 NVL72~15MI355X~19 | GB300 NVL72~4MI355X~8 |