Head-to-head AI inference benchmark comparison of B200 (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 β same interactions as the main inference chart.
Near the low end of the 14β110 tok/s/user interactivity band, at 37 tok/s/user on MiniMax M2.5/M2.7: B200 runs 10432 tok/s/chip at $0.05/M tokens, MI355X runs 5356 at $0.08/M. B200 is 69% cheaper per token; B200 delivers 95% more tok/s/chip.
Setting 62 tok/s/user as the target on MiniMax M2.5/M2.7, B200 produces 5345 tok/s/chip ($0.09 per million tokens) and MI355X produces 3138 ($0.13). B200 is 48% cheaper per token; B200 delivers 70% more tok/s/chip.
At 86 tok/s/user interactivity on MiniMax M2.5/M2.7, B200 delivers 3213 tok/s/chip at $0.15 per million tokens; MI355X delivers 1954 tok/s/chip at $0.21. B200 is 43% cheaper per token; B200 delivers 64% more tok/s/chip at this point. (Numbers reflect the default 8k/1k Β· fp8 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B200:10432.3MI355X:5355.6 | B200:5345.3MI355X:3138.1 | B200:3212.7MI355X:1953.9 |
| Cost ($/M tok) | B200:$0.046MI355X:$0.078 | B200:$0.090MI355X:$0.133 | B200:$0.150MI355X:$0.213 |
| tok/s/MW | B200:6100788MI355X:2562481 | B200:3125935MI355X:1501464 | B200:1878801MI355X:934901 |
| Concurrency | B200:~777MI355X:~67 | B200:~62MI355X:~11 | B200:~25MI355X:~5 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
TCO $/chip/hr:
Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate cost per million tokens per decode chip or per prefill chip, rather than per total chip count. This makes direct token cost comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate input throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct input throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate output throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct output throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate power per decode chip or per prefill chip, rather than per total chip count. This makes direct power comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate Joules per decode chip or per prefill chip, rather than per total chip count. This makes direct Joules per token comparison with aggregated configs not an apples-to-apples comparison.
Matching measurements exist, but their chip series are hidden.
Shift+Scroll to zoom β’ Drag to pan β’ Double-click to reset β’ Click a point to pin tooltip
1 The ATOM engine is promising, however it has yet to serve production tokens. It is still in its infant stage.
MiniMax M2.5/2.7 230B 8K / 1K Total Tokens per $1 TCO vs. Interactivity