Head-to-head AI inference benchmark comparison of MI355X (AMD CDNA 4) and Vera Rubin NVL72 (NVIDIA Vera Rubin) on MiniMax M3 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 →
At 194 tok/s/user interactivity on MiniMax M3 428B, MI355X delivers 30882 tok/s/chip at $0.01 per million tokens; Vera Rubin NVL72 delivers 140527 tok/s/chip at $0.01. Vera Rubin NVL72 is 89% cheaper per token; Vera Rubin NVL72 delivers 355% more tok/s/chip at this point.
MI355X posts 16561 tok/s/chip for $0.03 per million tokens at 286 tok/s/user on MiniMax M3 428B; Vera Rubin NVL72 posts 90654 tok/s/chip for $0.01. Vera Rubin NVL72 is 127% cheaper per token; Vera Rubin NVL72 delivers 447% more tok/s/chip.
Throughput at 377 tok/s/user on MiniMax M3 428B: MI355X hits 9082 tok/s/chip, Vera Rubin NVL72 hits 50731. Per-million costs land at $0.05 and $0.02 respectively. Vera Rubin NVL72 is 132% cheaper per token; Vera Rubin NVL72 delivers 459% more tok/s/chip. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | MI355X:30881.7Vera Rubin NVL72:140526.8 | MI355X:16561.3Vera Rubin NVL72:90653.6 | MI355X:9082.3Vera Rubin NVL72:50730.8 |
| Cost ($/M tok) | MI355X:$0.013Vera Rubin NVL72:$0.007 | MI355X:$0.025Vera Rubin NVL72:$0.011 | MI355X:$0.046Vera Rubin NVL72:$0.020 |
| tok/s/MW | MI355X:14775924Vera Rubin NVL72:42583889 | MI355X:7924084Vera Rubin NVL72:27470785 | MI355X:4345617Vera Rubin NVL72:15372978 |
| Concurrency | MI355X:~18Vera Rubin NVL72:~42 | MI355X:~9Vera Rubin NVL72:~26 | MI355X:~4Vera Rubin NVL72:~16 |
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 • Drag to pan • Double-click to reset • 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 M3 428B 8K / 1K Total Tokens per $1 TCO vs. Median Interactivity