Head-to-head AI inference benchmark comparison of MI355X (AMD CDNA 4) and Vera Rubin NVL72 (NVIDIA Vera Rubin) on DeepSeekv4 Pro 0813 1.6T. 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.
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 β
Setting 118 tok/s/user as the target on DeepSeekv4 Pro 0813 1.6T, MI355X produces 23523 tok/s/chip ($0.02 per million tokens) and Vera Rubin NVL72 produces 267359 ($0.00). Vera Rubin NVL72 is 372% cheaper per token; Vera Rubin NVL72 delivers 1037% more tok/s/chip.
At 161 tok/s/user interactivity on DeepSeekv4 Pro 0813 1.6T, MI355X delivers 15902 tok/s/chip at $0.03 per million tokens; Vera Rubin NVL72 delivers 204966 tok/s/chip at $0.00. Vera Rubin NVL72 is 436% cheaper per token; Vera Rubin NVL72 delivers 1189% more tok/s/chip at this point.
MI355X posts 5483 tok/s/chip for $0.08 per million tokens at 205 tok/s/user on DeepSeekv4 Pro 0813 1.6T; Vera Rubin NVL72 posts 77326 tok/s/chip for $0.01. Vera Rubin NVL72 is 486% cheaper per token; Vera Rubin NVL72 delivers 1310% more tok/s/chip. (Numbers reflect the default agentic-traces Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | MI355X:23523.4Vera Rubin NVL72:267359.4 | MI355X:15902.0Vera Rubin NVL72:204966.4 | MI355X:5482.8Vera Rubin NVL72:77326.2 |
| Cost ($/M tok) | MI355X:$0.018Vera Rubin NVL72:$0.004 | MI355X:$0.026Vera Rubin NVL72:$0.005 | MI355X:$0.076Vera Rubin NVL72:$0.013 |
| tok/s/MW | MI355X:11255227Vera Rubin NVL72:81018005 | MI355X:7608605Vera Rubin NVL72:62111039 | MI355X:2623372Vera Rubin NVL72:23432169 |
| Concurrency | MI355X:~48Vera Rubin NVL72:~2854 | MI355X:~28Vera Rubin NVL72:~2190 | MI355X:~9Vera Rubin NVL72:~625 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
