Head-to-head AI inference benchmark comparison of GB200 NVL72 (NVIDIA Blackwell) and MI355X (AMD CDNA 4) on Muse Spark Open 816B. 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
At 89 tok/s/user interactivity on Muse Spark Open 816B, GB200 NVL72 delivers 23657 tok/s/chip at $0.02 per million tokens; MI355X delivers 6964 tok/s/chip at $0.06. GB200 NVL72 is 174% cheaper per token; GB200 NVL72 delivers 240% more tok/s/chip at this point.
GB200 NVL72 posts 18444 tok/s/chip for $0.03 per million tokens at 121 tok/s/user on Muse Spark Open 816B; MI355X posts 4313 tok/s/chip for $0.10. GB200 NVL72 is 245% cheaper per token; GB200 NVL72 delivers 328% more tok/s/chip.
Throughput at 152 tok/s/user on Muse Spark Open 816B: GB200 NVL72 hits 15731 tok/s/chip, MI355X hits 2630. Per-million costs land at $0.03 and $0.16 respectively. GB200 NVL72 is 382% cheaper per token; GB200 NVL72 delivers 498% more tok/s/chip. (Numbers reflect the default agentic-traces Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB200 NVL72:23657.2MI355X:6963.7 | GB200 NVL72:18443.7MI355X:4313.2 | GB200 NVL72:15731.4MI355X:2630.4 |
| Cost ($/M tok) | GB200 NVL72:$0.022MI355X:$0.060 | GB200 NVL72:$0.028MI355X:$0.097 | GB200 NVL72:$0.033MI355X:$0.158 |
| tok/s/MW | GB200 NVL72:12650890MI355X:3331929 | GB200 NVL72:9862918MI355X:2063713 | GB200 NVL72:8412523MI355X:1258543 |
| Concurrency | GB200 NVL72:~18MI355X:~12 | GB200 NVL72:~13MI355X:~7 | GB200 NVL72:~11MI355X:~3 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
No measurements to plot for this selection. Review the benchmark controls above or adjust quick filters.
