Qwen 3.5 397B-A17B — GB200 NVL72 vs MI355X
Head-to-head AI inference benchmark comparison of GB200 NVL72 (NVIDIA Blackwell) and MI355X (AMD CDNA 4) on Qwen 3.5 397B-A17B. 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 101 tok/s/user on Qwen 3.5 397B-A17B, MI355X delivers 38865 tok/s/chip at $0.01 per million tokens; GB200 NVL72 hasn't been benchmarked at this target.
MI355X hits 20776 tok/s/chip for $0.02 per million tokens at 172 tok/s/user on Qwen 3.5 397B-A17B. No GB200 NVL72 data at this operating point.
MI355X: 6192 tok/s/chip, $0.07 per million tokens at 243 tok/s/user on Qwen 3.5 397B-A17B. GB200 NVL72 is unmeasured here. (Numbers reflect the default agentic-traces · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB200 NVL72:—MI355X:38865.3 | GB200 NVL72:—MI355X:20776.2 | GB200 NVL72:—MI355X:6191.8 |
| Cost ($/M tok) | GB200 NVL72:—MI355X:$0.011 | GB200 NVL72:—MI355X:$0.020 | GB200 NVL72:—MI355X:$0.067 |
| tok/s/MW | GB200 NVL72:—MI355X:18595855 | GB200 NVL72:—MI355X:9940742 | GB200 NVL72:—MI355X:2962575 |
| Concurrency | GB200 NVL72:—MI355X:~33 | GB200 NVL72:—MI355X:~15 | GB200 NVL72:—MI355X:~4 |