Qwen 3.5 397B-A17B · Chip comparison
Qwen 3.5 397B-A17B — GB300 NVL72 vs H200
Head-to-head AI inference benchmark comparison of GB300 NVL72 (NVIDIA Blackwell) and H200 (NVIDIA Hopper) 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 →
Near the low end of the 68–359 tok/s/user interactivity band, at 141 tok/s/user on Qwen 3.5 397B-A17B: GB300 NVL72 runs 59,758 tok/s/chip at $0.011/M tokens, H200 runs 16,049 at $0.021/M. GB300 NVL72 is 97% cheaper per token; GB300 NVL72 delivers 272% more tok/s/chip.
Setting 214 tok/s/user as the target on Qwen 3.5 397B-A17B, GB300 NVL72 produces 33,238 tok/s/chip ($0.019 per million tokens) and H200 produces 9,601 ($0.035). GB300 NVL72 is 83% cheaper per token; GB300 NVL72 delivers 246% more tok/s/chip.
At 287 tok/s/user interactivity on Qwen 3.5 397B-A17B, GB300 NVL72 delivers 24,883 tok/s/chip at $0.026 per million tokens; H200 delivers 5,556 tok/s/chip at $0.061. GB300 NVL72 is 137% cheaper per token; GB300 NVL72 delivers 348% more tok/s/chip at this point. (Numbers reflect the default agentic-traces · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB300 NVL7259,757.7H20016,048.7 | GB300 NVL7233,237.9H2009,600.7 | GB300 NVL7224,882.6H2005,555.6 |
| Cost ($/M tok) | GB300 NVL72$0.011H200$0.021 | GB300 NVL72$0.019H200$0.035 | GB300 NVL72$0.026H200$0.061 |
| tok/s/MW | GB300 NVL7228,187,615H20011,714,364 | GB300 NVL7215,678,249H2007,007,840 | GB300 NVL7211,737,088H2004,055,168 |
| Concurrency | GB300 NVL72~392H200~22 | GB300 NVL72~159H200~12 | GB300 NVL72~113H200~8 |