Qwen 3.5 397B-A17B — H100 vs RTX PRO 6000
Head-to-head AI inference benchmark comparison of H100 (NVIDIA Hopper) and RTX PRO 6000 (NVIDIA Blackwell) 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.
H100: 433 tok/s/GPU, $0.85 per million tokens at 64 tok/s/user on Qwen 3.5 397B-A17B. RTX PRO 6000 is unmeasured here.
At 100 tok/s/user on Qwen 3.5 397B-A17B, H100 delivers 311 tok/s/GPU at $1.15 per million tokens; RTX PRO 6000 hasn't been benchmarked at this target.
H100 hits 236 tok/s/GPU for $1.54 per million tokens at 135 tok/s/user on Qwen 3.5 397B-A17B. No RTX PRO 6000 data at this operating point. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | H100:433.3RTX PRO 6000:— | H100:310.5RTX PRO 6000:— | H100:236.1RTX PRO 6000:— |
| Cost ($/M tok) | H100:$0.848RTX PRO 6000:— | H100:$1.151RTX PRO 6000:— | H100:$1.536RTX PRO 6000:— |
| tok/s/MW | H100:316265RTX PRO 6000:— | H100:226676RTX PRO 6000:— | H100:172330RTX PRO 6000:— |
| Concurrency | H100:~28RTX PRO 6000:— | H100:~13RTX PRO 6000:— | H100:~7RTX PRO 6000:— |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.