Qwen 3.5 397B-A17B — GB200 NVL72 vs H100 Performance per Dollar
Cost per million tokens of GB200 NVL72 (NVIDIA Blackwell) versus H100 (NVIDIA Hopper) on Qwen 3.5 397B-A17B. Owning-hyperscaler TCO normalized by output tokens — performance per dollar across LLM workloads. Pick the more cost-efficient SKU at every target interactivity level. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
GB200 NVL72: $0.49 per million tokens. H100: $0.73. Both at 63 tok/s/user on Qwen 3.5 397B-A17B, with GB200 NVL72 48% cheaper.
Around the middle of the 29–165 tok/s/user interactivity band — at 97 tok/s/user — GB200 NVL72 runs $1.20 per million tokens on Qwen 3.5 397B-A17B while H100 runs $1.66. GB200 NVL72 is the cheaper choice by 39%.
On Qwen 3.5 397B-A17B at 131 tok/s/user, the per-million math comes out to $2.56 for GB200 NVL72 and $4.06 for H100; GB200 NVL72 delivers 59% more output per dollar. (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.)
GPU pricing (owning hyperscaler): GB200 NVL72 $2.21/GPU/hr · H100 $1.30/GPU/hr. Source: SemiAnalysis Market August 2025 Pricing Surveys & AI Cloud TCO Model.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Dollar per Million Tokens | GB200 NVL72:$0.489H100:$0.726 | GB200 NVL72:$1.201H100:$1.664 | GB200 NVL72:$2.557H100:$4.057 |
| Concurrency | GB200 NVL72:~129H100:~39 | GB200 NVL72:~22H100:~10 | GB200 NVL72:~8H100:~3 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.