Qwen 3.5 397B-A17B — B300 vs GB300 NVL72
Head-to-head AI inference benchmark comparison of B300 (NVIDIA Blackwell) and GB300 NVL72 (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.
B300 / GB300 NVL72 on Qwen 3.5 397B-A17B at 87 tok/s/user: 5095 / 10829 tok/s/chip, $0.12 / $0.06 per million tokens. GB300 NVL72 is 108% cheaper per token; GB300 NVL72 delivers 113% more tok/s/chip.
Around the middle of the 40–230 tok/s/user interactivity band, at 135 tok/s/user on Qwen 3.5 397B-A17B: B300 runs 3489 tok/s/chip at $0.18/M tokens, GB300 NVL72 runs 4862 at $0.13/M. GB300 NVL72 is 34% cheaper per token; GB300 NVL72 delivers 39% more tok/s/chip.
Setting 183 tok/s/user as the target on Qwen 3.5 397B-A17B, B300 produces 2574 tok/s/chip ($0.24 per million tokens) and GB300 NVL72 produces 2682 ($0.24). GB300 NVL72 is 2% cheaper per token; GB300 NVL72 delivers 4% more tok/s/chip. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B300:5095.1GB300 NVL72:10829.2 | B300:3489.2GB300 NVL72:4861.9 | B300:2573.9GB300 NVL72:2682.3 |
| Cost ($/M tok) | B300:$0.123GB300 NVL72:$0.059 | B300:$0.179GB300 NVL72:$0.134 | B300:$0.245GB300 NVL72:$0.239 |
| tok/s/MW | B300:2681641GB300 NVL72:5108120 | B300:1836413GB300 NVL72:2293329 | B300:1354709GB300 NVL72:1265234 |
| Concurrency | B300:~27GB300 NVL72:~1100 | B300:~12GB300 NVL72:~95 | B300:~7GB300 NVL72:~37 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.