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 90 tok/s/user: 9770 / 6025 tok/s/GPU, $0.07 / $0.12 per million tokens. B300 is 85% cheaper per token; B300 delivers 62% more tok/s/GPU.
Around the middle of the 44–228 tok/s/user interactivity band, at 136 tok/s/user on Qwen 3.5 397B-A17B: B300 runs 6823 tok/s/GPU at $0.10/M tokens, GB300 NVL72 runs 2801 at $0.26/M. B300 is 166% cheaper per token; B300 delivers 144% more tok/s/GPU.
Setting 183 tok/s/user as the target on Qwen 3.5 397B-A17B, B300 produces 4481 tok/s/GPU ($0.14 per million tokens) and GB300 NVL72 produces 1338 ($0.55). B300 is 292% cheaper per token; B300 delivers 235% more tok/s/GPU. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | B300:9770.0GB300 NVL72:6024.7 | B300:6823.1GB300 NVL72:2801.0 | B300:4481.0GB300 NVL72:1338.4 |
| Cost ($/M tok) | B300:$0.066GB300 NVL72:$0.122 | B300:$0.097GB300 NVL72:$0.256 | B300:$0.141GB300 NVL72:$0.553 |
| tok/s/MW | B300:4502289GB300 NVL72:2868920 | B300:3144307GB300 NVL72:1333831 | B300:2064984GB300 NVL72:637343 |
| Concurrency | B300:~26GB300 NVL72:~71 | B300:~12GB300 NVL72:~21 | B300:~6GB300 NVL72:~7 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.