🔌
New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
// RUN — FITS
Can the NVIDIA B200 run Qwen2.5-3B?
Yes — here's how fast and up to what context, with real measured numbers.
Qwen2.5-3B on the NVIDIA B200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 2578k | 945-1126.2 t/s | ~4.7-5.6 (slow) |
| Q8 | yes | 2625k | 1890-2252.4 t/s | ~9.4-11.3 (slow) |
| Q4 | yes | 2646k | 3374.9-4022.2 t/s | ~16.9-20.1 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to system RAM offload (depends on your system RAM) and speed drops to the "beyond" figure.
FAQ
Can the NVIDIA B200 run Qwen2.5-3B?▶
Yes. At Q4 it generates ~3374.9-4022.2 tok/s and fits up to 2646k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~16.9-20.1 tok/s.
How much context fits?▶
At Q4, up to about 2646k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.