🔌
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 MiniCPM-3-4B?
Yes — here's how fast and up to what context, with real measured numbers.
MiniCPM-3-4B on the NVIDIA B200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 2550k | 730-870 t/s | ~3.7-4.4 (slow) |
| Q8 | yes | 2611k | 1460-1740 t/s | ~7.3-8.7 (slow) |
| Q4 | yes | 2638k | 2607.1-3107.1 t/s | ~13-15.5 (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 MiniCPM-3-4B?▶
Yes. At Q4 it generates ~2607.1-3107.1 tok/s and fits up to 2638k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~13-15.5 tok/s.
How much context fits?▶
At Q4, up to about 2638k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.