🔌
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 phi-3.5-mini?
Yes — here's how fast and up to what context, with real measured numbers.
phi-3.5-mini on the NVIDIA B200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 2556k | 768.4-915.8 t/s | ~3.8-4.6 (slow) |
| Q8 | yes | 2614k | 1536.8-1831.6 t/s | ~7.7-9.2 (slow) |
| Q4 | yes | 2640k | 2744.4-3270.7 t/s | ~13.7-16.4 (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 phi-3.5-mini?▶
Yes. At Q4 it generates ~2744.4-3270.7 tok/s and fits up to 2640k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~13.7-16.4 tok/s.
How much context fits?▶
At Q4, up to about 2640k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.