🔌
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 H200 run phi-3-small 7.4B?
Yes — here's how fast and up to what context, with real measured numbers.
phi-3-small 7.4B on the NVIDIA H200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 865k | 236.8-282.2 t/s | ~2-2.4 (slow) |
| Q8 | yes | 922k | 473.5-564.3 t/s | ~3.9-4.7 (slow) |
| Q4 | yes | 947k | 845.6-1007.7 t/s | ~7-8.4 (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 H200 run phi-3-small 7.4B?▶
Yes. At Q4 it generates ~845.6-1007.7 tok/s and fits up to 947k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~7-8.4 tok/s.
How much context fits?▶
At Q4, up to about 947k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.