🔌
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 SmolLM-1.7B?
Yes — here's how fast and up to what context, with real measured numbers.
SmolLM-1.7B on the NVIDIA B200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 2620k | 1707.6-2035.1 t/s | ~8.5-10.2 (slow) |
| Q8 | yes | 2646k | 3415.2-4070.2 t/s | ~17.1-20.4 (slow) |
| Q4 | yes | 2658k | 6098.6-7268.2 t/s | ~30.5-36.3 (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 SmolLM-1.7B?▶
Yes. At Q4 it generates ~6098.6-7268.2 tok/s and fits up to 2658k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~30.5-36.3 tok/s.
How much context fits?▶
At Q4, up to about 2658k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.