🔌
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 StableLM-2-1.6B?
Yes — here's how fast and up to what context, with real measured numbers.
StableLM-2-1.6B on the NVIDIA H200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1906k | 1065.4-1269.8 t/s | ~8.9-10.6 (slow) |
| Q8 | yes | 1931k | 2130.8-2539.5 t/s | ~17.8-21.2 (slow) |
| Q4 | yes | 1942k | 3805.1-4534.8 t/s | ~31.7-37.8 (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 StableLM-2-1.6B?▶
Yes. At Q4 it generates ~3805.1-4534.8 tok/s and fits up to 1942k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~31.7-37.8 tok/s.
How much context fits?▶
At Q4, up to about 1942k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.