🔌
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 GenSLM?
Yes — here's how fast and up to what context, with real measured numbers.
GenSLM on the NVIDIA H200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 298k | 70.1-83.5 t/s | ~0.6-0.7 (slow) |
| Q8 | yes | 394k | 140.2-167 t/s | ~1.2-1.4 (slow) |
| Q4 | yes | 436k | 250.3-298.3 t/s | ~2.1-2.5 (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 H200 run GenSLM?▶
Yes. At Q4 it generates ~250.3-298.3 tok/s and fits up to 436k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~2.1-2.5 tok/s.
How much context fits?▶
At Q4, up to about 436k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.