🔌
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 Persimmon-8B?
Yes — here's how fast and up to what context, with real measured numbers.
Persimmon-8B on the NVIDIA H200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 836k | 188.4-224.5 t/s | ~1.6-1.9 (slow) |
| Q8 | yes | 907k | 376.8-449 t/s | ~3.1-3.7 (slow) |
| Q4 | yes | 939k | 672.8-801.8 t/s | ~5.6-6.7 (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 Persimmon-8B?▶
Yes. At Q4 it generates ~672.8-801.8 tok/s and fits up to 939k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~5.6-6.7 tok/s.
How much context fits?▶
At Q4, up to about 939k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.