🔌
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 ProGen2-base?
Yes — here's how fast and up to what context, with real measured numbers.
ProGen2-base on the NVIDIA H200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1933k | 2293.2-2733 t/s | ~19.1-22.8 (slow) |
| Q8 | yes | 1945k | 4586.4-5466 t/s | ~38.2-45.5 (slow) |
| Q4 | yes | 1950k | 8190-9760.7 t/s | ~68.2-81.3 (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 H200 run ProGen2-base?▶
Yes. At Q4 it generates ~8190-9760.7 tok/s and fits up to 1950k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~68.2-81.3 tok/s.
How much context fits?▶
At Q4, up to about 1950k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.