🔌
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 DGX Spark run ProGen2-base?
Yes — here's how fast and up to what context, with real measured numbers.
ProGen2-base on the NVIDIA DGX Spark
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1712k | 130.4-155.4 t/s | ~4.8-5.7 (slow) |
| Q8 | yes | 1723k | 260.9-310.9 t/s | ~9.6-11.4 (slow) |
| Q4 | yes | 1728k | 465.8-555.1 t/s | ~17.1-20.3 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to SSD swap and speed drops to the "beyond" figure.
FAQ
Can the NVIDIA DGX Spark run ProGen2-base?▶
Yes. At Q4 it generates ~465.8-555.1 tok/s and fits up to 1728k context; beyond that it spills to SSD swap and slows to ~17.1-20.3 tok/s.
How much context fits?▶
At Q4, up to about 1728k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.