🔌
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 SANA 1.5 4.8B?
Yes — here's how fast and up to what context, with real measured numbers.
SANA 1.5 4.8B on the NVIDIA DGX Spark
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1588k | 20.8-24.7 t/s | ~0.8-0.9 (slow) |
| Q8 | yes | 1662k | 41.5-49.5 t/s | ~1.5-1.8 (slow) |
| Q4 | yes | 1694k | 74.1-88.4 t/s | ~2.7-3.2 (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 SANA 1.5 4.8B?▶
Yes. At Q4 it generates ~74.1-88.4 tok/s and fits up to 1694k context; beyond that it spills to SSD swap and slows to ~2.7-3.2 tok/s.
How much context fits?▶
At Q4, up to about 1694k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.