🔌
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 Minitron 4B?
Yes — here's how fast and up to what context, with real measured numbers.
Minitron 4B on the NVIDIA DGX Spark
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1607k | 23.7-28.3 t/s | ~0.9-1 (slow) |
| Q8 | yes | 1671k | 47.5-56.6 t/s | ~1.7-2.1 (slow) |
| Q4 | yes | 1699k | 84.7-101 t/s | ~3.1-3.7 (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 Minitron 4B?▶
Yes. At Q4 it generates ~84.7-101 tok/s and fits up to 1699k context; beyond that it spills to SSD swap and slows to ~3.1-3.7 tok/s.
How much context fits?▶
At Q4, up to about 1699k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.