🔌
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 Nemotron-H 56B?
Yes — here's how fast and up to what context, with real measured numbers.
Nemotron-H 56B on the NVIDIA DGX Spark
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 5k | 1.8-2.1 t/s | ~0.1-0.1 (slow) |
| Q8 | yes | 176k | 3.6-4.2 t/s | ~0.1-0.2 (slow) |
| Q4 | yes | 251k | 6.4-7.6 t/s | ~0.2-0.3 (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 Nemotron-H 56B?▶
Yes. At Q4 it generates ~6.4-7.6 tok/s and fits up to 251k context; beyond that it spills to SSD swap and slows to ~0.2-0.3 tok/s.
How much context fits?▶
At Q4, up to about 251k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.