🔌
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 SGPT BE 5.8B?
Yes — here's how fast and up to what context, with real measured numbers.
SGPT BE 5.8B on the NVIDIA DGX Spark
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 779k | 17.2-20.5 t/s | ~0.6-0.8 (slow) |
| Q8 | yes | 823k | 34.4-41 t/s | ~1.3-1.5 (slow) |
| Q4 | yes | 843k | 61.4-73.1 t/s | ~2.2-2.7 (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 SGPT BE 5.8B?▶
Yes. At Q4 it generates ~61.4-73.1 tok/s and fits up to 843k context; beyond that it spills to SSD swap and slows to ~2.2-2.7 tok/s.
How much context fits?▶
At Q4, up to about 843k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.