🔌
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 KoGPT?
Yes — here's how fast and up to what context, with real measured numbers.
KoGPT on the NVIDIA DGX Spark
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 773k | 16.2-19.3 t/s | ~0.6-0.7 (slow) |
| Q8 | yes | 820k | 32.3-38.5 t/s | ~1.2-1.4 (slow) |
| Q4 | yes | 841k | 57.7-68.8 t/s | ~2.1-2.5 (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 KoGPT?▶
Yes. At Q4 it generates ~57.7-68.8 tok/s and fits up to 841k context; beyond that it spills to SSD swap and slows to ~2.1-2.5 tok/s.
How much context fits?▶
At Q4, up to about 841k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.