🔌
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 FRED-T5-XL?
Yes — here's how fast and up to what context, with real measured numbers.
FRED-T5-XL on the NVIDIA DGX Spark
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1682k | 57.3-68.3 t/s | ~2.1-2.5 (slow) |
| Q8 | yes | 1708k | 114.5-136.5 t/s | ~4.2-5 (slow) |
| Q4 | yes | 1720k | 204.5-243.8 t/s | ~7.5-8.9 (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 FRED-T5-XL?▶
Yes. At Q4 it generates ~204.5-243.8 tok/s and fits up to 1720k context; beyond that it spills to SSD swap and slows to ~7.5-8.9 tok/s.
How much context fits?▶
At Q4, up to about 1720k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.