🔌
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 Prithvi WxC?
Yes — here's how fast and up to what context, with real measured numbers.
Prithvi WxC on the NVIDIA DGX Spark
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1665k | 43.3-51.6 t/s | ~1.6-1.9 (slow) |
| Q8 | yes | 1700k | 86.6-103.3 t/s | ~3.2-3.8 (slow) |
| Q4 | yes | 1715k | 154.7-184.4 t/s | ~5.7-6.8 (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 Prithvi WxC?▶
Yes. At Q4 it generates ~154.7-184.4 tok/s and fits up to 1715k context; beyond that it spills to SSD swap and slows to ~5.7-6.8 tok/s.
How much context fits?▶
At Q4, up to about 1715k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.