🔌
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 FuriosaAI RNGD run Cosmos-Predict2-14B-Text2Image?
Yes — here's how fast and up to what context, with real measured numbers.
Cosmos-Predict2-14B-Text2Image on the FuriosaAI RNGD
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 89k | 34.8-42.3 t/s | ~0.9-1.1 (slow) |
| Q8 | yes | 175k | 69.6-84.6 t/s | ~1.9-2.3 (slow) |
| Q4 | yes | 213k | 124.4-151.1 t/s | ~3.3-4 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to system RAM offload (depends on your system RAM) and speed drops to the "beyond" figure.
FAQ
Can the FuriosaAI RNGD run Cosmos-Predict2-14B-Text2Image?▶
Yes. At Q4 it generates ~124.4-151.1 tok/s and fits up to 213k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~3.3-4 tok/s.
How much context fits?▶
At Q4, up to about 213k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.