🔌
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 LUKE?
Yes — here's how fast and up to what context, with real measured numbers.
LUKE on the FuriosaAI RNGD
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 636k | 1009.3-1226.7 t/s | ~26.9-32.7 (slow) |
| Q8 | yes | 644k | 2018.6-2453.4 t/s | ~53.8-65.4 (slow) |
| Q4 | yes | 647k | 3604.7-4381.1 t/s | ~96.1-116.8 (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 LUKE?▶
Yes. At Q4 it generates ~3604.7-4381.1 tok/s and fits up to 647k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~96.1-116.8 tok/s.
How much context fits?▶
At Q4, up to about 647k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.