// RUN โ€” FITS

Can the FuriosaAI RNGD run Falcon-OCR?

Yes โ€” here's how fast and up to what context, with real measured numbers.

Falcon-OCR on the FuriosaAI RNGD

QuantFits?Max contextSpeedBeyond context
FP16 yes 642k 1625-1975 t/s ~43.3-52.7 (slow)
Q8 yes 646k 3250-3950 t/s ~86.7-105.3 (slow)
Q4 yes 648k 5803.6-7053.6 t/s ~154.8-188.1 (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 Falcon-OCR?โ–ถ

Yes. At Q4 it generates ~5803.6-7053.6 tok/s and fits up to 648k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~154.8-188.1 tok/s.

How much context fits?โ–ถ

At Q4, up to about 648k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.