// RUN โ€” FITS

Can the FuriosaAI RNGD run MegaMolBART?

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

MegaMolBART on the FuriosaAI RNGD

QuantFits?Max contextSpeedBeyond context
FP16 yes 650k 10833.3-13166.7 t/s ~288.9-351.1 (slow)
Q8 yes 650k 21666.7-26333.3 t/s ~577.8-702.2 (slow)
Q4 yes 651k 38690.5-47023.8 t/s ~1031.7-1254 (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 MegaMolBART?โ–ถ

Yes. At Q4 it generates ~38690.5-47023.8 tok/s and fits up to 651k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~1031.7-1254 tok/s.

How much context fits?โ–ถ

At Q4, up to about 651k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.