// RUN โ€” FITS

Can the FuriosaAI RNGD run ImageBind?

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

ImageBind on the FuriosaAI RNGD

QuantFits?Max contextSpeedBeyond context
FP16 yes 622k 523.1-635.7 t/s ~13.9-17 (slow)
Q8 yes 637k 1046.1-1271.5 t/s ~27.9-33.9 (slow)
Q4 yes 643k 1868.1-2270.5 t/s ~49.8-60.5 (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 ImageBind?โ–ถ

Yes. At Q4 it generates ~1868.1-2270.5 tok/s and fits up to 643k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~49.8-60.5 tok/s.

How much context fits?โ–ถ

At Q4, up to about 643k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.