// RUN โ€” FITS

Can the FuriosaAI RNGD run HyenaDNA?

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

HyenaDNA on the FuriosaAI RNGD

QuantFits?Max contextSpeedBeyond context
FP16 yes 651k 73863.6-89772.7 t/s ~1969.7-2393.9 (slow)
Q8 yes 651k 147727.3-179545.5 t/s ~3939.4-4787.9 (slow)
Q4 yes 651k 263798.7-320616.9 t/s ~7034.6-8549.8 (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 HyenaDNA?โ–ถ

Yes. At Q4 it generates ~263798.7-320616.9 tok/s and fits up to 651k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~7034.6-8549.8 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.