// RUN โ€” FITS

Can the Instinct MI430X run HiFi - NN?

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

HiFi - NN on the Instinct MI430X

QuantFits?Max contextSpeedBeyond context
FP16 yes 6041k 2446500-2990166.7 t/s ~4200-5133.3 (slow)
Q8 yes 6042k 4893000-5980333.3 t/s ~8400-10266.7 (slow)
Q4 yes 6042k 8737500-10679166.7 t/s ~15000-18333.3 (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 Instinct MI430X run HiFi - NN?โ–ถ

Yes. At Q4 it generates ~8737500-10679166.7 tok/s and fits up to 6042k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~15000-18333.3 tok/s.

How much context fits?โ–ถ

At Q4, up to about 6042k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.