// RUN โ€” FITS

Can the Instinct MI430X run Step 3.7 Flash?

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

Step 3.7 Flash on the Instinct MI430X

QuantFits?Max contextSpeedBeyond context
FP16 no โ€” โ€” โ€”
Q8 yes 378k 74.1-90.6 t/s ~0.1-0.2 (slow)
Q4 yes 544k 132.4-161.8 t/s ~0.2-0.3 (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 Step 3.7 Flash?โ–ถ

Yes. At Q4 it generates ~132.4-161.8 tok/s and fits up to 544k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~0.2-0.3 tok/s.

How much context fits?โ–ถ

At Q4, up to about 544k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.