// RUN โ€” FITS

Can the Instinct MI430X run LUKE?

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

LUKE on the Instinct MI430X

QuantFits?Max contextSpeedBeyond context
FP16 yes 6027k 15195.7-18572.5 t/s ~26.1-31.9 (slow)
Q8 yes 6034k 30391.3-37144.9 t/s ~52.2-63.8 (slow)
Q4 yes 6037k 54270.2-66330.2 t/s ~93.2-113.9 (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 LUKE?โ–ถ

Yes. At Q4 it generates ~54270.2-66330.2 tok/s and fits up to 6037k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~93.2-113.9 tok/s.

How much context fits?โ–ถ

At Q4, up to about 6037k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.