// RUN โ€” FITS

Can the AMD Radeon RX 7900 XT run MoEFold2D?

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

MoEFold2D on the AMD Radeon RX 7900 XT

QuantFits?Max contextSpeedBeyond context
FP16 yes 258k 262500-320833.3 t/s ~13125-16041.7 (slow)
Q8 yes 258k 525000-641666.7 t/s ~26250-32083.3 (slow)
Q4 yes 258k 937500-1145833.3 t/s ~46875-57291.7 (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 AMD Radeon RX 7900 XT run MoEFold2D?โ–ถ

Yes. At Q4 it generates ~937500-1145833.3 tok/s and fits up to 258k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~46875-57291.7 tok/s.

How much context fits?โ–ถ

At Q4, up to about 258k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.