// RUN โ€” FITS

Can the AMD Radeon RX 7900 XT run CLIP (ViT L/14@336px)?

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

CLIP (ViT L/14@336px) on the AMD Radeon RX 7900 XT

QuantFits?Max contextSpeedBeyond context
FP16 yes 247k 681.1-832.4 t/s ~34.1-41.6 (slow)
Q8 yes 252k 1362.2-1664.9 t/s ~68.1-83.2 (slow)
Q4 yes 255k 2432.4-2973 t/s ~121.6-148.6 (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 CLIP (ViT L/14@336px)?โ–ถ

Yes. At Q4 it generates ~2432.4-2973 tok/s and fits up to 255k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~121.6-148.6 tok/s.

How much context fits?โ–ถ

At Q4, up to about 255k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.