// RUN โ€” FITS

Can the NVIDIA L40S 48GB run StyleGAN3-T?

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

StyleGAN3-T on the NVIDIA L40S 48GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 651k 141417-168538.1 t/s ~6547.1-7802.7 (slow)
Q8 yes 651k 282834.1-337076.2 t/s ~13094.2-15605.4 (slow)
Q4 yes 651k 505060.9-601921.8 t/s ~23382.4-27866.8 (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 NVIDIA L40S 48GB run StyleGAN3-T?โ–ถ

Yes. At Q4 it generates ~505060.9-601921.8 tok/s and fits up to 651k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~23382.4-27866.8 tok/s.

How much context fits?โ–ถ

At Q4, up to about 651k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.