// RUN โ€” FITS

Can the NVIDIA L40S 48GB run Engine-Base (NE)?

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

Engine-Base (NE) on the NVIDIA L40S 48GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 647k 2543.2-3031 t/s ~117.7-140.3 (slow)
Q8 yes 649k 5086.5-6061.9 t/s ~235.5-280.6 (slow)
Q4 yes 650k 9082.9-10824.9 t/s ~420.5-501.2 (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 Engine-Base (NE)?โ–ถ

Yes. At Q4 it generates ~9082.9-10824.9 tok/s and fits up to 650k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~420.5-501.2 tok/s.

How much context fits?โ–ถ

At Q4, up to about 650k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.