// RUN โ€” FITS

Can the NVIDIA L40S 48GB run Yi-1.5-9B?

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

Yi-1.5-9B on the NVIDIA L40S 48GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 191k 35.7-42.6 t/s ~1.7-2 (slow)
Q8 yes 258k 71.4-85.1 t/s ~3.3-3.9 (slow)
Q4 yes 288k 127.6-152 t/s ~5.9-7 (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 NVIDIA L40S 48GB run Yi-1.5-9B?โ–ถ

Yes. At Q4 it generates ~127.6-152 tok/s and fits up to 288k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~5.9-7 tok/s.

How much context fits?โ–ถ

At Q4, up to about 288k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.