// RUN โ€” FITS

Can the NVIDIA GeForce RTX 3090 run GLIDE?

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

GLIDE on the NVIDIA GeForce RTX 3090

QuantFits?Max contextSpeedBeyond context
FP16 yes 207k 97.6-116.3 t/s ~4.2-5 (slow)
Q8 yes 261k 195.2-232.7 t/s ~8.3-9.9 (slow)
Q4 yes 284k 348.6-415.5 t/s ~14.9-17.8 (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 GeForce RTX 3090 run GLIDE?โ–ถ

Yes. At Q4 it generates ~348.6-415.5 tok/s and fits up to 284k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~14.9-17.8 tok/s.

How much context fits?โ–ถ

At Q4, up to about 284k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.