// RUN โ€” FITS

Can the NVIDIA GeForce RTX 3090 Ti run ViT + DINO?

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

ViT + DINO on the NVIDIA GeForce RTX 3090 Ti

QuantFits?Max contextSpeedBeyond context
FP16 yes 311k 4328.5-5158.6 t/s ~171.8-204.7 (slow)
Q8 yes 313k 8656.9-10317.2 t/s ~343.5-409.4 (slow)
Q4 yes 313k 15458.8-18423.5 t/s ~613.4-731.1 (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 Ti run ViT + DINO?โ–ถ

Yes. At Q4 it generates ~15458.8-18423.5 tok/s and fits up to 313k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~613.4-731.1 tok/s.

How much context fits?โ–ถ

At Q4, up to about 313k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.