// RUN โ€” FITS

Can the Intel Arc Pro B70 run BEIT-3?

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

BEIT-3 on the Intel Arc Pro B70

QuantFits?Max contextSpeedBeyond context
FP16 yes 368k 92.8-115.2 t/s ~6.1-7.6 (slow)
Q8 yes 397k 185.6-230.4 t/s ~12.2-15.2 (slow)
Q4 yes 410k 331.4-411.4 t/s ~21.8-27.1 (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 Intel Arc Pro B70 run BEIT-3?โ–ถ

Yes. At Q4 it generates ~331.4-411.4 tok/s and fits up to 410k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~21.8-27.1 tok/s.

How much context fits?โ–ถ

At Q4, up to about 410k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.