// RUN โ€” FITS

Can the Intel Gaudi 3 run AraBERT LArge v2?

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

AraBERT LArge v2 on the Intel Gaudi 3

QuantFits?Max contextSpeedBeyond context
FP16 yes 1763k 2892.2-3590.3 t/s ~31.3-38.8 (slow)
Q8 yes 1768k 5784.4-7180.6 t/s ~62.5-77.6 (slow)
Q4 yes 1771k 10329.2-12822.5 t/s ~111.7-138.6 (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 Gaudi 3 run AraBERT LArge v2?โ–ถ

Yes. At Q4 it generates ~10329.2-12822.5 tok/s and fits up to 1771k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~111.7-138.6 tok/s.

How much context fits?โ–ถ

At Q4, up to about 1771k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.