// RUN โ€” FITS

Can the Intel Arc A770 16GB run Pythia-1b?

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

Pythia-1b on the Intel Arc A770 16GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 171k 162.4-201.6 t/s ~11.6-14.4 (slow)
Q8 yes 186k 324.8-403.2 t/s ~23.2-28.8 (slow)
Q4 yes 193k 580-720 t/s ~41.4-51.4 (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 A770 16GB run Pythia-1b?โ–ถ

Yes. At Q4 it generates ~580-720 tok/s and fits up to 193k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~41.4-51.4 tok/s.

How much context fits?โ–ถ

At Q4, up to about 193k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.