// RUN โ€” FITS

Can the Apple M1 Ultra 128GB run ProteinBERT?

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

ProteinBERT on the Apple M1 Ultra 128GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 1734k 17000-20500 t/s ~212.5-256.3 (slow)
Q8 yes 1735k 34000-41000 t/s ~425-512.5 (slow)
Q4 yes 1735k 60714.3-73214.3 t/s ~758.9-915.2 (slow)

Estimate (memory-bound). Beyond "max context" the KV cache spills to SSD swap and speed drops to the "beyond" figure.

FAQ

Can the Apple M1 Ultra 128GB run ProteinBERT?โ–ถ

Yes. At Q4 it generates ~60714.3-73214.3 tok/s and fits up to 1735k context; beyond that it spills to SSD swap and slows to ~758.9-915.2 tok/s.

How much context fits?โ–ถ

At Q4, up to about 1735k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.