// RUN — FITS

Can the Apple M3 Max 36GB run Very Deep VAEs (ImageNet-64)?

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

Very Deep VAEs (ImageNet-64) on the Apple M3 Max 36GB

QuantFits?Max contextSpeedBeyond context
FP16 yes 468k 816-984 t/s ~27.2-32.8 (slow)
Q8 yes 470k 1632-1968 t/s ~54.4-65.6 (slow)
Q4 yes 470k 2914.3-3514.3 t/s ~97.1-117.1 (slow)

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

→ All models on the Apple M3 Max 36GB → Very Deep VAEs (ImageNet-64) specs → Other hardware

FAQ

Can the Apple M3 Max 36GB run Very Deep VAEs (ImageNet-64)?▶

Yes. At Q4 it generates ~2914.3-3514.3 tok/s and fits up to 470k context; beyond that it spills to SSD swap and slows to ~97.1-117.1 tok/s.

How much context fits?▶

At Q4, up to about 470k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.