I'm curious why they wouldn't opt for Mac studios -

Smaller size = more efficient cluster layout?

RAM is the only care so skip the beefy studio CPUS and max ram on something cheaper?

Perhaps someone closer to this space could illuminate me

Does anyone knows why? I thought MacMinis shortage was all about OpenClaws but it seems to be a bit more than that happening.

What startup? a better source than some Twitter account?

Whatever this article is about, it's probably not about economic advantage of local hosting.

The cloud GPUs are 170 thousand a year.

1000 minis are going to cost 1.5 million + floor space, power, setup and administration over their lifetime.

Amortizing the minis against the cloud savings looks to require 10 or more years.

But the entire genAI industry has manifest in less than 10 years-- AppleSi based compute modules aren't that old!