Cost, Benefit, No
Apple won't climb Mt Everest because a spreadsheet says no.
Podcast: Accidental Tech Podcast
Episode: 666: We Have Nothing That Hot
John
In particular, like setting aside when they screwed up the GPU thing by under-resourcing some buffer or something, the fact that they insist on building their Ultras out of two Maxes is, it’s a clever idea if it had worked out for them, but I think that hurt the scaling more than anything else. Whereas if they had made a dedicated chip, it wouldn’t.But if they made a dedicated chip, they can’t justify the expense. That’s been the whole thing with the Extreme, you know, with the Quad thing. They can’t justify the expense. We’re not going to sell, we’re going to sell this many of them and it’s going to cost this much to make. It’ll be our most expensive, you know, like Intel had this problem to some. degree but not really because they were sold to server farms and everything like this chip this whatever this high-end chip is going to cost the most and be the most complicated chip that we make and we’re going to sell a tiny tiny fraction of the number we sell of our other chips like why are we doing this again?
and i would point them to my case for a true mac pro successor article and saying you’re you know mount everest you’re climbing because it’s there blah blah blah but apple doesn’t apple doesn’t have that they don’t they don’t have that in them they can’t justify it someone comes down and says cost benefit no