Re:Apple M3 Ultra
by nightski on Wednesday March 05, 2025 @11:00PM
(#43276270)
$10k to run a 4 bit quantized model. Ouch.
[ Parent
| on HN ]
and low TOPS, but da novices will go deep in debt for dem. (Apple employee up/down voters, please --I have a lifetime of experience and your company might have failed w/o me and my million dollar pos at juuust the righ time, m'kay?)
M3 Ultra is 17-36 TOPS, GB10 (Project DIGITS) is more like 270 TOPS. Two Project DIGITS will cost about $4K less than an M3 Ultra Studio. This is a fact. GB10 is "only" 128GB @ 500Mbps tho.
Downvote away Cupetino Tea drinkers!
are you comparing the same models? How did you calculate the TOPS for M3 Ultra?
An M3 Ultra is two M3 Max chips connected via fabric, so physics.
Did not mean to shit on anyone's parade, but it's a trap for novices, with the caveat that you reportedly can't buy a GB10 until "May 2025" and the expectation that it will be severely supply constrained. For some (overfunded startups running on AI monkey code? Youtube Influencers?), that timing is an unacceptable risk, so I do expect these things to fly off the shelves and then hit eBay this Summer.
Re:Apple M3 Ultra
by fredoliveira on Thursday March 06, 2025 @10:19AM
(#43281230)
what?
Sorry, an apostrophe got lost in "PO's"
Re:Apple M3 Ultra
by titanomachy on Thursday March 06, 2025 @01:42PM
(#43283621)
I'm downvoting you because your use of language is so annoying, not because I work for Apple.
So, Microsoft?
The M4 MacBook Pro 128GB can run a 32B perimeter model with an 8 bit quantized model just fine
Re:Apple M3 Ultra
by OriginalMrPink on Thursday March 06, 2025 @08:03AM
(#43279737)
That's today. What about tomorrow?