Our "world" build is slightly faster on my M1 Max.
https://twitter.com/kiratpandya/status/1457438725680480257
The 3990x runs a bit faster on the initial compile stage but the linking is single threaded and the M1 Max catches up at that point. I expect the M1 Ultra to crush the 3990x on compile time.
There is Mold (https://github.com/rui314/mold) as an alternative but it doesn’t support MacOS yet and that’s a blocker for our team.
None of that swapability is available on any of the Apple systems, so when they stagnate there's nothing that can be done.
I love the idea of component machines, but at this time of my life, I would be just fine to never have build another PC. With that, these systems of plug in the cables and go to work are appealing. Having said that, it still rubs me the wrong way on how non-upgradeable these things are.
How have they stagnated?
https://www.dell.com/en-us/work/shop/workstations-isv-certif...
or the HP equivalent - they're still equipping them with Xeon's from 2 1/2 years ago. So pretty stagnant.
The CPU performance is impressive but I wonder how well it'd fare against say an AMD ThreadRipper or a dual socket system with an equivalent price? It'd be also interesting to see a deep learning benchmark against Nvidia. Branding and RAM limits suggest this is geared towards video processing, but might also be useful for some other domains.