4 ms·
Apple used to contribute to the PyTorch MPS backend, but decided to create their own framework (MLX) instead, fragmenting the ecosystem for very little gain. (M
by ipsum2 1y ago
Apple used to contribute to the PyTorch MPS backend, but decided to create their own framework (MLX) instead, fragmenting the ecosystem for very little gain. (MLX is basically PyTorch, but invented-at-apple)
Meta, the creator and main contributor to PyTorch, does not use Macs for their day-to-day ML work (they focus on GPUs and CPUs), so the MPS backend is sadly incomplete and has errors like the one you see here.
- almostgotcaught 1y agonone of this is correct (except the part where FB doesn't use apple in prod). EDIT: for the downvoters - i'll repeat, this is not a correct assessment of the relationship between Apple and PyTorch. but you can keep downvoting if you want <shrug>
- ipsum2 1y agoPlease be specific if you have anything to say. By the way, the co-creator and core maintainer of PyTorch has the same opinion as me. https://x.com/soumithchintala/status/1978848796953161754 https://x.com/soumithchintala/status/1978848796953161754 "MacStudio you ask? Apple Engineering's *actual* time spent on PyTorch support has't given me confidence that PyTorch Mac experience would get anywhere close to NVIDIA's any time soon, if ever. The Meta engineers continue to do a huge amount of heavy-lifting for improving the MPS backend, including feeling the responsibility for the Mac experience. Apple's priorities keep changing, the number of engineering hours they contribute keeps changing and their interest in actually and wholly owning the PyTorch MPS backend keeps varying. If Apple wants MacStudio to become an actual AI devbox, and not just an AI inference machine, then prioritizing software support for PyTorch (>90% marketshare in AI) would probably be a good idea."
- hedgehog 1y agoApple has never cared about ML research on their hardware. I've never been able to pin down a specific reason why, best I can figure out is they don't see it bringing enough additional hardware sales to be a focus.
- almostgotcaught 1y agolol @ quoting soumith - the guy's sole job responsibility is tweeting.
- ipsum2 1y agoIf you have more knowledge than the core maintainer of pytorch, why are you unwilling to share, instead of snarking?
- almostgotcaught 1y agoHe's not a core maintainer and hasn't been for years - pytorch's contributors are completely public https://github.com/pytorch/pytorch/graphs/contributors https://github.com/pytorch/pytorch/graphs/contributors
- ipsum2 1y agoGot it, you're a bad troll. He's listed as the "Lead Core Maintainer" on PyTorch.
- almostgotcaught 1y agoi'm sure mark zuckerberg is as well <shrug>
- Q6T46nT668w6i3m 1y agoEveryone I know at Meta uses a Mac
- ipsum2 1y agoNo one at Meta runs local inference on a Mac, unless its for fun.
- sampton 1y agoMLX and MPS are 2 completely different teams within Apple. It's more like MPS team doesn't have control or visibility into PyTorch roadmap and can only contribute so much from their side.