undefined | Better HN

0 pointssamwillis2y ago0 comments

I very much doubt that.

0 comments

3 comments · 1 top-level

smoldesu2y ago· 2 in thread

If they want to own the whole stack, I don't think they have much to work with. Their highest-end server chip is a duplex laptop SOC, with maxed-out memory that doesn't even match the lowest-end Grace CPU you can buy (nevermind a fully-networked GH200). Their consumer offerings are competitive, but I don't think Apple Silicon or CoreML is ready to seriously compete with Grace and CUDA.

samwillisOP2y ago

While Apple silicone may not be there for training, I think it's probably there for inference. I expect next years device models to launch with exclusive support for Apples own LLM based Siri.

smoldesu2y ago

Sure. Haswell CPUs from 2014 are "there" for inference if they have AVX support and 8gb of RAM. Inferencing isn't the problem though, not on M1 or Macbooks from 2016. Scaling a desirable (and hopefully open) GPGPU programming interface is. This is bottlenecked by both hardware and software decisions Apple has made, making a "home grown" competitive model much more unlikely in my eyes.

I agree that there is an incentive to put AI models on your OS. I just don't think Apple can own the whole stack if they want to play ball right now.

j / k navigate · click thread line to collapse