Muskyinhere

You are on hn so let me ask you this: Are/were you a nerd? Did you play around with systems, use linux for fun etc.?

If yes, you just might be depressed or haven't realized that a job might be shit anyway.

If no, interesting though that you are aware of hn (i have to assume a certain interest because i know plenty of software people avg ones, not being aware of hn at all).

But hf at the navy.

Customers want choices. They just sell all models.

No one is running LLMs on consumer NVidia GPUs or apple MacBooks.

A dev, if they want to run local models, probably run something which just fits on a proper GPU. For everything else, everyone uses an API key from whatever because its fundamentaly faster.

IF a affordable intel GPU would be relevant faster for inferencing, is not clear at all.

A 4090 is at least double the speed of Apples GPU.

NVidia Blackwell is not just a GPU. Its a Rack with a interconnect through a custom Nvidia based Network.

And it needs liquid cooling.

You don't just plugin intel cards 'out of the box'.

Because if they could just do that and it would rival what NVidia has, they would just do it.

But obvoiusly they don't.

And for reasons: NVidia has worked on CUDA for ages, do you believe they just replace this whole thing in no time?