Nvidia doesn‘t deliver LLMs, if that is what you‘re associating with AI. But have a deeper look into e.g. „physical AI“ and robotics. Take a dive into NVidia „Cosmos“ which is OSS by the way
https://www.nvidia.com/en-us/ai/cosmos/
Have a look at the role of CUDA when developing AI and MLP.
Yes it is foremost hardware you need to train and run AI solutions, but by banning NVIDIA from its platforms, it also banned AI development. Only one reason why I switched a lot of my workload to Linux. By the way as a Mac user, you‘ll loce the native Gnome experience on Linux.
And if we talk about the future and the Mac Studio with 512GB for $15K, do yourself a favor and look at NVidia Digits. An AI supercomputer with 128GB RAM for $3K. All workflows in the future will replaced by AI agents, AI in video and photo editing. One may say the Mac Studio is targeted at a different audience, but in terms of AI capabilities it is dust.
Btw. It is smaller compared to the Mac Mini - and stackable …
https://www.nvidia.com/en-us/project-digits/