Note that AMD has an abandoned K12 plan. They still have an ARM license AFAIU. And like what Fujitsu did to their SPARC to produce A64FX, swapping the frontend may make it possible to support an ARM instruction set without much modifications to the backend. So if they will, AMD would make an ARM chip
ARM: maybe. Intel would be smart to make an ARM CPU. M1: probably not. Apple never really knew how to sell servers, and having a premium brand means they can't sell low-end laptops, so the market for something between a Chromebook, iPad mini, and low-end Macbook Air will still be x86 (maybe Intel, maybe AMD, whichever's cheapest) until ARM Windows is a thing.
ARM in the datacenter should be Intel's main fear.
Exactly. It would be a bit of a gamble, as it may undercut their x86 lines, but as a hedge against the success of ARM/M[1,2,...], they may do themselves a favor to attempt a project to implement RISC-V.
They don't have much to lose, as AMD is already killing their x86 lines anyway.
I've been hoping Intel or AMD would try RISC-V. If the ISA is inherently more efficient (not saying it is) they should be able to prove it by having already optimized so much core and cache hardware.
And do what, shoot both x86 feet they're standing on? If they wanted to optimize technical/engineering success, it's absolutely the best move. But to optimize shareholder value or anything like it, it's impossible.
If money was the way forward AMD wouldn't be kicking Intel's behind. It is more about the right people and the right culture. IMO Intel doesn't have either and haven't in quote a bit of years. Sure they have very good people but that is not worth much if they can't work in the culture they are in (which they clearly can't).
