Of course there's been business motivation for OEMs to diversify beyond Intel. It's the same reason why the phone vendors are eager for Intel to succeed in phones, so they have diversification beyond Qualcomm.
But in the case of Apple bringing up ARM to compete in the PC space, the chances of it happening are beyond even the visible horizon. The most advanced core on ARM's roadmap is the Cortex-A57 - targeted in 2014 on the 20nm process - with a SPEC_int 2000 score of about 1250.
And with super-high 2.5GHz clocking (along with plenty of power dissipation to get there), it will maybe scale another 40% or so. So we'll give them credit for 1750. Just to put that into perspective, a Merom based Core 2 Duo running at 2.93GHz scores 3,034.
Sadly, SPEC2k is sorely outdated, else we might have some Sandy or Ivy Bridge submissions likely showing scaling beyond 5k in this benchmark.
So... sorry for being skeptical, but I really don't see ARM getting within a "good enough" mile of Intel's scalar performance in the PC space. Not even when Intel scales that performance down to power envelopes to enable super-thin convertibles and detachable systems.
I'm guessing Even a 1.5GHz 7W Haswell chip would beat a 2.5GHz ARM Cortex-A57 (likely also requiring 7W or more) quite handily.
Market Data 
Markets 
AI
