Today at BUILD Intel announced several new initiatives and partnerships aimed at improving their position in mobile computing. They promised Medfield tablets and phones, a stronger focus on low wattage cpus as well as a commitment from Google to make Android sing on x86. Thing is, none of it matters.
Pundits and industry watchers seem to be of the opinion that Intel’s failure in mobile so far rests entirely on their inability to make a competitive chip. That the power requirements are too high and that the software support wasn’t there. The corollary of this is that, if Intel were to correct these things, they would be able to take on ARM. The problem with this line of thinking is that it ignores ARM’s greatest asset: the ARM ecosystem.
The ARM ecosystem is a Utopia. If you want to make a product, you have dozens of chip vendors willing to sell to you, all priced competitively in order to garner you interest. The diversity of SoCs is incredible. If that’s not enough, you can work with one of these vendors to help create an even more optimized solution. Furthermore, you even have the option to license the ARM instruction set and design references and just make your own chip. You are never tied down to a single vendor and never stuck waiting for the next yearly wave of updates. These are things that Intel can never match, and no OEM that has tasted paradise will be willing to leave.
The fundamental problem with Intel is not its chips but its business model. Without vendor lock-in via Windows or some other meaningful form of integration, a single vendor solution can never compete with a free, functioning and diverse market. Merely being competitive on performance is not nearly enough.
rvqs 