Doug Siebert (foo@bar.bar) on 1/26/12 wrote:
---------------------------
>Honestly I think Intel's biggest problem is going to be their decision to support
>translating ARM code. If it performs poorly, it makes Intel phones look crappy
>and they become what Cyrix was in the early 90s. If performs good enough, then
>there's no incentive for anyone to ever write x86 versions of their apps. Intel's
>process advantage would always be totally negated by the translation overhead.
>Intel is about two years too late, and IMHO the best they can possibly hope for
>is to be the AMD of the smartphone market. Even that is only possible if Windows
>Phone becomes a strong third player in the smartphone market - I think they'll be
>third but probably not a strong third. Basically, third only because RIM will probably
>be bought out (possibly by Microsoft) within 18 months for their patents and Exchange integration technology.
>
>

I agree that Intel's decision to support ARM code through emulation carries with it significant risks, especially if it eases off the pressure on developers to provide native x86 binaries (for those applications that contain native ARM code today). However that could be alleviated through the leverage that Intel seems to have on Google and the vast resources that both companies have available to sift through the applications on Android Market and translate the native code in them (with the objective of creating a fully transparent experience to the user as far as different ISAs go). That is as long as Intel takes the problem seriously, which I suspect they do.