Has anyone (not under NDA) actually used WOA? The promises quoted sound pretty much the same as promises made for Wn95, 98, CE, Me, XP, Vista, 7...
Until it's proven, I think it's pretty premature to be saying that Intel's in for a heap of trouble. It may end up that various ARM providers give Intel a run for their money, but based on history, I'd guess that a genuine challenge is still a few WOA versions away.
Given Microsoft's track record of running Windows on Intel x86, I doubt that Windows on ARM would be a great user experience. Sure, battery probably lasts longer, but the user probably ends up waiting longer for an app to finish.
Until Windows on ARM has similar performance as on x86, it is too early to conclude that Intel is done.
Application dependency on specific CPU architecture is at real risk these days. If the host OS can handle virtual environments, there is no real reason that all of those environments have to use the same instruction set. Java's VM was a good start in that direction. If the OS can provide a common platform on a number of CPU's then it can more completely commoditize the CPU.
This is significant step MS has taken towards shrugging off x86 bottle neck and dependence for its growth. This should bode well for MS in the mobile segment in the next 2 years. As far as ARM is concrned it is well making the most of the prevailing situation and seems quite focussed on where it wants to be.
Intel will not feel any heat until this marriage succeeds and even then still not for a few years.
Blog Doing Math in FPGAs Tom Burke 24 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...