Dana Blankernhorn writes:
Intel’s decision to turn away from straight-ahead development of its Pentium IV and Xeon lines, in favor of putting all its eggs in low-power chips, is a big, big deal.
For starters it’s another illustration of Moore’s Second Law, which holds that as chips get more complex they get more expensive to make. Even Intel can’t do it all any more.
Intel has decided to go all-out on low power. The laptop metaphor, and all that follows it, means that Intel is leaving the electrical grid behind. Instead the grid will become back-up power. Batteries will now matter more.
This is also an important endorsement of Always-On. I’ve written here many times about how wireless LANs, as an application platform, need a battery-powered “black box” at their center.
This is where Intel is moving, toward a low-power platform that is still expandable. The “PC” at the heart of its future can’t be a desktop, because this announcement has closed that off, and it can’t be a laptop, because laptops aren’t expandable.
Some new central reference design is needed. More likely, several such designs will come to contend to the future. I think the interface-free black box, separating the interface from the workings of the device, is a contender in the home market, but we’ll see.
What Intel also needs to do is $50 thin clients and sell them for $1-2 a month. Computing has to become a service and Intel, along with Microsoft, needs to recognise this.