CES 2010 The Story So Far: 'I'm a PC' is in jeopardy

If you're a 2.0 GHz quad-core notebook PC running Windows 7, the place you do not want to be showing yourself right now is the Consumer Electronics Show. Like any other year, CES is about platform evolution, sometimes revolution. Where the most concepts evolve is where the action revolves, and this year it's around the growing confidence that manufacturers have finally found a small form factor they can invest in and make money on.
Netbooks, throughout their brief history, have been cheap, rubber-stamp PCs. That won't be the case for very long, as both Intel and Qualcomm now have argued strong cases in favor of multicore connectivity devices with limited, though functional, computing power for the consumer. There will be strong competition in this field, with Intel's Atom head-to-head in the consumer space against Qualcomm's Snapdragon, and with OEMs everywhere testing both brands and favoring neither over the other at this early stage.

What makes these platforms lucrative is that OEMs can keep their handle on them after the devices are sold. This will especially be the case for netbooks running Google's Android and Intel's Moblin flavors of Linux. During his evening keynote yesterday, Intel's Paul Otellini demonstrated the developmental state of Moblin by way of its app store, which Intel is calling "AppUp." OEMs will be able to put their own flavor on these stores, however; so Dell, for instance, will have a different storefront from HP, even though many of the same apps will be available both places.

Dell's flavor of AppUp storefront, part of Intel's beta program for Moblin apps shown during the Thursday evening Intel keynote.

Intel's "fusion" (shamelessly borrowing the metaphor from AMD) of its integrated graphics chip onto the same die as its CPU, for consumer-grade Core i3, i5, and i7 models, is that company's effort to further cement its consumer processors into a cohesive platform. But by driving costs down, and thereby prices, Intel kicks value-priced PCs more into the netbook space, effectively making an i3-powered system a "premium" or "best" alternative to an OEM's Atom-based "midrange" or "better" version. That means Core may play with Atom in the Moblin space as well.

Despite the fact that Windows XP is the most installed operating system on netbooks today, with Windows 7 catching up fast, this week's news from CES is very, very bad for Microsoft. During the heady days of the 1990s and early 2000s, when it appeared Windows could be leveraged onto every device with a CPU as far as the eye could see, Microsoft relied on its relationships with Intel and AMD to cement itself as the operating system provider for the x86 platform. But that's when there was something called "the x86 platform," and that's not the situation today.

Yesterday, Intel presented its vision of the x86 platform in the consumer space, and it was not running Windows. The AppUp app store is not for Windows; and for the devices you use to play movies in your car, get directions to the meeting place, find the best prices for shoes, and set the thermostat for your house, Intel wants customers to go through AppUp. There is no equivalent for Windows, not yet; and even if Microsoft comes up with a similar environment for Windows Mobile 7 next month, it won't (immediately) extend to Windows 7 for PCs.

The best that Ballmer could come up with: a ten-second peek at an HTC phone that already exists, the HD2 running Windows Mobile 6.5.

Perhaps more importantly than all that: "Apps" as a term is taking on a definition that is separate from "applications." On new, smaller, more flexible, and more portable platforms, an app is representative of an everyday life transaction: finding that pair of shoes, getting a stock quote, checking out the fishing at area lakes. Yes, the Apple iPhone started that trend, but the number of followers is increasing rapidly. Windows does not appear to be among them; it's still the platform of "applications," of Word and Excel, of getting business done. And it may still be that platform for many years to come, but that's a whole lot of no fun. It keeps Microsoft bottled up in the realm of enterprise licensing and servers, which is still profitable but not enough for a company that has invested as much as it has in recent years on developing a consumer-oriented brand.

Ironically, it was Microsoft's Ray Ozzie who pronounced last November at PDC that platforms will come to be defined in consumers' minds more by apps than by operating system brands (a statement that was widely misquoted elsewhere), and that consumers will choose an OS based on what apps it runs. Without a viable retail model for making everyday apps visible, as opposed to conventional applications, Microsoft is exposing a widening cavity in its product portfolio -- a vacancy that includes Windows Mobile 7, but which now extends into its core competencies.

Which makes the absence of WM7 news at CES this year an indicator of a much broader problem for that company, but also a serious shift in the axis around which all technology revolves. The last time I remember Microsoft being so widely discussed as an also-ran needing to catch up to Apple's lead, we were running VisiCalc on an Apple II. Maybe most importantly of all, however, with consumers likely to enjoy huge benefits in the coming months, perhaps the only folks who see this as a problem will work for Microsoft.