For those of you who have been hibernating all winter, the most significant tech story since Thanksgiving is IBM's sale of its Personal Computing division to Lenovo, China's largest PC maker and the most recognizable PC brand throughout much of Asia. It's a little simplistic to call it a sale; it resembles more of a joint venture than anything. Under the terms of the deal, IBM retains an 18.9 percent equity stake in Lenovo, which will make IBM-branded PCs for at least five years. And IBM will use this deal as a foot in the door to sell its servers, services, and software to the fastest growing market in the world--China.

But for all intents and purposes, IBM is exiting the Wintel PC business. When the founder of an industry exits the industry, it's big news. And the more I think about it, the more under-reported facets of this story appear.

To paraphrase Sam Palmisano, CEO of IBM, the company chose to exit the PC business because it is mature with a capital 'M', meaning it's increasingly difficult to make money through innovation. Those who win the Wintel PC business will benefit from huge economies of scale rather than from innovative product designs. IBM is all about high-margin businesses that allow for high-end R&D and healthy profits.

The conventional wisdom, spun by The Wall Street Journal, is that it's a little early to exit the PC business. As we have been reporting for almost a quarter century, the PC is often a maddening device. And it continues to leave a lot of room for improvement. It would seem that IBM could continue to innovate by making PCs more secure and reliable, by making files easier to find and work with, and so on.

What the WSJ op-ed writers fail to recognize is that IBM has used innovation to make PCs better, but it hasn't helped sales all that much. For example, its ThinkPad laptops feature all kinds of unique features, such as one-touch system recovery. But they still sell in lower volumes than Dell's and HP's. Innovation gets IBM in the game, but it can't help it win. The only thing that wins this game is low price, which means innovation is too expensive for IBM to fund and still make money.

Other analysts have focused on Palmisano's comments that the PC market is taking on features of the consumer electronics market. They point out that IBM is basically a business-to-business company--after all, Business is its middle name. They state that IBM wanted to exit the PC business because it is not interested in consumer businesses. These analysts are mistaken. IBM is interested in any technology market in which it can make money through innovative engineering and services. Thus, a week before it sold the PC business, it announced the Cell chip, a specialized, Power-based digital media processor that will appear in Sony and Toshiba consumer products later this year. The features of the industry Palmisano alluded to are high volumes and low margins, not consumer versus business.

If you read Palmisano's memo to employees carefully, you'll see that he's committed to continuing to innovate in the PC arena. He points out that IBM is investing heavily in developing the computing platforms of the future, featuring its Power line of processors, open architectures, and open-source developments, such as Linux. He is careful to extend this vision from mobile devices through PCs, servers, mainframes, and supercomputers. IBM is the market leader in those last three types of computers. The news here is that the vision extends down to smaller-scale devices, which currently occupy low-margin business segments. Reading between the lines, it seems IBM is poised to develop new kinds of Power-based personal computers.

Why exit one PC business only to develop new PC businesses? If you can't win the PC game, you have to change the game. Look at the nature of the game itself to see why the current crop of PC technology is an innovation dead end, and to see how you can change the game to once again innovate.

The first thing to note is that the bulk of the money for a new PC goes to two manufacturers: Microsoft and Intel. For IBM, this is an unfortunate remnant of its first PC, which it developed through partnerships with two small companies--Microsoft and Intel, of course--which became huge companies at IBM's expense. After they take their cuts, there's precious little left over for a PC maker. But what if you could develop a PC without Microsoft and Intel dipping into the pot? This is essentially what Apple does. And because Apple controls most of the margins on its PC business, it can still innovate and make money, as is demonstrated by its iMac line. The only supplier that takes a cut of Apple's PC margins is IBM, the developer and manufacturer of the Mac's PowerPC processors.

The second thing to note is that the majority of the innovation for existing PCs can only come from Microsoft and Intel. The latter company controls the hardware architecture. Microsoft controls the software architecture. In this game, all innovation can do is put Band-Aids on outmoded architectures. One-touch system recovery is nice, but the only reason it is needed is because of the inherent instability of running bloatware on overclocked, single-threaded processors.

If you could run your own breed of software under Linux on a cool, multithreaded, low-power processor with graphics and communications on the chip, you could really innovate with the PC. And given IBM's ever-expanding economies of scale with the Power line of chips, it will soon be in a position to offer such a computer at a price point businesses can embrace. One customer that's already using a version of these workstations is the New York Stock Exchange. The workstations are a part of a larger deal in which IBM creates the entire trading infrastructure for NYSE--from the transaction software running on a backend mainframe through the workstations to custom handhelds for the traders. This is just the kind of high-margin, end-to-end deal on which IBM wants to build its future business. And it could not have been as successful for either IBM or NYSE if IBM were forced to use off-the-shelf Wintel PCs and handhelds.

If IBM can develop custom PCs and handhelds for one customer, it can commercialize those products for similar customers. If it gets enough customers, it can begin to offer lower prices through economies of scale. At some point, it can compete with the Wintel price points, while offering much more stable, secure, and speedy systems. This is essentially what it has done with its server business.

The announcements of new markets are coming faster than we can report. As I mentioned, IBM partnered with Sony and Toshiba to create the Cell chip architecture, which will enable all kinds of new consumer devices. How about a digital TV with an integrated PC that lets users pan and zoom on live TV as well as perform real-time editing on paused TV before playing it? Using the same chip, how about a digital camera that lets you edit your photos with a touch-screen, Photoshop-like application before bringing them into a lab for printing? Why not partner with Apple to create new consumer PCs for specialized purposes? Imagine a Tivo-like device that can store a terabyte of programming and that's as easy to use as an iPod.

Once IBM stopped beating its head against the wall by trying to innovate with Wintel PCs, it opened all kinds of doors to new innovations. If I were you, I wouldn't go back into hibernation just yet, even if you do see your own shadow. You might miss out on the remaking of the PC industry.

James Mathewson is editor at large for ComputerUser and a contract software engineer for IBM.