The Observer: Grid Lions

It's odd: In many ways, the computer industry has become a business where advances arrive incrementally, rather than in dramatic leaps. At times, though, the stars align and a single news item jumps out and screams that things are changing in a big way.

So it was, early one morning a couple weeks ago, when upon taking a break from paint pots and spackling knives, I saw a small piece in The Providence Journal about grid computing. There's nothing new about the grid concept--Novell and UtiliCorp United made an early, abortive pass at it, in literal form, back in 1996, and it's been theorized about since the 1960s.

But when I saw the name Irving Wladawsky-Berger mentioned in the story, my attention was considerably piqued. After investigating a little more, I became convinced that this technology promises to be one of the events that will redefine enterprise computing for the next decade.

First, a little background. In its basic form, the grid concept amasses vast amounts of computing power that are, in turn, made available on networks that mimic the basic function of the electrical grid. The grid is to a conventional network as the North American electrical grid is to a single power plant. The latter is perfectly functional, but has limitations of scale; the grid, on the other hand, provides power on a colossal scale.

In theory, those massive resources are made available to individual users when and where they're needed. (For the record, one of the reasons Novell and UtiliCorp's concept fell apart is that they tried to use the actual electrical grid as the backbone of their contemplated network, but that's a post mortem for another time.)

The significance of the grid? Merely the ability for organizations to create a massive, almost unimaginably powerful computing backbone that adds an entire new dimension to the Internet. Moreover, that backbone will allow for as-needed (private or shared) access to huge database files, the instantaneous distribution and/or sharing of applications, and collaboration among individuals and organizations on a scale that can't be achieved today.

Perhaps most significantly, the grid, by definition, must be an open environment. To be sure, individual vendor organizations will have stakes in its propagation and success, but it will be beyond the undue influence of any single vendor.

And that's where seeing Mr. Wladawsky-Berger's name in the story crystallized things for me. He is IBM's vice president for technology and strategy. He also is, in my opinion, one of the few people in the world who really and truly gets it when it comes to technology and its potential.

In more than 30 years at IBM, he has overseen, among other things, the revitalization of IBM's mainframe business, the growth of its workstation division, and the refinement of its E-business strategy. More significantly, he is one of IBM's most vocal, most visible proponents of Linux. If there's anyone, anywhere who understands logical next steps and the linkage of timing and theory in a successful initiative, it is he. In sum, if anyone can make the grid workable and valuable for commercial interests, it's Irving.

The Web was born of initiatives that had their roots in government, military, and university research centers. The grid concept is following the same path. In the United States, the Departments of Defense and Energy, NASA, and the National Science Foundation all have grid projects underway.

In Europe, things are even further along (so far as we know, given the pedigree of some of the U.S. "gridders"). No accident, then, that IBM is actively involved in building Britain's national grid, which will link nine research centers, and that those centers are being linked based on an open-code model.

None of this is to suggest that grid computing will be ubiquitous in weeks or months. The incremental process, after all, takes time. But Unilever, Pfizer, Glaxo, Smith-Kline, BMW, Ericsson, Hitachi, and others aren't willing to wait. All are funding grid initiatives and others will surely join the list in short order.

The business of holding a finger up to the wind and trying to divine which way it's blowing is an inexact science. But as was mentioned at the outset, sometimes events conspire to make prognostication a touch easier. This, plainly, is one of those times.

The grid, quite unmistakably, is upon us. The challenge now becomes dealing with the questions of the whens and whys for its widespread commercial use. What's beyond question, though, is that the grid will--if it hasn't already--redefine the landscape.