Business Technology: Grid Computing: Too Big To Be Ignored

Here's a question that I suspect most of you won't have to think too terribly hard about: If you could spend $100,000 on an application of emerging technology and realize savings of $21 million directly from that investment, would you pull the trigger? Or would you refrain because the underlying architectural strategy is still not 100% proven? Or would you just say that you wouldn't spend it because this anecdote clearly falls under the category of "if it seems too good to be true, then it probably isn't true"?

Would you believe me if I told you the emergent architecture is grid computing? Or would it sound too much like the latest version of bubble-memory-based supercomputers in every pot?

I like to think that I am, simultaneously, not only skeptical of wild new proclamations of spectacular breakthroughs but also not jaded enough to be completely close-minded. And I'd be pretty darned skeptical of the whole story if all this had come from the mouth of the hypertalkative person in the seat next to me on my last plane trip who was introducing himself and trying to shake my hand while I was putting my suitcase in the overhead and who kept talking as I got off the plane and went into the nearest rest room as he handed me a paper towel and who didn't shut up until I shut the taxi door in his face (where do these people come from?). But no, my source for this and a lot of similarly optimistic discussions of the grid philosophy was not some traveling lunatic--rather, it was a reseller partner of Oracle, and his general impressions were endorsed with great enthusiasm and real-world clarity by two of Oracle's customers moving toward the grid.

In addition, Oracle itself is remarkably bullish on the technology--so much so that it's shifting the letter used in the names of its database products from "i" (e.g., Oracle 8i, 9i, etc.) to "g" in the forthcoming release of 10g. That naming convention shouldn't be taken to mean more than it's supposed to; on the other hand, Oracle didn't make this decision lightly.

And the Oracle reseller mentioned above stressed that the move to the grid shouldn't be taken lightly, either. Ian Baird, chief business architect at Platform Computing, said that "it's key to remember that grid computing is built, not bought," and that customers should start with one app in one department and then accelerate. But Baird pulled no punches in front of a few dozen financial analysts about the longer-term value grid computing holds for customers. "Dozens of our customers are moving to grid computing," he said, noting that J.P. Morgan Chase has "lowered its hardware expenses by 40%" since taking the plunge. Many others, he said, have cut hardware expenses by 50% to 60% and are realizing big savings in manageability, overhead, and other areas. As for the company that stands to save $21 million by spending $100,000, Baird would say only that it's in the auto industry.

Oracle customer Frederick Limp, a professor in the Geosciences and Environmental Dynamics department at the University of Arkansas, added this enthusiastic perspective: "When Oracle got behind grid," he said, "I thought it was like the move from Mosaic to Netscape." And the result, he said, is that grid computing has reached a "tipping point of acceptance" among large organizations whose databases must handle huge files from disparate sources and whose underlying architectures must be secure, scalable, and increasingly cost-effective.

Now, we're all old enough to recall some of the overhyped next-ware can't-miss silver bullets that somehow turned out to be more like tinfoil. But grid computing seems to be racking up some numbers that can no longer be ignored.

Bob Evans,
Editor in Chief
[email protected]

To discuss this column with other readers, please visit Bob Evans's forum on the Listening Post.

To find out more about Bob Evans, please visit his page on the Listening Post.