At first glance, a story about Dell supplying the computers to power a 2,008-system cluster at the State University of New York in Buffalo appears quite remarkable. When a company maker better known for its business model than its R&D scores a major victory in the clustering arena, that's big news- but not for the reasons you might think.

In the big picture, this is not a story about how Dell has suddenly and miraculously turned into a clustering player. Given Dell's model, the transformation was inevitable. If the company didn't push itself there, it would have been sucked into the space anyway.

This is yet another episode in the continuing saga of the fall from grace of proprietary technologies, the commoditization of processing power, and the difference between the must-haves and the nice-to-haves in budget-constrained times. It's also a signal that the high-end technical computing space is no longer a sanctuary for vendors of premium-priced boutique systems.

With the economy in the toilet and HP and Dell fighting tooth and nail for every one-hundredth of a percent of share in the Intel-based server market, Dell is properly gloating over its big win. The SUNY Buffalo victory came directly on the heels of another multi-million dollar victory with RackSpace, a company known for its Linux-based Web hosting business.

In the face of growing criticism over a merger-induced slump, HP has countered with stories about recent successes like the one with General Mills and a year-old success with the U.S. Postal Service.

Even Gartner and IDC, each of whom has researched conflicting market share numbers, are caught in the credibility crossfire. While IDC's last Quarterly Tracker put Dell on top of HP (29.7 percent vs. 29.6 percent) in the U.S. Intel-based server market share column, Gartner says it's the other way around, with HP at 28.7 percent and Dell at 26.7 percent. The race would not be nearly as close were it not for the HP-Compaq merger.

As the two companies round the far-corner of the racetrack, deals like SUNY Buffalo or General Mills can lower the chances of a photo-finish at the end of each quarter. As insane as it sounds, such bragging rights matter to potential customers. But, to me, Dell's victory in a clustering configuration may be an early warning to HP (and third-place IBM) that the frills typically associated with their systems are becoming less important. More important now is the ability to consistently deliver cost-competitive, reliable systems based on relatively standard (or in the case of Intel, ubiquitous) and extremely low-cost building blocks such as IA-32 and Linux.

Poor man's supercomputer
HP is well aware of this principle. About a year ago, HP partnered with scientists in France and built a cluster using Linux and 255 stripped down, low-cost HP PCs (rather than a bunch of tricked out servers). That cluster's benchmarks earned it a spot in the world's top 500 supercomputer list. SUNY Buffalo's $4 million cluster, which one of the university's professors called a poor man's supercomputer, is expected to make it onto that list as well. Competing supercomputers cost many times more. Sooner or later, in the face of workable alternatives, the CFO is going to ask for the extra expense to be justified, and that won't be easy.

Already, Dell's conservative model of producing systems for high-volume markets by integrating standard parts was working its "magic" on the regular Intel server market. Given a cluster's ability to solve the same problems (i.e.: performance, fault-tolerance) that systems with expensive proprietary architectures like IBM's EXA or HP's ZX/1 can solve, I'm beginning to wonder about the value of such boutique technologies.

More importantly, Dell's commanding pre-HP/Compaq merger lead was built primarily on non-Linux operating systems. But now that clusters are available that run Linux (an OS to which most Unix applications are easily ported), the higher, compute-intensive end of the market will start to buckle under the same pressure that the rest of the server market is already feeling. That pressure comes from the commoditizing effect that low-cost, high volume, standard building-block based systems have on processing power. And Linux is one of those building blocks.

Dell's cluster success is a story about how standard building blocks are heading upstream and why, if it's good for the guys who build clusters to rely on as few proprietary technologies as possible, it's probably good for you too.

Where is your head at? When it comes to building high performance high availability systems, will you pay the premium for the secret sauce in boutique systems? Or, do you see yourself going the more standard, building-block route by turning to vendors like Dell? TalkBack below or write to me at david.berlind@cnet.com.