Apple, with the iPhone, changed the game in mobile devices by pulling together previously disparate elements of architecture, convenience, and technology. Software and services were the keys to new levels of integration, better interfaces and a comprehensive user experience.

The result has lead to a tectonic market shift that combines stunning customer adoption, whole new types of user productivity, a thriving third-party developer community—and mobile and PC market boundaries that are swiftly blurring. Doing the advance work of pulling together elements of the full solution—so that the users or channel players or consultants do not—has worked well for Apple. It was bold, risky, and it worked.

Carriers could never pull off the iPhone integration value for users. Indeed, the way carriers go to market practically forbids it. It took an outsider and new entrant to the field to change the game, to remove the complexity and cost of integration—and pass along both the savings and seductive leap in functionality to the buyers.

With today's announcement of the Cisco Unified Computing System—along with a deep partnership with VMware on software and management—Cisco Systems is attempting a similar solution-level value play as Apple with the iPhone. The solution may be at the other end of the IT spectrum—but the potential leap in value, and therefore the disruption, may be as impactful.

We're seeing a whole new packaging of the modern data center in a way that may very well change the market. It's bold, and it's risky. Cisco—as an entrant to the full data center solution field, but with a firm command of certain key elements (like the network)—may be able to do what the incumbent data center providers—along with the ecology of support armies—have not. One-stop shopping for data centers has been only a goal, never fully realized. In fact, many enterprises probably don't want any one vendor to have such control, especially when standards are in short supply. But they need lower costs and lower complexity.

Cisco, therefore, is using the latest software and standards (to SOME degree at least) to integrate the major elements of "compute, network, storage access and virtualization into a cohesive system," according to Cisco. They go on to claim this leads to "IT as a service" when combined with VMware's upcoming vSphere generation of data center virtualization and management products. I'd like to see more open source software choices in the mix, too. Perhaps the marker will demand this?

The concept remains appealing, though. Rather than have a systems integrator, or outsourcer, or major vendor, or your own IT department (or all of the above) cobble these complex data center elements together—at high initial and ongoing monstrous cost ad infinitum—the "integration is the data center" (as distinct from the network is the computer) has a nice ring to it.

Cisco is proposing that the next-generation data center, then, is actually an appliance—or a series of like appliances. Drop in, turn on, tune in and run your applications and services faster, better, cheaper. Works if it works. This may be too much for most seasoned IT professionals to stomach, but it's worth a try, I suppose.

And this will, of course, greatly appeal during a prolonged period of economic stress and uncertainty. Say hello to 2010. And the approach could be appealing to enterprises, carriers, hosting companies, and a variety of what are loosely called cloud providers. Indeed, the more common the data center architecture approaches across all of these players, the more likely for higher-order efficiencies and process-level integrations. Federating, sharing, tiering, cost-sharing—all of these become more possible to the heightened productivity of the community of participants.

The cloud of clouds needs a common architecture to reach it's potential. Remember Metcalfe's Law on the network's value based on number of participants on it? Well, supplant "node" and participant with "data center" and the Law and the network gain entirely new levels of value if the interoperability is broad and deep.

Make no mistake, the next generation data center business is a very large, multi-tens-of-billions of dollars market, and the competition is global, well-positioned, cash-secure and tough. Selling these data center appliances and "IT as a service" into individual accounts will be a huge challenge, especially if they are perceived as replacements alone. The Cisco solution needs to work well inside, alongside and inclusive of the other stuff, and the integrators have deep claws into the very accounts Cisco must enter.

We'll need to see the Cisco Unified Computing System act as a data center of data centers first. It's appeal, then, must be breathtaking to supplant the frisky incumbents, all of which also understand the importance of virtualization and low-cost hardware.

IBM, HP, Oracle, EMC, Microsoft, Sun, and the global SIs—all will see any market game changing by Cisco as disruptive in perhaps the wrong way. But the enterprise IT market is ripe for major better ways of doing things, just like the buyers of iPhone have been for the last two years.

At the very least, Cisco's salvo will accelerate the shifts already under way in the next generation data center market toward highly-efficient on-premises clouds, complete and integrated applications support solutions, a deep adoption of virtualization—and probably to a lot less total cost, real estate use, and energy demand as a result. The move by Cisco could also spur the embrace of open source software, along with standards, standards, standards. It's hard to see the economics working without them.

Already, Red Hat and Cisco announced a global OEM partnership. Cisco will sell and support Red Hat Enterprise Linux as part of its Unified Computing System, and will also support the newly announced Red Hat Enterprise Virtualization portfolio when it ships.

"Combined, Red Hat and Cisco will offer customers next-generation computing beyond RISC, beyond UNIX, beyond yesterday's legacy solutions for both virtualized and non-virtualized systems," says the statement.

Cisco and VMware are leaders in their areas, for sure, but they will need a community of global partners like Red Hat to pull this off. How about the larger open source universe? Unlike with Apple, it's a lot harder to create a data center support ecology than an app store. So the risks here are pretty huge. The enemy of my enemy is my friend effect may well kick in… or not.

Or even more weirdness may ensue. What if Microsoft wanted in in a big way, given where it needs to go? What if Windows became the default virtualized container in Cisco's shiny new data center appliance? Disruption can be, well, disruptive.

Cisco has been seeking a way for many years now to extend its networking successes into new businesses. It has bought, it's built, and it's partnered—but not to great effect in the past. Could this be the big one? The one that works? Is this the new $20 billion business that Cisco so desperately needs?