The IT industry is going through a historic realignment. Product vendors want to be service vendors, and those without good services arms are seeing too much margin pressure to compete. But in the rush away from product specialization, I think some major players are misreading what the market wants -- or more precisely, misreading just how big certain markets really are.

The most recent to venture down this path is Brocade, which is at least rumored to be up for sale. The storage specialist had looked to broaden its networking footprint by buying Foundry, but that acquisition was too little, too late. Brocade's main customers are storage vendors, and in a world where Hewlett-Packard buys EDS and Dell buys Perot, it's unlikely that Brocade will ever have a solid relationship with enterprise buyers, so why not sell the company to someone who does? HP seems the odds-on favorite to buy Brocade, but don't count out Dell, which wants to extend its enterprise service footprint in a bad way; Juniper, which has cash and whose product line lacks pretty much everything Brocade sells (its nascent switching line notwithstanding); and EMC, which, while a longer shot, has a good enough history of preserving the independence of acquisitions that competitors still do business with companies it buys. Oracle is a possibility, since Brocade would fit with Sun once that acquisition is completed, but Oracle simply isn't a hardware company.

For all these vendors, it's pretty much a case of IBM envy. Big Blue decided it wasn't up for low-margin businesses that didn't "make a difference" and sold its networking and end-user systems divisions. It then went on to concentrate on its global services arm, where it not only makes fat profit margins, but it also sells lots of its own enterprise software with associated fat margins. While IBM didn't want to be in the networking business, that's central to the enterprise data center, and thus why HP, Dell, EMC, and Juniper see great value in adding Brocade's product set.

So even as those companies rush to crash IBM's party, I think they're misreading the overall appetite for service-based enterprise systems and software offerings. While I'm not generally a fan of cloud computing in all its incarnations, I think that software-as-a-service providers will get a huge boost from this consolidation. When you have a hammer, you see nails. When you have 10,000 consultants, you see big, expensive integration projects. SaaS is going to be a popular alternative

See our report:
Data Center Services

Register to see all InformationWeek Analytics reports

The other likely winner here is Microsoft. Windows 7 seems to be what Vista should have been, and for a large percentage of sticker-shocked CIOs and CFOs, the ability to buy Exchange, SharePoint, a pile of enterprise apps, server software, and endpoint operating systems all from one place, all integrated with Active Directory, and all deployable without a hoard of integrators ... well, let's just say that's going to sound darn attractive. Throw in "just good enough" Windows 7 functions like BranchCache, BitLocker, AppLocker, DirectAccess, and even XP mode, and the more contrite Microsoft looks pretty smart and like a pretty good deal.

Art Wittmann is director of InformationWeek Analytics. Write to him at [email protected].

To find out more about Art Wittmann, please visit his page.

Register to see all reports at InformationWeek Analytics.