So, when will Oracle buy Brocade?

Well, they might buy Juniper, Alcatel or maybe even pick-up Nortel on a distress sale, but I’m pretty sure they will jump on Brocade to secure their data center dominance.

Several vendors have been assembling stacks in attempts to become one-stop shopping centers for CTOs.  Cisco shocked many by getting into the server business, HP by buying 3Com, and Oracle buying Sun and not realizing the mess they had gotten themselves into.  IBM is oddly the laggard, evidently content to rake in huge margins on services instead.

But IBM may well buy what Oracle doesn’t.

Let’s ignore applications, if for no other reason than to give Larry Ellison a reason to panic and read this blog.  The major components of the data center infrastructure stack include servers, storage, operating systems, management suites and networking.  HP has all of the above, and wedding ProCurve, 3Com and Openview will possibly put them ahead of Cisco as a favored network gear vendor.  Cisco has most of the stack sans storage, though not many folks are jumping through hoops over Cisco’s server solutions.  Oracle is missing the networking piece almost entirely, and given their new market position for servers and storage, not having data center networking gear seems to be a hole in their whole.

Whole product definition that is.

This is a rather gaping gulf for a company that, upon completing the Sun acquisition, said they were the 1960’s IBM of the new millennia.  Given how their competitors are maneuvering, Oracle will eventually need to get into the networking gear business, though they likely will focus only on data center bit pipes.  As with SPARC and x64 systems, Oracle is optimizing gear for the huge database and application servers many enterprises need.

So the question is which network hardware vendor has an adequate product suite for Oracle and can be had for a good price. Many alternatives are simply too big or have too many non-data center products to sooth Ellison’s digestion.  Juniper has a market cap equal to Oracle’s cash, so they are out.  Nortel could be had on the cheap though they remain radio active and have a lot of telco products that Oracle wouldn’t need or want.  Extreme isn’t, lacking core technologies essential for mega servers.  That leaves Alcatel and Brocade, with Alcatel being twice as expensive mainly because it is holding a lot of cash.

So Brocade has the best product line and price match.  With most of that nasty little backdating scandal is behind them, they don’t present unreasonable risk. The only real question is if IBM will move before Oracle or the other way around.

Market trends and marketing go hand-in-paw.  The trend in IT is consolidation, at least in the hardware space.  Some of the big boys (Cisco, HP, Oracle) are attempting to offer solid stacks.  This trend cannot be ignored, and the eventual winners will be those who offer something beyond hardware.  After all, in an x64 and Ethernet world, raw horsepower differentiation is disappearing.  Other factors, such as service, support and reputation will sway buyers (and herein, HP may have a great advantage since nearly everybody believes they make unbreakable hardware).

The wild card — as always — is Oracle.  They dominate the application space, and since hardware only exists to run applications, this factor cannot be ignored.  Early Sun integration shows an Oracle propensity for optimizing the stack between components.  It remains to be seen how obscene application-to-hardware optimization might become.  Aside from lower prices (which means lower margins, which means an instant veto from Larry) application/hardware optimization becomes a core hardware differentiator.

If When Oracle buys a network gear vendor, we may see the largest new barrier to entry for anyone in the enterprise IT space since IBM dominated everything.  Innovators and start-ups will need to find ways of making their products a strategic add-on to HP, Cisco, Oracle product lines with acquisition exit strategies.  The days of enterprise IT IPOs may be behind us, though the future of corporate venture capital is looking mighty good.