A Merger that Happens Once in a Big Blue Sun
The news broke this morning, March 18th, that IBM is talking to Sun about buying the company. Sources from both companies tell me that such a deal is in the works and it may be completed as early as this week.
Sun's pricetag may be as high as $6.5-billion with a large part of the deal being made with IBM stock. Sources indicated that what IBM wants is Sun's software businesses, not its x86 and SPARC server lines.
There are three possible hardware plays here. One is that part of Sun will remain as a hardware server firm. Another is that IBM will go ahead and buy the hardware line and merge Sun's x86s into its System x line and place SPARC within its System p division. Finally, some or all of Sun's hardware may go to a third company, such as Fujitsu, which is already in the SPARC business
Of course, in deals like this, there's many a slip between the cup and the lip. Still, I, for one, am sure that the deal will go through. While Sun CEO Jonathan Schwartz recently pointed out Sun's strong points--"over $3 billion in cash, and a nearly two decade history of generating positive cash flow. We've also got a set of technologies and people that continue to play an ever more vital role in the economy."--he didn't really address Sun's serious problems.
What Schwartz didn't mention was that Sun had laid off from 15% to 18% of its employees in November 2008, that's between 5,000 and 6,000 employees, after a quarter which saw a $1.68 billion quarterly loss. Even before the economy started its nose-dive, Sun had been bleeding red-ink for several quarters.
Things haven't changed for the better for Sun since then. In its latest quarter, even after the mass layoffs, Sun's revenue dropped 11%. Sun reported a net loss for the quarter of $209 million. With red-ink like this even $3-billion won't last forever.
What's more important is that Sun embraced open source too late to reap its benefits while at the same time it was saddled with its expensive SPARC server line. When the few companies that are still buying server hardware are buying x86 servers and blades, there's simply not enough buyers left for Sun.
If the deal does come through, this is what I see happening. First, I don't see IBM's support for Linux wavering for one moment. Linux has been very, very good for IBM. That said, IBM has already shown that it has no problem with supporting its own house brand of Unix, AIX. Because of this I'm sure IBM won't mind supporting OpenSolaris/Solaris for these operating systems' existing customer base. In addition, by certifying Solaris on IBM's BladeCenter, IBM may end up giving Solaris a new lease on life.
And, if former Sun customers do want to move to Linux? Then IBM, which just bought Transitive, which specialized in moving Solaris programs to Linux, is ready to help them make the jump.
I've heard some people wonder why, with all of IBM's interest in Linux, "Why are they buying Sun, when they could buy Novell?" The reason is simple. IBM has never had, and continues to have no interest in being a Linux distributor. IBM Global Services, which delivers integration, support, and service to Linux customers, is how IBM makes it billions from Linux. IBM is perfectly happy to let Red Hat, Canonical,and all the rest have the distribution and first-line support business.
IBM will also get more control over the powerful Java programming language. IBM has long supported Java through Eclipse. Encouraging developers to use Java on Linux and the AIX/Solaris family will only help IBM keep Microsoft out of server development circles.
Sun mishandled its new acquisition. MySQL's top programmers left. MySQL's founder and still Sun CTO for its MySQL division, Michael 'Monty' Widenius said publicly that the GA (general availability) version of MySQL 5.1 was filled with bugs.
With the right management, say IBM's DB2 division, though, MySQL should still be a star. MySQL is the DBMS (database management system) of Web 2.0 businesses such as Twitter, Facebook, and WordPress.
By now, you've gotten the point. There's a lot of good stuff inside of Sun, which would work hand-in-glove with IBM's existing projects. As for Sun, they've been unable to fend off Linux and commodity servers in its server lines, and the company has also failed to successfully monetize Java, OpenSolaris, OpenOffice or MySQL.
Frankly, I'm only surprised that it's taken this long for a deal like this to materialize.