What I'd give to be a fly on the wall in IBM CEO Sam Palmisano's office right now. Assuming he's the one who ultimately made the decision to play hardball in IBM's talks to acquire Sun Microsystems, he must be wondering how he could possibly have let Sun slip through his fingers and into the clutches of Oracle CEO Larry Ellison. When he got the word that Sun had eloped with Oracle and secured a $7.4 billion dowry, he was probably livid at having been left standing at the altar.
You can bet that if Palmisano had it to do over again, he wouldn't have been so smug when he was haggling with Sun over the price. He should have known that Ellison was lurking in the shadows. After all, Ellison and Scott McNealy, Sun's chairman and former CEO, have been friends for years, and it is quintessentially Ellison-esque to seize the spotlight with a dramatic, landscape-altering, scene-stealing appearance.
[ InfoWorld's Neil McAllister correctly predicted Oracle's Sun takeover. Find out why he thought the deal would make sense. | For full coverage on the Oracle-Sun deal, see InfoWorld's special report. ]
IBM and Sun could have made beautiful music together, with IBM's global services strength making Sun's impressive work in the realm of cloud computing strategically viable for corporate IT. Instead, Oracle will gain that intellectual property and will most certainly capitalize on it by reviving Ellison's Network Computer dream.
I still remember a conversation I had with Ellison in 1997, when I asked him what was the smartest business decision he'd made in the previous two years.
"I suppose the introduction of the NC -- the network computer architecture," he responded. The reply would become a standing joke, because Ellison failed miserably in getting corporate IT to adopt the NC, and it quickly died an inglorious death. But he was clearly on to something.
Ellison recounted a meeting he'd had with then-President Clinton about bringing low-cost computers to market, with McNealy and Apple's Michael Spindler also present.
"I asked the President to challenge our industry to build a computer that's inexpensive enough to put on the desks of all children. And we just picked a number -- $500 seemed like a magic threshold," Ellison said. "And Scott McNealy got very sarcastic and said, 'The heck with $500, how about $200?' Shut up, Scott."
But what McNealy understood was that it wasn't in the DNA of enterprise IT vendors like Oracle and Sun to market low-cost computers for school kids. The Network Computer needed to be a thin client for corporate networks. Two years later, in September 1999, Sun introduced its Sun Ray thin client. And that's what will fulfill Ellison's hardware dream.
Sun's server business does absolutely nothing for Oracle. Unix servers are a dying breed, and Intel/AMD-based servers are a commodity. Oracle will almost certainly ratchet that business down as quickly as possible, or sell it off outright, perhaps to Fujitsu. But the thin-client business is a different story. It's not difficult to see that the future of corporate computing lies in the cloud, and that a thin-client architecture will seed it.
When it does, it will no doubt rain on Sam Palmisano's parade.