The IBM-Sun Saga Continues--Or Rather, Doesn't

Published: April 13, 2009

by Timothy Prickett Morgan

Well, so much for that idea. About a day after I wrote about the impending acquisition of Unix server maker Sun Microsystems by IBM for last week's issue, the deal came apart. And why? Apparently--and we all have to say apparently out here because no one has said anything officially--because the two parties could not agree on money or the firmness of their commitments.

This seems to be a typical bad engagement gone worse before the nuptials are even exchanged. Which is when you want things to go bad, particularly if you are most Sun employees, people who want to keep their jobs and who want Sun to remain a standalone company and compete in the IT marketplace. I think at this point, even if Sun and IBM do come to some sort of agreement, there pretty much will not be anything resembling a honeymoon.

The longer the talks between Sun and IBM have been going on, the lower the price IBM was reportedly offering to acquire its rival. The original story in the Wall Street Journal that broke the news on March 18 indicated that IBM was initially offering somewhere between $10 and $11 per share for Sun, and then by late Thursday (March 26), the price had slipped to $9.50 or $9.55 per share according to reports in the Journal and elsewhere.

Late Sunday, after last week's issue of The Four Hundred had gone to bed, the Wall Street Journal, once again, broke the news that the deal was possibly falling apart, and not just because IBM was thinking of offering somewhere between $9.10 to $9.40 per share for Sun, but because IBM didn't have enough conditions in its offer to make Sun feel comfortable that it was truly committed to the acquisition, including fighting antitrust regulators in the United States and Europe who might object to the deal.

There was some talk that the top brass at Sun were not pleased with their golden parachutes, which is just so uncouth right now given that whole AIG bonus mess. If Sun has to sell itself to survive, what these managers deserve is a doorknob hitting them in the ass, and quite hard, on the way out. And maybe some clawbacks on performance-based bonuses and stock options before the doorknob makes its bruise.

By Monday morning, the Journal was reporting that IBM had actually made an offer on Saturday for $9.40 a share or lower (yes, I know that is vague), and that Sun's board had split into two factions. Those who wanted to do the deal, led by president and chief executive officer, Jonathan Schwartz, were outvoted by those who did not want to do the deal, who were lead by chairman and Sun co-founder, Scott McNealy. Sun's shareholders--particularly the equity firms and big pension plans who own sizable stakes in Sun--are going to be furious that Sun didn't take the deal, particularly considering that on March 17, Sun's shares were trading at $4.97 a pop and the company's stock was trading at $8.50 a share before the deal went south two weekends ago. The stock market was closed in New York on Good Friday, but Sun's shares had fallen to $6.61 by the close of the market on Thursday (April 9), giving Sun a market capitalization of $4.97 billion. It is hard to believe Sun's stock has fallen this low, but the fact is, it has been lower. Like $2.60 a pop back in November 2008, making Sun worth $1.9 billion, or just a little more than the cash it had in the bank. I said it before, and I will say it again. Sun should have taken itself private right then and there and then told Wall Street to go jump off the same cliff they took the economy over in 2008.

But Sun's top brass are too proud for that, and they believe their own optimism despite evidence to the contrary. And now, Sun can't afford to get Wall Street off its back. The wonder, are far as I am concerned, is why Sun's share price hasn't fallen to $2 a pop, given the fact that no one wants to buy it and the IBM deal is going nowhere. That the shares are holding up at all says that investors still think Big Blue is going to do the deal, somehow, somewhen.

To make matters worse, on Wednesday, Sun confirmed that Marc Tremblay, the chief technology officer of Sun's Microelectronics division within its Systems Group, had left to take a job at an elite unit of Microsoft. Tremblay was the co-architect of the UltraSparc-I processor, a breakthrough product for Sun that put it in the server market for real in 1995, and was the designer of the UltraSparc-II processor that made Sun the dot in dot-com. Among a number of disappointing projects, Tremblay was a key designer (among a handful) for the "Niagara" Sparc T series of multicore, multithreaded chips that Sun has built a $1.4 billion business around and was the key designer of the forthcoming "Rock" UltraSparc-RK processors, which pack 16 cores and 32 threads into a single chip and which include some innovative technologies. Rock machines were supposed to ship in the second half of last year (earlier if you want to be honest), and were delayed until the second half of 2009. Now, everyone is wondering if Rock chips and their "Supernova" servers will see the light of day, with Tremblay leaving. Without knowing anything, I would guess that the Rock chips are nearly done, the Supernovas will ship later this year, but Sun is calling into question its ability to continue to design chips like Rocks. The trouble is, there may be no Rock-II chips, which would compel Tremblay to leave, perhaps and that will make it hard for Sun to sell Rock-I systems, no matter how much bang for the buck and bells and whistles they offer.

Tremblay, according to Microsoft, has been tapped to join Microsoft's Strategic Software/Silicon Architectures group, and will report to KD Hallman, the unit's general manager. "Marc will help oversee cross-company technical task forces and strategic direction for the company's software and semiconductor technologies," Microsoft said in a statement they emailed to me. It seems unlikely that Microsoft will design its own chips for servers, but then again, it did create a three-core variant of the PowerPC Cell processor, code--named "Xenon" and known as the XCPU, for the Xbox game console. It wouldn't surprise me at all if Microsoft suddenly decided that it wanted to create Xserver appliances with various Windows Server features embedded and hardened. If I was Microsoft, that's what I'd do.

RELATED STORIES

IBM Poised to Buy Sun, Rumors Say

Wall Street Makes IBM, Sun Strange Bedfellows?

                     Post this story to del.icio.us
               Post this story to Digg
    Post this story to Slashdot