Unix market braces for IBM acquisition of Sun
March 18, 2009
The Wall Street Journal has reported that IBM is in talks to purchase Sun Microsystems, a $6.5 billion deal that could give Big Blue a big edge in the enterprise Unix derby. Hewlett-Packard might hope that the deal is only a rumor — or that at worst, it has as much impact as HP's purchase of Digital did seven years ago.
Technically, HP didn't buy Digital outright, but it acquired Compaq, which had bought Digital. By the time Digital products like OpenVMS showed up on HP price lists, their impact was only on the installed base of Digital customers. HP killed off the Digital True64 implementation of Unix and axed one of the best processors ever sold in Alpha. You could argue Compaq did more to hobble Digital's Unix than HP in the three years Compaq owned Digital before the HP-Compaq merger.
The consolidation dance is supposed to help a vendor gain market share. Sometimes the edge shows up when a salesperson can say, "Oh, instead of buying a Sun system, why not move toward our Series p [IBM Unix] line? Unix is Unix, after all, and we'll be melding product lines before too long."
That Unix mind-meld has been more of a fantasy than a dream over the last decade. It's useful to have a Unix application already for sale when you want to offer it on another version of Unix. But unlike Sun's Java, Unix was never "write once, run everywhere." Unix differs too much to ever be the "open system" it promised 20 years ago.
Any assimilation of Sun into IBM is going to cause a ripple in HP-UX server business, however. IBM will pursue Unix enterprise server business more effectively than Sun did. Big Blue will use the same selling edge that HP enjoys: "We offer everything in computing that a company could want, from services to software to systems, of all sizes and capital costs." This could be troubling news for the customer who's hoping that HP can keep its HP-UX market share from sliding. (Industry-standard Windows servers make up all of HP's server growth these days.) Everyone in the 3000 community knows what happens to an HP product which experiences declining revenue growth.
The IBM-Sun deal is far from official. Even the WSJ noted that while the talks are underway, the acquisition may not happen. Neither IBM or Sun officials are commenting, a standard practice while you negotiate the sale of your company to a competitor. The WSJ story claims that HP was offered Sun and passed on the deal, having spent $13 billion on EDS last year to beef up higher-profit services business.
If the deal makes, however, it would mark a milestone far more important than Digital becoming HP. Sun grew up thumbing its nose at the established computing industry, growing strong on the dot-com boom while firms ramped up their Internet and Web resources. Sun was one of the last standing players in the processor derby, too, offering its SPARC architecture until it embraced Intel like nearly all other computer makers.
IBM remains the exception to these conceding companies. Big Blue still does a strong business in its POWER line of processors, sold for the Unix and AS-400 (Series i) systems. The most challenging part of IBM's effort to take in Sun would be combining cultures. Nothing could be more different than melding the culture of a company whose CEO wears a ponytail with the button-down mind of IBM. The clash of wills between Compaq and HP cultures comes to mind. Some HP analysts say the smaller of those two companies's cultures achieved dominance once that deal was finalized.