by Martin Veitch

IBM + Sun merger = Unix + processor rationalisation?

Mar 19, 20093 mins
IT Strategy

For the last seven years or so, there have been occasional bursts of speculation about the future of Sun Microsystems as an independent entity. In the first dot-com economic period, the company enjoyed a golden era occasioned by every startup buying a Starfire server as a prerequisite investment in e-commerce but since then, Sun has struggled to reposition itself and has frequently appeared gimmicky, playing around with open-source investments, Linux relationships, software licensing manoeuvres and off-the-shelf processors from Intel. Under former CEO Scott McNealy, whatever question you asked Sun received the same answer: you need the Solaris flavour of Unix running on Sparc processors. That is no longer the case and the company often appears rudderless.

That said, Sun has still been one of the most watchable companies in IT. Depending on innovation and differentiation rather than commoditisation, supply-chain excellence or lowball pricing, it is the polar opposite of Dell. However, reports of a deal for IBM to buy Sun suggest even that legacy attraction might soon be heading south.

Firstly though, are the reports true? The answer is ‘yes, probably’. The report appeared in The Wall Street Journal, an outlet that has an unparalleled record for seeking out technology M&A deals, and it has elicited no straight denials from either party, not even from that source of promised ‘transparency’, CEO Jonathan Schwartz’s blog. The deal could still fall apart, most likely based on arguments over price, of course.

Second, would it be good news for CIOs? The answer is ‘no, probably not’. It’s hard to envisage any great desire on IBM’s behalf to perpetuate Sparc processors and, if it did, that would likely be at the cost of the Power processor family. An eventual rationalisation of Solaris and AIX Unixes would also be likely with one relegated to, at best, organ-donor status. Both outcomes would be net negatives for buyers who would suffer from potential ensuing disruption.

Third, there would also be disruption at other levels. Sun-Oracle, one of the most trusted alliances of the last 10 years, would be vulnerable as IBM would doubtless favour its own DB2 database, although MySQL could be less affected. There is also significant overlap in middleware and of course in people, including sales and technical consulting reps, and support staff.

All in all, IBM would build a dominant position in proprietary Unix servers but do no great favours to Sun customers. However, the deal is not done yet and it is possible that an alternative buyer could emerge, the most likely being Fujitsu Siemens which has a long-running relationship with Sun and has a bunch of key customers that also stand to be hurt by an IBM-Sun combination.