More on Sun vs. IBM: Why destroy SPARC

To really understand why IBM might want to cripple Sun through take-over talks all you need to know is that a Sun 5440 will outperform a P570 by about 2:1 on most business tasks - and costs roughly ten cents on the IBM dollar, inclusive of storage and RDBMS licensing.
Written by Paul Murphy, Contributor

As I write this (Sunday March 22) I don't know whether IBM will commit to buying Sun or not - but by the time this runs on the 26th things should be a bit clearer.

What's very clear already, however, is that there are no significant synergies between Sun and IBM: meaning that any IBM purchase of Sun can only be motivated by IBM's desire to shut the upstart down - after all, destroying SPARC would give IBM an effective monopoly on RISC server sales at prices that are very nearly ten times Sun's per unit of throughput delivered.

Notice too that a deal held up in the courts would in many ways be more effective than an actual purchase in terms of destroying Sun's credibility as a long term supplier.

There are some alternatives:

  1. Sun could sell only its x86 unit to IBM while committing to long term support for the OpenSolaris community.

    That would actually make perfect strategic and business sense for both companies - but is far too good an outcome to be likely.

  2. Someone could play white knight, trump IBM's bid, and leave Sun as the open source and processor design leader in its markets. I believe that Fujitsu at one time agreed to backstop Sun in this role, but that agreement (assuming it was formalized) would long since be history.
  3. anti-trust litigation could stop this thing dead in its tracks - perhaps retroactively justifying Sun's StorageTek purchase by allowing a dead competitor to save the company.As I understand it this was the show stopper for early discussions between Southeastern and IBM, but change has come to Washington and while anti-trust may become an excuse for failing to follow through, it is no longer a real issue.
  4. Sun's own employees could mount a counter-offer and take the company private.That's another winner - and for two reasons:
    1. Sun should have done this years ago - precisely to avoid this kind of situation because even if IBM backs off this time, there will always be another trier; and,
    2. the employees should be highly motivated about now: because only a very few people in sales and support are likely to have jobs after this is over - and they'll be starting again at the bottom of the totem pole.

    So how would that work?

    Simple in principle, painful in practice:

    • someone, like Schwartz or McNealy, would have to front an employee takeover offer to the major shareholders - starting with Southeastern because that forces them to choose publically between taking money from IBM to kill the company or working with the employees to strengthen it.
    • Sun's board would have decide what its role will be: with options ranging from fighting IBM (and thus Southeastern - whose board representation would then be between rocks and hard places) by issuing more shares and using IBM's valuation of Sun to leverage the buyout, to fighting its own employees and founders.
    • eventually (once the employees own the company) normal issuance and valuation rules would supplant the valuations derived from IBM's offer and all of us would continue to benefit from marketplace competition and the open source boom; and,
    • Sun would have to create a legal unit tasked entirely with ensuring that IBM does not directly misuse the competitive information it gained during the failed talks.

So will an employee buyout happen? In a rational world it would be part of the right answer pretty much no matter what IBM does - even if it does the smart thing for both companies: buy only Sun's x64 business and port OpenSolaris to its own PowerRISC lines.

Editorial standards