% fortune -ae paul murphy

A dumb prediction: IBM will buy HP's Unix customers

Just over a year ago Scott McNealy offered to have Sun extend support for HP-UX and begin a process to migrate HP's remaining Unix customers to Solaris. Most of the press laughed, and HP's people, presumably too busy bugging journalists to watch what was going on in the Unix part of their business, essentially did nothing.

Now Sun's T2 is out and it's pretty much the world beater they promised - 30% faster on SPEC then IBM's 4.7Ghz Dual core Power6 and, more significantly, one third the cost and somewhere between two and three times the throughput of the Itanium.

Sun, however, seems unlikely to repeat their offer to HP on the same terms because what they were willing to pay for in 2005/6 is coming their way now at no attributable cost - and that leaves HP between rocks and hard places of its own making.

The right engineering solution is probably to drop Itanium, start buying silicon from Sun, and switch the customer base to Solaris on SPARC with an HP brand name on the box -while offering customers both Linux and Solaris on HP branded x86. Unfortunately making good engineering decisions isn't one of HP's current strengths and I don't see them doing this.

Instead, I think the people who want HP to be the best Compaq it can be will prevail and therefore that they'll put their Unix customer base up for sale to the highest bidder - and that won't be Sun because anyone still buying HP-UX and Itanium after Rock comes out will be doing it because they hate Sun and are quietly hoping for a miracle - just as DEC's partisans (and HP's own MPE customer base) did before them.

It could be IBM - because those people largely trust IBM, are sitting ducks for a Linux conversion, are used to spending big for support, and will needs lots of it.

One of the things IBM's people will have to consider in deciding on this is whether making this deal will eventually force them to choose between continuing with AIX or dropping it in favor of building toward Linux compatibility on the HP-UX code base.

I've used both, and they're both terrible - but in very different ways. HP-UX has, ahem, cough, "idiosyncracies", but it works, and quite reliably too. AIX, in contrast, has so much data processing crap glued onto a fairly lightweight System V core that it can be very difficult and unforgiving to work with - and is correspondingly fragile in large scale production use where shared control creates the potential for technical conflict.

So, bottom line, how dumb is this prediction? It's a bet on how smart IBM is - and, really, how can that be a loser?


Paul Murphy wrote and published The Unix Guide to Defenestration. Murphy is a 25-year veteran of the I.T. consulting industry, specializing in Unix and Unix-related management issues.