By DaveLevy on Dec 14, 2006
I have been researching Sun's competitive position against HP. It's clear to me that HP, when they think about IT as opposed to ink, agree with Nick Carr and his thesis "IT doesn't matter". Mark Hurd (HP CEO) in the highlights of his keynote speech [.mp3] to Oracle World earlier this year (November) stated that their IT would contain
no proprietary infrastructure...... consisting of proliant and "c" class blades, linux, network storage, dynamic smart cooling, integrated racks
No room for HP/UX! Nor for Itanium! No wonder the HP/UX roadmap is so ephemeral. My colleague Matias Alonso, has also written about this direction from HP and finishes his article with the comment that even HP prefer the cost of open source to their own proprietary fees, which presumambly they rebate, so the TCO must be pretty poor.
It's already been damaged as a platform by the requirement to recompile or develop in order to transition from HP PA RISC to Itanium, and neither HP/UX nor Itanium have the volumes to retain and recruit developers. This discontinuity is a true tippining point; if you have to spend money on the application, you'd prefer (as Hurd himself says earlier in the speech) to spend it on improving the business logic, not extending its life.
HP-UX 11i v3 has slipped again to next year  and is now over 2½ years late. Given it only runs on HP's Itanium, perhaps the demand's not there! A prediction of this sorry state of affairs was made Sun when it published over 15 months ago an article "Reality Check: HP-UX has no future!", on its web site.
The chart below has been built from an article at IT jungle, analysing HP's most recent earnings call. Now that HP/UX is only available on Itanium, it is part of 'Business Critical Systems' (which is the extruded segment.)
Business Critical Systems (BCS) earned $987m during the quarter selling Itanium, PA-RISC, MIPS and Alpha based systems The Itanium revenues grew (representing 45% of the revenue) but HP do not publish the Linux/HP-UX breakdown any more. These earnings represent a year on year decline because the growth in Itanium systems is insufficient to compensate HP for the decline in the older proprietary PA-RISC and Alpha systems. The one thing we do know is that HP-UX only represents a part of the BCS revenues. Is this enough to keep HP interested? Is it enough to keep the ISVs interested? ISV's that want an Itanium port, will probably choose Linux. I think HP-UX's future is destined to add to HP's growing reputation as the IT industry's operating systems undertaker and join Ultrix, True64 and VMS in the hospice.
Hurd's Oracleworld speech streaming audio [.mp3]: the quote is at 3:10, you can find the whole speech with other keynote speakers on the Oracle World 2006's keynote speaker page. We can see HP's historic delivery of their roadmap, HP's recent release history.