Oracle Says No To Itanium – Embarrassment For Intel, Big Problem For HP

Oracle announced today that it is going to cease development for Itanium across its product line, stating that itbelieved, after consultation with Intel management, that x86 was Intel’s strategic platform. Intel of course responded with a press release that specifically stated that there were at least two additional Itanium products in active development – Poulsen (which has seen its initial specifications, if not availability, announced), and Kittson, of which little is known.

This is a huge move, and one that seems like a kick carefully aimed at the you know what’s of HP’s Itanium-based server business, which competes directly with Oracle’s SPARC-based Unix servers. If Oracle stays the course in the face of what will certainly be immense pressure from HP, mild censure from Intel, and consternation on the part of many large customers, the consequences are pretty obvious:

  • Intel loses prestige, credibility for Itanium, and a potential drop-off of business from its only large Itanium customer. Nonetheless, the majority of Intel’s server business is x86, and it will, in the end, suffer only a token loss of revenue. Intel’s response to this move by Oracle will be muted – public defense of Itanium, but no fireworks.
  • HP gets hit. Oracle’s database is a major workload on HP’s Itanium servers, with possibly up to 50% of HP’s flagship Superdome servers running Oracle. A public statement from Oracle that it will no longer develop their database software will be a major drag on sales for anyone considering a new Oracle project on HP-UX and will make customers and prospects nervous about other key ISV packages as well. We should note that this does not equate to either wholesale abandonment of HP platforms or an immediate dip in revenue. Oracle versions tend to live for years after their successors are announced, so many HP customers with current Oracle versions (which Oracle will continue to support on HP-UX/Itanium) will still buy additional capacity for years after the next non-Itanium versions are introduced.
  • IBM can potentially profit with a combination of their P-Series servers running AIX as an alternative to HP-UX. However, since they also compete with Oracle’s servers and their volumes are not that much different from HP’s, they could be the next victim. IBM will probably use this opportunity to ramp up their DB2 marketing on AIX as well.
  • Major enterprise HP-UX customers will see their options constrained and will probably incur additional costs out of current budget plans. Since migrating from the Oracle database is difficult and expensive, many of them may be forced to prematurely refresh their systems if they need a newer version of Oracle under on a Unix platform. And guess who will be prepared to sell them such a system, on a CPU architecture that has even less market share than either Itanium or Power CPUs?
  • Microsoft can come out ahead on this as well by aggressively pushing SQL server into the uncertainty caused by Oracle’s move.

All in all a very rough move on Oracle’s part. Oracle claims that the move was not motivated by competitive issues with HP, and HP has been very vocal in decrying it as destructive and unfair to major enterprise Oracle customers. While it is always difficult to unravel the decision process of major shifts like this one, my gut feeling is that HP’s claim probably has some merit. Oracle was in a position where an  arguably reasonable decision on long-term strategy also had a potentially negative tactical impact on a major competitor, and as such was probably much easier to make.

As a closing thought, note that I say “tactical” and not “strategic” in regards to the impact of this decision on HP. HP has the technology and the resources to port HP-UX to an x86 platform, and today even has x86 systems such as the DL980 that are worthy platforms for HP-UX, as well as the ability to build a Superdome variant based on x86. HP has resisted this move for many years as x86 matured and out of a desire to preserve Itanium system revenues. Oracle’s thrown gauntlet may be the trigger HP needs. From a customer perspective, there is no real need to consider abandoning a strategic HP relationship. The company has a good track record in bridging generations of hardware for its HP-UX users, and there is no reason to think that the transition to an x86 system would be different – current Itanium customers should expect multiple yeas of parallel availability, and HP has said they will support Itanium systems for at least a decade.


HP customers do have another choice

Richard, this is great analysis. I think mainstream IT media missed the point that this really nothing but an Oracle hunt for HP business. I blogged about it earlier today
HP customers don't have to take a cue from Oracle and abandon their HP hardware investment. IBM's DB2 v9.7 runs on HP hardware and has a nifty Oracle compatibility feature that makes database migration from Oracle a fairly simple proposition. Most applications coded to Oracle APIs can be migrated in weeks if not days. Oracle DBA and programing skills are also preserved and no change over of hardware is required.

Regarding the option of

Regarding the option of keeping database workloads on HP/Itanium by switching to IBM DB2... there is actual some recent Forrester research in this area at Simpler Database Migrations Have Arrived!

HP has always maintained

HP has always maintained they'd have some technical difficulty porting HP-UX to X86. HP-UX is big endian, and X86 is little endian. Sun managed to port Solaris (also big endian on SPARC) to X86, but it was an expensive and time consuming project. Moreover, even if HP-UX somehow landed on X86, what could it run? It still couldn't/wouldn't run any Oracle products. At least HP-UX on Itanium runs Oracle versions up through March, 2011. Would IBM, SAP, Tibco, and other software vendors show up to an HP-UX on X86 party? Very, very doubtful.

Another aspect to this problem is that Oracle also controls critically important middleware on VMS (Rdb) and NonStop (Tuxedo), the two other operating systems that HP offers on Itanium. Oracle's announcement was really aimed at all three of HP's enterprise operating systems, and it delivered a head wound to all three.

I think HP ought to ring IBM (if they haven't already) and strike a deal focused on the Power and System z CPUs. IBM owns Transitive, the company that did such a fantastic job creating Rosetta for Apple, which runs PowerPC (big endian) Macintosh binary code within Intel (little endian) versions of Mac OS X. IBM and HP could take the same approach with HP-UX, folding HP-UX into AIX as a subsystem (that HP sells and supports, probably) together with some Transitive code to provide Itanium binary support on AIX Power for those applications that need it. Then all the AIX middleware becomes available to HP-UX customers within the same operating system. VMS would have to run as a separate operating system instance, again with some Transitive magic. (Power is bi-endian and can run in either mode.) NonStop probably ought to move to System z CPUs, given the demanding qualities of service required for that environment and its customers. NonStop and System z are both big endian, and the NonStop Kernel would run in a separate partition or virtual machine on System z. Again, Transitive could work their magic to provide heritage Tandem/MIPS/Itanium binary support on the System z CPU. IBM would probably agree to have HP resell System z hardware to NonStop customers. (Hitachi got that deal in the early 2000s.)

That's the hardware side. On the software front, IBM is now Itanium's biggest middleware provider that still has continuing development. HP needs to strike a deal there, too, to help move its customers to IBM middleware while HP and IBM execute on the hardware strategy to protect HP's enterprise customers.

Or maybe HP just cashes out and sells the entire HP-UX, NonStop, and VMS businesses to IBM, then goes and takes that money to try to build a software business essentially from scratch. Maybe HP's ex-SAP CEO would want to buy SAP if he could afford it.

Good Article

Hi Richard,

Good Article... particularly interested with the HPUX on x86 option... I had actually lost a bet with one of my friends who happened to be in HP Pre-Sales where I had predicted that HP will come out with HPUX on x86 by Jan 1, 2011. Seems I could have managed with some more time..... attaching a link where Intel as come out with a statement on RAS capabilities of its new Nehalem EX processors being similar to Itanium.

The other side to the HPUX on x86 option is that if it works (it would really be tough) then it would seem that oracle has shot itself in its foot and taken IBM along with them as this would really change the dynamics of the RISC market.

I would add to your last line where HP need not buy SAP but get into an agreement where HP sells Sybase as a DB solution offering in all its Solution deals and gets to have SAP's next ECC version with Sybase ported in its Itanium and HPUX boxes. SAP anyways sells a lot of HP software solutions in its stack. Leo can definately add his clout there.

Prasad Patil