Death of HP-UX and Itanium

I used to sell Itanium based systems running HP-UX back during my time at HP in the early 2000s so I have fond memories of how I was able to compete with confidence, knowing that I had a robust solution that clients liked and respected in my armory. Heck, I even did a huge $5m deal at the time for Superdome refresh. How times have changed.  I decided to take a walk down memory lane and google HP-UX and obviously, I landed on a HPE website pretty rapidly.  The white paper I found speaks volumes.  Not by the copy, as that is what I remember fondly, but by the dates on the citations. Time seems to be frozen in 2014.  Check it out for yourself here.  Perhaps my senses should have been alerted when the HPE website didn’t ask me to fill in my details before downloading the online brochure.  Sad how the mighty have fallen.

Let’s take a trip down memory lane.

History of Itanium

The Itanium line had a complex history, the initial bold direction was set when the aspiring server chip giant, Intel, partnered with the old Hewlett Packard to merge the concepts of very long instruction word (VLIW) and explicit parallelism (EPIC) in its future PA-RISC chips into Intel’s future silicon.

Think back to the 1990s, a time when the datacenter was undergoing explosive growth and almost daily transformation. The 1990’s was a time when Sun Microsystems ruled the Unix space (if you believed their marketing) and the transaction processing space. This was time when AMD was just starting to think about getting into the server market. how times have changed. The server space was a lot less homogenous then than it is now, and it was a lot smaller market, too. The risks that chip makers took were proportionately bigger but at the same time, the options seemed to be wider.

HP started development on the ideas of EPIC back in 1989 as it was clear that its PA-RISC line was going to run out of runway.  Then in 1994, HP and Intel formed the Itanium chip partnership, and a year later the RISC/Unix crowd were all starting to commit to porting their operating systems to Itanium they hype at the time made the transition seem somewhat inevitable.

This infamous Itanium server revenue forecast on Wikipedia encapsulates a hype/reality curve better than any I have seen recently.

By the time Intel got Itanium working reasonably well in the 2008-2009 timeframe roadmap churn had set in.  The issues were compounded by delayed processor deliveries several times and then things got really interesting.

Oracle Drops the bomb.

Oracle unilaterally decided to drop support for Itanium systems running HP’s HP-UX operating system in 2011 this may have had something to do with Oracle’s acquisition of Sun Microsystems a little over a year earlier, but that would be pure speculation on my part obviously…. HP sued, obviously, claiming that Oracle was in breach of a 2010 contract between the two companies in which the database firm promised to support HP’s Itanium systems. Then in June of 2016 a San Jose jury awarded what was now Hewlett-Packard Enterprise (HPE) $3 billion (£2.25B) in damages.  The payout due to Oracle’s breached contract to provide Itanium support in its namesake database and Linux distribution. If only HPE was as good at acquisition strategy as its lawyers were in court at the time. Autonomy anyone???

Oracle had a point, even if the court didn’t agree. AIX and Solaris never made it to Itanium, and with Red Hat Linux decommitting in 2009 and Microsoft pulling Windows and SQL server support in 2010 they were not exactly alone in their intent.

What Happened to HP-UX?

HP was still defending HP-UX as late as 2014 which is admirable, given the writing must have been on the wall since 2009-2010. With the rise of Linux operating system and opensource in recent years, the time of HP-UX and the bespoke Itanium processor seems a long time ago.

What started out in 1984 as HP’s proprietary Unix operating system culminated in V11.3 released in 2007.

The Final Bell

Intel’s Itanium family of processors for enterprise servers has spent the better part of a decade as the walking dead. The platform is now derisively referred to as “Itanic.” The final steps of Itanium’s slow march toward eventual demise will end on July 29, 2021-just over two years away.

Intel’s order deadline for the parts was earlier this year, on January 30, 2020, though this deadline was only relevant for the sole Itanium customer, HPE. Support for HPE’s Itanium-powered Integrity servers, and HP-UX 11i v3, will come to an end on December 31, 2025.

What are you to do if you are still running Itanium and HP-UX?

You still need the major qualities that led you to HP in the first place:

  • Performance
  • Availability
  • Scalability
  • Security

However, you want to embrace the benefits of Linux.  Well of course Red Hat have some helpful content on their website which can be found here.

But what to do about the hardware layer now you have the OS sorted?  Commodity x86 is one approach, but you had that choice back in the 2000s and you actively chose a UNIX system instead.  You wanted more.  You still do.  LinuxONE is the answer.  LinuxONE high-end Linux system is built for the demands of mission-critical workloads.  LinuxONE runs an IBM developed chip architecture that has a rich history before it found its way into the LinuxONE server line up launched in mid-2015.  With a 5.2Ghz processor and a huge cache LinuxONE is a beast and typically runs open-source software such as MongoDB and PostgreSQL 2x faster than commodity x86 servers.


While the HP-UX and Itanium processors are behind the need for a high-end system designed for the most demanding workloads remains.  LinuxONE is the answer.  To find out more about LinuxONE read more here.

One thought on “Death of HP-UX and Itanium

  1. AMD played an extremely important role in the Itanium processor’s history. Intel positioned Itanium as its true 64-bit processor, supporting direct addressing of more than 4 GB of memory. Intel maintained that its X86 processors would remain with 32-bit addressing, albeit with PAE (Physical Address Extension) support, introduced in 1995. Itanium processors could technically run X86 instructions (with 32-bit addressing), but the performance was never terrific, and I’m being kind. This market positioning arrangement made a great deal of business sense for Intel, HP, and the more junior members of the Itanium family. It was an attempt to push the market, including software developers, in a certain direction while making the transition less painful but not painless. Unfortunately for Intel and HP (mostly HP), a competitor had a different idea.

    In 2000, AMD announced the AMD64 architecture which built on PAE but was/is a true expansion of the X86 architecture to 64-bit addressing. (We now also sometimes refer to this architecture as “X86-64.”) AMD released its first AMD64 processors in 2003. And they were rather good, certainly far better running 32-bit X86 code than Itanium (“IA64″) processors were, plus they allowed straightforward, evolutionary expansion into 64-bit programming with decent performance to start and more to come. This architectural development forced Intel to respond at least to hedge its bets, and Intel shipped its first EM64T/”Intel 64” processor in mid-2004, fully software compatible with the AMD64 processors. Within the next couple years “Intel 64” (actually AMD64) architecture trickled down into almost all of Intel’s X86 product line. In other words, AMD defined the new-but-evolutionary-and-familiar 64-bit architecture for X86 processors, and Intel followed. How reluctantly or enthusiastically Intel followed, we don’t know, but they did. And the rest is history.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.