Updates to this story
“At length did cross an Albatross:
Through the fog it came;
As if it had been a Christian soul,
We hailed it in God’s name.” – Rime of the Ancient Mariner, Coleridge
Oracle could never be really described as shy and retiring. So we weren’t surprised to see the software giant and Sun system server telling the world that it was phasing out Itanium support – Microsoft and Red Hat made the decision to do that in the last two years. (See Intel, HP slam Oracle over Itanium support and Oracle puts bullet into Itanium corpse)
But what was interesting is that in this statement Oracle said that it was clear that the Itanium is coming to the end of its life as a processor, after discussions with senior management at Intel.
There’s no more senior manager at Intel than its CEO, Paul Otellini – and he was quoted in HP’s response to Oracle’s announcement – saying that the Itanium very much had a future.
HP, of course, was a fellow traveller with Intel on the road to develop a CPU based on EPIC architecture and rather controversially ported the very robust OpenVMS to the Itanium architecture. Intel has spent vast amounts of money on creating and developing the Itanium but over the last 10 years products have been delayed, support for it has been somewhat lukewarm from the rest of the industry “egosystem” and it’s not entirely clear what its fate is to be.
There was a time when Intel had very high hopes for this microprocessor, believing that widespread support from the PC industry would propel it into being one of the most important CPUs it ever designed. However, as Intel continued to grow market share and adopt 64 bit support for its X86 microprocessors, continuing questions about the future of the Itanium proliferated – and were met by senior Intel executives with shuffling of the seat, embarrassing coughs and “we’ll take that question offline” – a euphemism for “we’re not going to answer that question”.
HP’s reaction to the Oracle move was to attack it because it was trying to force its customers into buying Sun SPARC servers, desperate to maintain market share for its own big tin. But finding out what customers want is a tricky business. HP’s apparently over the top reaction to the Oracle announcement conceals an uncomfortable fact for both HP and Intel.
HP has a number of US government customers and contracts that it must fulfil, stretching years ahead into the future. It can’t just drop those customers without any consequences and Intel needs for several reasons to stay committed to the Itanium too. No one knows how many Itaniums Intel makes and sells in a year, but the fact is that its Xeon microprocessors vastly outnumber the EPIC-based chip in the corporate server market.
It’s rather hard to believe that Intel has ever made a profit or broken even on the Itanium family of chips – the amount of money it’s thrown into development, into partnership schemes and marketing is impossible for anyone on the outside to calculate.
No doubt the Itanium is something of an embarrassment for Intel – but it’s an embarrassment that it has to bear. It just doesn’t have to talk to anyone outside the charmed circle about this damnable chip that it just can’t get rid of – a bit like the albatross in the Rime of the Ancient Mariner, by Coleridge.
Intel killed pretty much all x86 alternatives (incl. Alpha) and wanted to be 'the' 'future instruction chip maker' rather than competing in a multi vendor risc market.
Now it's own exit is prevented because it is so central to the that market, rather than just part of the multi vendor market.
Microsoft with IE6 did everything it could (subsidised flagship activeX projects) to tie the browser to it's particular desktop.
Now it spends 70% of research budget on cloud research, but it is hamstrung (by IE6 tight integration to the desktop, and IE6 woeful standards compliance) from growing that cloud business at the rate it desires.
( Hence Microsoft's own recent campaign to encourage folks to migrate from IE6 )
Judging by these links, Microsoft would love to be able to exit from anything to do with activeX also:
http://bit.ly/fvFvAU
http://www.google.com/search?q=activeX+error#q=activeX+error&hl=en&tbo=1&output=search&source=lnt&tbs=qdr:y&sa=X
This Itanium graphic is the best:
http://www.wikipedia.org/wiki/File:Itanium_Sales_Forecasts_edit.png
Here is an useful extract about the chip development history:
"Because this chip [ Itanium ] was supposed to radically change the way computers work and become the driving force behind all systems in the future, one promising project after another was dropped. The MIPS chip, the DEC Alpha (perhaps the fastest chip of its era), and anything else in the pipeline were all cancelled or deemphasized. Why? Because Itanium was the future for all computing. Why bother wasting money on good ideas that didn't include it?"
http://www.pcmag.com/article2/0,2817,2339629,00.asp