Many companies backing Intel's Itanium processor are planning to announce a new alliance in September to try to make it easier for customers to adopt systems using the high-end chip, ZDNet UK sister site CNET News.com has learned.
The group, called the Itanium Solutions Alliance, has several plans to make Itanium more useful, said a source involved with the outfit. The alliance will sponsor porting events to help programmers bring their software to the processor, set up porting centres where such work can take place and create catalogues so customers in specific industries can find combinations of Itanium hardware and software for various tasks.
In addition to Intel and HP — the co-developer of Itanium and the top seller of Itanium servers — the alliance includes server makers NEC, SGI, Unisys, Hitachi, Fujitsu and Bull, and software makers Microsoft, Red Hat, Novell, Oracle, SAP and SAS, the source said.
The alliance is a new step in a years-long effort to encourage broad Itanium use and is an indication of just how much effort its backers are willing to put into getting the processor to catch on. But some believe the effort is late.
"It sure would have been nice to have this several years ago," Sageza Group analyst Clay Ryder said. "What happened here is Intel seriously underestimated the support the marketplace would have for an incompatible platform and believed 'if we say it's industry standard, everybody will come.' Clearly that didn't happen."
Intel and HP declined to comment for this story. However, Intel spokeswoman Erica Fields said the chipmaker has been working hard to attract hardware and software partners and now there are more than 5,000 software applications and five operating systems available for Itanium.
There are plenty of companies selling and buying Itanium-related products, but not nearly the number Intel and HP projected when the chip was under development in the 1990s.
Itanium's rocky past
The chip's debut was marred by delays, low initial performance and incompatibility with software for x86 processors such as Pentium and Xeon, which are vastly more widely used. In 2004, Intel missed Itanium shipment goals and HP scrapped its Itanium workstation. In 2005, IBM backed off plans to sell its own Itanium servers.
In response to the troubles, Intel redefined Itanium as a chip chiefly intended for high-end multiprocessor servers, models that today usually use RISC (reduced instruction set computing) processors such as IBM's Power, Sun' UltraSparc, Fujitsu's Sparc64 and HP's PA-RISC.
In line with that high-end positioning, Microsoft's planned update to its Windows Server 2003 operating system — called R2 — won't be available for Itanium. Microsoft's rationale for the move is that R2 is geared toward smaller servers. The Window Server 2003 successor due in 2007, code-named Longhorn Server, will support Itanium, however.
Itanium hasn't yet reached heir-apparent status by most estimates. HP — the server maker that has pushed the chip most aggressively — sold $108m in Itanium-based Unix servers in the second quarter of 2005, compared with $1.1bn in the PA-RISC-based Unix servers (£59.9m and £610m respectively) they're intended to replace, according to Gartner figures.
But Intel has convinced Unisys, NEC, SGI, Fujitsu and Hitachi to join HP in designing mammoth Itanium systems and bringing them to the market.
"The industry has lined up — all but one — with Itanium as the platform of choice for RISC replacement and mainframe platforms for the future," Pat Gelsinger, general manager of Intel's digital enterprise group, said last week at the Intel Developer Forum in San Francisco.
The lone holdout Gelsinger referred to is IBM, whose Power processor is the second horse in what Intel likes to call a two-horse race, but it should be noted that Sun also shuns Itanium in favour of high-end systems built around Sparc processors.
Intel's Itanium plans still extend years into the future. Next up will be a model code-named Montecito, due to be released by the end of this year. It employs dual cores, an approach already used by high-end server chips from Sun, HP and IBM.
"Montecito will deliver a huge step up in performance, more than doubling the performance capabilities of today's Madison family of products," Gelsinger said last week.
After Montecito comes a close relative called Montvale, in 2006. Then in 2007, Intel plans to introduce a four-core Itanium called Tukwila, a design that will be succeeded by another four-core chip, code-named Poulson.
For years, one of Itanium's advantages over Xeon was a 64-bit architecture that permitted easy use of vast amounts of memory. AMD's Opteron brought that feature to x86 processors in 2003, though, and Intel followed suit with 64-bit Xeons in 2004.
The Xeons have the advantage because they run the vast amount of software already available for x86 processors, but efforts to expand the pool of available software should help Itanium's prospects, Ryder said.
"Most people do not buy systems because of floating-point performance, [and] the war for basic processing capability is largely over. There's more than enough for most people," Ryder said. "They buy systems based on applications."






Talkback
The Itanium processor is Intel effort to yet again enforce their will for their own gain I would like to add on the computing world. Three cheers for AMD and there efforts to give the customer what they really want; instead of forcing us into a computing bait and switch. This is where they tell us how great their product is; show us all the facts and figures they can massage, and once you get it your spend the next year plus in support hell trying to get the damn thing to do what they claimed it would do when you purchased. Oooops! That's Microsofts' territory! Intel better watch out. Bill may develop his own chip with internal security and virus scanning:):) I work in the Enterprise computing world, and it will be quite some time before Intel can rise to the level of performance demanded in today's telecom data centers. The thing that really gets me is all the rhetoric and lies that's in the IT news about this under-performing processor that you have to find sofware to run on it, spend time helping them develop so they can in-turn sell it back to at 10 times the cost. Get lost Intel!!!
This is all just smoke and mirrors. After 5 years of failed efforts for Itanium to survive, the ISA needed to bet another $10BN. IDC's forecasts on Itaniums success was over 95% wrong and they now come out with a survey that's even more flawed, clearly trying to get some of that $10BN in marketing funds! Its time to do a reality check folks! Itanium is flatlining, volumes are decreasing (just ask IDC), roadmaps have slipped, the #2 and #3 vendors have dropped it and the new #2 is ready for bankruptcy because of the slips! Heres a great and respectable analysis on Itaniums futures. Everyone should read it! http://valleyviewventures.com/content/00/01/03/59/23/userimages/Itanium%20--%20Is%20This%20a%20Wise%20Investment%202-17-06.PDF
The King is dead, long live the King!