HP to the rescue!
IDC and Gartner don't normally break Itanium server shipments and sales out as a separate category, so you have to infer how well or poorly Itanium is doing by using Hewlett-Packard as a kind of yardstick. And if you have watched HP over the years, the transition from the collection of PA-RISC, Alpha, and NonStop servers to a converged Itanium-based Integrity platform has seen a slight decline in sales.
What I mean is that as sales of those older systems have contracted, Integrity sales have risen, but not enough to offset the declines. In the fourth quarter, by IDC's reckoning, HP's sales of RISC and EPIC (that's the shorthand name for the Itanium instruction set) servers fell by 11 per cent to $1.42bn, while IBM's sales of Power Systems (including machines that run AIX, Linux, or i) fell by 10 per cent to $2.48bn. Sun's RISC server sales (it doesn't do Itanium) fell by 18 per cent to $1.08bn. The remaining RISC+EPIC server sales amounted to $325m in sales, down 26 per cent, with Bull doing both Power and Itanium boxes and Fujitsu doing both Sparc and Itanium boxes and NEC peddling Itanium.
If money matters more than shipments, that doesn't exactly seem like a position of strength for Itanium if HP's numbers are any guide (assuming that the vast majority of RISC+EPIC sales at HP were for Itanium boxes). Moreover, Sun abandoned Itanium before it even shipped (perhaps one of the smart things Sun did during the dot-com bust) and IBM and Dell dumped Itanium years ago.
Unisys has all-but abandoned Itanium processors, and in the wake of its acquisition of the IT business of Siemens, Fujitsu has been pretty clear that its x64-based Primergy line is going to be primary in Europe, with Sparc and mainframes getting honorable mentions and no one saying much about the PrimeQuest Itanium boxes. NEC is still committed to Itanium, and Bull seems to be, too, but that still leaves HP as its main champion.
The alliance wanted to point out that there are some 14,000 applications available on Itanium, provided you double, triple, or even quadruple count those software programs as they are compatible with HP-UX, the Linuxes from Red Hat and Novell, Windows, OpenVMS, or NonStop. (I would guess that there are only several thousand unique applications on the Itanium platform once you stop the multiple counting.)
Itanium enthusiasts are particularly jazzed about Novell's SUSE Linux Enterprise Server 11 supporting Itanium and the upcoming R2 release of Windows Server 2008 Itanium Edition that will scale across 256 threads, along with the future "Kilimanjaro" SQL Server release, which will also span 256 threads. The current Windows Server 2008 and SQL Server are both limited to 64 threads.
The real question is not whether Itanium has a future as an HP-UX and Windows database server, but if the future eight-core "Nehalem EX" Xeon 7500s will basically eat all of the business when they ship later this year or early next.
A four-socket Nehalem EX machine pack 64 threads and will cost a lot less than an equivalently powerful Itanium machine. And that probably means Itanium will be pushed up into an even tighter niche of customers who want more than 64 threads to run databases. Unless you are HP-UX customers, in which case you won't have a choice but to buy Integrity boxes. HP has said again and again it has no intention of porting HP-UX to x64 chips. ®
For *very* large databases Nonstop on Itanium is worth considering. Nonstop supports running a single Nonstop SQL database over 4000 cpus (8000 cores) with each cpu having 48GB non shared memory (Nonstop is a parallel shared nothing architecture). Not that anyone would ever need to go that big, but running several hundred cpus is realistic for some applications. For Microsoft SQL Server I can't see much benefit in going the Itanium route over x86.
Re: Let's try to understand the processor market a bit
> leibniz Computing Centre in Munich.
> NASA also have Altrix systems running Linux with 2K+ Itanium 2 chips in them
Real "scale up" market's acid test is Finance (bank of america), retail(walmart), telco (verizon) and manufacturing (Unilever) segments. Not research or academia.
I doubt if NASA or LRZ even have CIOs... probably it's a professor/scientist making the decision (not a CIO like in walmart's case). But don't get me wrong - I am not criticising Linux's capabilities. All I am saying is that "real datacenters" take time before adapting a new technology. For NASA the user is more educated and hands on. They also have a bigger risk appetite than the real world (both due to self confidence and lack of business pressure). For the real world CIOs - NASA, universities, PARC, CERN - it all sounds like "lab experiment" to their ears. (not yours, I understand that part).
And yes, overtime these instances will prove Linux's ability as a single instance champion. That's when HPs and DELLs of the world will start pushing big iron linux as a readily available solution. It will take a good amount of time before we see Linux kicking out Solaris, HP-UX and AIX from the big iron niche.
Oops I think I got my coat aleady.
Re: Let's try to understand the processor market a bit
--" While Linux on itanium has scaled to a single image 512 processors in a SGI lab; it'll be a long, long while(even a decade) before customers start using linux on x86 beyond 4/8 processors. CIOs like to stay one step behind. And they matter, not a lab result "--
Perhaps you should take a trip to visit the Leibniz Computing Centre in Munich. LRZ houses Germany's National Supercomputer System, and runs an Altix 4700 with 4,096 Itanium 2 processors, 17TB of global shared memory on a single Linux instance. That got installed in 2006. NASA also have Altrix systems running Linux with 2K+ Itanium 2 chips in them. Big Linux left the lab and proved itself a long time ago.