It’s Intel’s most powerful processor ever. It has the ability to take on IBM, sink Sun, make or break HP, and crush or revive AMD. It’s keeping every CEO in computing up at night. And it’s just getting started. The multibillion-dollar battle between Itanium 2 and its rivals has begun.
See This Chip?
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSAlert.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
This article is aimed at the largely non-technical readership of Fortune, and for that reason alone will probably be rather uninteresting to most of OSAlert’s readers. However, just skimming through it I noticed a number of glaring technical inaccuracies which further invalidate it as a worthwhile read.
The article does go a great detail into the various corporate relationships and rivalries which underlie modern big iron servers. If you don’t know about these, perhaps you’ll be interested in this article, but if you’re looking for a technical perspective on the battle for big iron, I’d suggest you look elsewhere.
That’s a pretty good article, there. And I’d have to agree with the guy from IBM who says “backwards compatibility always wins”.
That is one of the reasons that Microsoft was so successful with Windows 95. They maintained the ability to run MS-DOS programs inside of a new system.
I am a little biased-I’ve never had an Intel machine, only AMD, and never had any problems with them. And so cheap!
I read exact same stuff about sinking Sun, etc. in exact same sort of publications a few years ago. And still there are no indications of Itanium gaining momentum and I doubt that it ever will. The article sounds more like same old tired FUD sponsored by Intel. Sun is still gaining market share in high end 64-bit server space and Itanic is still going nowhere fast.
From what I have read and seen, the CPU isn’t the problem, Intel is. For example, lets look at IBM and the Power4 processor. They are able to produce chips CHEAPER than Intel, yet, their microprocessor side of their business is not even close to the size of Intel.
Yesterday I had a chat to an Intel sales representative, and from what he gathers the first two releases were the “testing the water”, and Itanium 3 (aka Deerfield and sub sequent releases) will be the “cheaper” version which may even replace the Xeon. If that is the case I would be more than willing to spend $AUS1500 on an Itanium processor and a few hundred on a motherboard, HOWEVER, unless they come down in price people are always going to look at Power4, PA-RISC, Alpha, heck, even the old Ultra Sparc as a cheaper and superior solution to Itanium.
I’ve done pricing and it is still cheaper for me to purchase a Ultra Sparc motherboard and processor. Until Intel starts to see the writing on the wall and lower their prices they will be relegated to the “obscure and weird” corner of the IT world.
I have been wonder, AMD is moving into 64 bit for desktops. But The Itanium 2 and beyond look like they will not be desktop cpu’s. Is intel planning another 32bit cpu after the P4? Or will they come out with a cheap version of the itanium for desktops?
yes .. the pentium 5 should be released in Q3
Break HP? HP is planning on moveing from their pa-risc to itanium.
Uh-oh, nice stuff in the article. Since when has the Toshiba R5900 32-bit MIPS CPU in PS2 been a variant of IBM’s 64-bit POWER?
After I read the documentation from Intel, the EPIC architecture looks to me a bit crazy. I can’t see any real argument how a compiler can take avantage from that extrem complicated design in a better way as it can be in other (parallel) architectures.
Arto, you mentioned PS2, but what about N64? The MIPS R4300 chip inside would be a power derivative, uh?
And doesn’t this fantasic processor only run at 1Ghz?
Try explaining how this is so wonderful to joe sixpack when he has a 3Ghz for his business or try telling him the difference between the processors.
I didn’t mention Nintendo, since I didn’t know what it runs on, and I had a distant memory of *some* console that used IBM processor.
That’d be Nintendo’s Gmecube with a 405MHz PowerPC in it…
Big Iron currently use 64bit processors. Most common system allow up to 32 processors each running at about 1GHz. They also don’t run windows… These machines have been in production use for over 4 years and they have very solid OSs that are only rebooted for upgrading and even then it has to be a major upgrade.
We have 2 systems were I work:
IBM AS/400
– patches: most can be installed with a live system.
– Hard Disks: can be relaced or added with a live system. they can also be attached to the current partition with a live system.
– memory: needs a reboot
– OS upgrade: needs a reboot
Stratus:
– patches: can be installed with a live system.
– hard disks: can be installed with a live system.
– memory: can be installed with a live system.
– OS upgrade: can be installed with a live system.
Lets look at Windows:
– patches: please reboot
– hard disk: please reboot
– memory: please reboot
– OS upgrade: please reboot 8 times until it installs correctly.
If Intel makes it any where with the new chip, it’ll take Linux and *BSD to make them productive servers.
HP can switch; since they are talking about taking their good Unix product over which will let all there software vendors to just do a recompile.
MSFT is taking Windows over to it, but will will have very slow or no backwards compatabilty with the old software. Don’t expect to see anything but back office software ported any time soon. Until there’s a market, almost no tools vendors will be switching.
If Intel makes it any where with the new chip, it’ll take Linux and *BSD to make them productive servers.
No, it doesn’t. You do yourself mention HPUX, and VMS first saw the light of day on Intel last friday. Both are big iron systems.
Since you can insert and remove floppies, ZIP disks and
CDs in Windows, why would hard drives be any different? Is it really in practice? Has anyone tried?