Intel pushes back schedule for Itanium

[utopia]

http://news.cnet.com/news/0-1003-200-2284759.html?tag=st.ne.1002.tgif.ni

 

[Eug]

They should just call it "Unobtainium".

 

[apoppin]

Eventually the Itanic will have its maiden voyage . . .

 

[zippy]

a real shocker...

 

[Burn]

Given Intel's track record with these sort of things I am extremely shocked...lol

 

[Scorpion]

Why even bother now. They might as well just hold of for the McKinley.

Hell the Merced(Itanium) was supposed to come out in Late '98 I believe. This pushback comes at no surprise. It lost all it's spark a long time ago. I don't know anyone who even cares about this chip anymore.

 

[DDad]

I'm surprised Intel didn't try to claim it available as a "special edition" or "limited availability" or some other paper claim

 

[Mday]

if intel can push their schedule back even more, the rambus contract would expire

 

[MagnusPAH]

Ah, I love to hear news like this...Now all I am interested in is to find out the plummeting rate of intel's pc market share. I will be happiest when intel's share is down to 40% or so. This news about Itanium will no doubt help that a little.

 

[Vich]

Why cant all of you people just be quite. They want to iron out the bugs and make it the best they can, so then you people who buy it wont go whinning about this and that, and how this locks up and how this overheats ect. Go rag on 3dfx lol

 

[Czar]

Eventually the Itanic will have its maiden voyage . . . and it will sail the open sea untill it hits an Iceberg that has been in that sea for quite a longer time tima than the Itanic. And somehow the Iceberg has a big AMD logo on it

 

[Crypticburn]

lol... yeah, AMD.... although I like intel chips... AMD is making a mockery of intel

 

[thorin]

Merced/Itanium can go be Hell, to little, to late, to bad. And SledgeHamer a 64bit extension of x86, yes I can see this for compatability reasons, but that's where the advantages end in my eyes. I'm waiting for McKinley it's gonna kick some major farking ass. McKinley will be out after the RDRam BS is over, so say hi to QDR etc for McKinley.

Recap:
Sledgehammer bleh! x86-64 BLEH!

McKinley Yeah! (EPIC) Explicit Parallelism YEAH!!!!!!!!!!!!!!!

And yes I know blah blah blah, new compilers and sh*t. Do you not all remember the 680x0 to PPC move that Apple did. Yes on a far smaller scope then performing the same move in the WIntel world, but no biggy. Everyone needs to lookup this wonderful thing they came up with during that move called Fat Binary.

Here's a def'n for you all:

Fat Binary

When the Apple Macintosh stopped using the original family of Motorola "brains," the 68000 series of CPUs, it moved on to the PowerPC platform, a much faster and more powerful chip. The bad news was that to take advantage of this new platform, Mac developers had to rewrite the code for their programs. The good news is that both versions can peacefully coexist in one application.

English-like programming source code that is compiled into something that a computer understands is called a binary. Because there are two sets of source code compiled into the one application (one for older Macintosh and one for newer PowerPC Macs), the size of the file is almost double, and hence is called a "fat binary."

Products such as Spring Cleaning from Aladdin Systems can neatly remove the code you don't need (your Macintosh is either PowerPC or it isn't...it can't be both) in your fat binary applications. You can sometimes reclaim a good chunk of hard drive space when you trim the "fat!"

--------------------------------------------------

Now why would it be such a big deal to do something like that on a WIntel platform. It wouldn't be, ppl are just to stuck in their ways. Software companies would only need to provide Fat Binary packages for a revision or two (ie: version 5 and mabe version 6). They can also for those who like to complain about HD space and who only support on or the other CPU time in their company. Provide a non accelerated (standard code) version and an accelerated (coded for newer instructionset/CPU) version. Expecially since Intel is providing compilers etc....

Thorin

 

[LocutusX]

Intel has so long been a company that introduced new products by building upon the previous work of the engineering scientists, that they are probably finding it difficult to come up with something 100% new. I'm sure when Mot,IBM,Apple first set out to make the PowerPC it took them a similar length of time (maybe not). Or maybe EPIC is just harder to develop, i.e. they have nothing to base their work on, they're starting from scratch, whatever.

 

[pm]

Actually, Locutus, EPIC is based on research that HP did.

And, Thorin, I agree with you: McKinley will rock.

Patrick Mahoney
IA64 Microprocessor Design
Intel Corp.

* not speaking for Intel *

 

[Buddha Bart]

Fat Binary....

hehehe.. shouldn't be hard to convince MS to bloat its code a little more

bart

 

[LocutusX]

Patrick,

Although HP might have set the ball rolling on EPIC, they didn't actually produce any "production" units did they? And according to your signature, you seem to be involved in Intel's IA64 project - what role do you play?

Personally, I don't think anyone on this board will have a real interest in IA64, apart from the CPU architecture freaks (you know who you are ) simply because Intel is really aiming for the high-end with the Merced/McKinley and I think we'd have to be lucky just to see a Q3A port for it.

 

[pm]

Lucutus: I do cache circuitry design on the McKinley project. Unfortunately, I can't answer any questions at all about McKinley - nor can I comment in any way on Itanium. About the only thing that I can talk about related to IA64 is the EPIC instruction set. As far as EPIC, you are right... HP didn't create any prototype or production units that I'm aware of.

I'm kinda hoping for a port of the next Doom game that iD are doing.

* Speaking only for myself *