BMW, thread, Magee latest on Itanium, some clips and my comments (in bold):
Even if Compaq manages to get Proliants out of the door this side of 2002, that will be eight years since Intel first floated the idea. It only took the White Star Line five years to produce the unsinkable Titanic.
Compaq has been shipping ProLiants for four years. If he means an Itanium populated one, he ought to say it. Careless reporting like that may explain this
We'd ask HP and Compaq, but for reasons best known to themselves, they don't seem to want to talk to the INQUIRER at the moment
Don't show this to Constantine:
We know that Intel had a so-called "skunkworks" running an alternative IA-64 plan up in Oregon, just in case the Itanic sunk. That, we believe, has been dissolved after Alpha became Santa Clara's property.
We think that it's just mere speculation that Intel is running another "skunkworks" for X86-64 architecture just in case it has to.
Unless they're canceled, below are supposed to go Itanium, along with the Tandem non-stop machines (which won't be canceled):
The slide published elsewhere today, is from a recent presentation (October this year), but we see little place here for the great Compaq shift for its customers for Tru64 and for OpenVMS.
Something like what Scott McNeally said yesterday:
Is there absolutely no forward and parallel planning going on here for when the Great Union of Minds comes with HP and Compaq? What must HP's and Compaq's customers think in the interregnum?
In entirety:
Intel's Itanic sails into 2002 fog
Marketing icebergs everywhere
By Mike Magee, 28/11/2001 08:12:43 BST
IN THE SECOND QUARTER of next year, Intel is set to start piloting its McKinley 64-bit processor - the followup to Merced.
It would be very unkind to call Merced an expensive and unmitigated disaster, but certainly it would only be a dyed-in-the-wool enthusiast that could hail it as a runaway success.
The Itanium is still unproved technology but thank god it will still be able to run all of our old DOS software, as Andy Grove pledged all those years back.
Even if Compaq manages to get Proliants out of the door this side of 2002, that will be eight years since Intel first floated the idea. It only took the White Star Line five years to produce the unsinkable Titanic.
This is far from unusual in the chip business, even though the press and marketing wings of the semi firms would have us all believe it's just a simple matter of hey presto, hooplah!
For example, we heard Mr Jerry Sanders +++ of AMD back in 1992 telling us about roadmaps that meant the K9 would be delivered by the year 2000.
This, of course, was before everything changed as it realistically must, despite the armchair visionaries dreams.
AMD's Hammer is delayed but maybe that doesn't matter too much - getting all of us in a feeding frenzy will eventually pay dividends, presumably.
We know that Intel had a so-called "skunkworks" running an alternative IA-64 plan up in Oregon, just in case the Itanic sunk. That, we believe, has been dissolved after Alpha became Santa Clara's property.
We think that it's just mere speculation that Intel is running another "skunkworks" for X86-64 architecture just in case it has to.
The only place you'll find quadruped skunks in the UK is in zoos, but they do smell very aromatic from miles away, we understand.
McKinley, which will at least run at 1GHz, according to Intel roadmaps, is being hailed as the real mccoy of IA-64 architecture and although it was originally supposed to have been fabbed on a .13 micron process, it's still very unclear whether Intel will manage that or not.
Madison, which is supposed to have a 6MB L3 cache, has started appearing on Intel roadmaps for 2003, and that is where everything - as if it wasn't already cloudy - gets cloudier still - mostly because we doubt very much if those Intel engineers have entirely sorted out what they want to do with this Alpha stuff.
The Powerpoint slides are there, but the silicon might be more difficult to fabricate.
For HP, the points of the arrow now seem to converge on the Itanium, but the PA-RISC chips, which Intel has been known to fab for them in the past, now seem to stretch right up to Madison and Deerfield, while OSes seem to be NT, HP/UX and Linux and Novell. This is almost as puzzling to us as it may be to you. The slide published elsewhere today, is from a recent presentation (October this year), but we see little place here for the great Compaq shift for its customers for Tru64 and for OpenVMS.
Is there absolutely no forward and parallel planning going on here for when the Great Union of Minds comes with HP and Compaq? What must HP's and Compaq's customers think in the interregnum?
The PA-8800 is on a timeline close to Madison and Deerfield, and we know from the Intel roadmaps we've seen that Madison is slated for the first half of 2003. So what conjuring trick is going to happen if the merger goes through?
We'd ask HP and Compaq, but for reasons best known to themselves, they don't seem to want to talk to the INQUIRER at the moment and the PR situation is as fuzzy and illogical here in the UK as the state of the Great Union of Minds, with no-one really knowing how to brief, who to talk to, or what to say.... about practically anything. To describe it as chaos would be to overegg the cake but its certainly pretty unstructured. The PR bunnies are even briefing against each other, believe it or not. No names here - we don't want to upset their clients too much... µ
See Also
HP: all roads lead to IA-64
Compaq's secret VMS plans
Birth of the Titanic
theinquirer.net |
