Intel COO signals willingness to go with AMD64!!

http://biz.yahoo.com/rc/040128/tech_intel_64bit_1.html

On Thu, 29 Jan 2004 02:04:13 GMT, "Yousuf Khan"
wrote:

http://biz.yahoo.com/rc/040128/tech_intel_64bit_1.html


Nothing new here except the words of an analyst:

"Otellini's comments now suggest that Intel intends to release a
desktop chip similar to and compatible with AMD's 64-bit offering,
Brookwood said."

Far as I can tell, Otellini said not a word about compatibility.
Maybe, maybe not. Either the Reuters correspondent wrote a lousy
article, putting the important words into the mouth of someone who
doesn't count, it was edited badly, or Brookwood drew an unwarranted
conclusion.

RM

Far as I can tell, Otellini said not a word about compatibility.
Maybe, maybe not. Either the Reuters correspondent wrote a lousy
article, putting the important words into the mouth of someone who
doesn't count, it was edited badly, or Brookwood drew an unwarranted
conclusion.

IMO, an Intel-branded 64-bit X86 (whether fully-campatible with AMD's
64-bit X86 or not) is more realistic than an average-joe desktop
version of IA64. By the time of the 2008 presidential primaries
(which is a long time in the tech world) IA64 will probably still be
stuck in the high-end server/workstation segment. So much for Intel's
grand vision for IA64 as an successor to X86 Intel just doesn't
want to admit that they'd been beaten by AMD as far as general
widespread acceptable of a 64-bit capable platform. And it would take
a long time for Intel to catch up with then if they wish to go the
IA64 route. So Intel cowers in muted secrecy on what 64-bit X86
support they may be working on.

On Wed, 28 Jan 2004 23:15:00 -0600, lyon_wonder
wrote:


IMO, an Intel-branded 64-bit X86 (whether fully-campatible with AMD's
64-bit X86 or not) is more realistic than an average-joe desktop
version of IA64. By the time of the 2008 presidential primaries
(which is a long time in the tech world) IA64 will probably still be
stuck in the high-end server/workstation segment. So much for Intel's
grand vision for IA64 as an successor to X86 Intel just doesn't
want to admit that they'd been beaten by AMD as far as general
widespread acceptable of a 64-bit capable platform. And it would take
a long time for Intel to catch up with then if they wish to go the
IA64 route. So Intel cowers in muted secrecy on what 64-bit X86
support they may be working on.

Intel said, when it rolled out the P4, that the technology would scale
to 10GHz, so I don't think they had any near-term plans to abandon
x86. By the time the P4 *does* reach 10GHz (if, indeed, it makes it),
the world of computers is likely to have changed beyond recognition,
and I don't think Intel ever had IA64 targeted at desktops. What used
to be desktops will be divided into home/office appliances with
low-power, low-heat, low-noise chips and workstations, where Xeon,
AMD64, Opteron, and Itanium are going to be fighting over a
modest-sized market.

What has happened that _has_ taken Intel by surprise, is that AMD has
successfully invaded a space for corporate servers it thought it
owned. They had expected to have that space nailed down with Itanium
and Xeon with performance out of reach for AMD. My own read is that
IBM's willingness to put the best of its process technology at AMD's
disposal tipped the balance of performance in a way that Intel never
expected.

Intel plainly feels a little pressured, but not so pressured that they
have made anything but P4/Xeon easy to buy. Just try to buy an
Itanium to put into your SuperMicro board or a Pentium-M to put into
your ITX board. Somebody out there beside me must be really
interested in what might be possible with Pentium-M if Intel didn't
keep such a tight grip on it, and Intel is more eager to keep Itanium
from being equated in any way with Opteron than it is to sell them.
Just my read.

RM

"Yousuf Khan" wrote in message
.rogers.com...


http://biz.yahoo.com/rc/040128/tech_intel_64bit_1.html


Nothing in the article or elsewhere even remotely suggests that Intel
will use AMD64 or anything that is similar to it in any way other than it
will run 64-bit software (source code compatability in high-level
languages).

DS

"David Schwartz" wrote in message
...

"Yousuf Khan" wrote in message
.rogers.com...


http://biz.yahoo.com/rc/040128/tech_intel_64bit_1.html


Nothing in the article or elsewhere even remotely suggests that Intel
will use AMD64 or anything that is similar to it in any way other than it
will run 64-bit software (source code compatability in high-level
languages).

I think it's one of those reading of the political tea leaves sort of
exercises. When Intel says that it's going to have processors ready to take
advantage of 64-bit software when that software is ready, the only software
that can be ready at that point is AMD64 software.

Obviously, source code compatibility hasn't resulted in a lot of
cross-platform applications coming out, for example between Itanium or
Opteron. Nor between those two and any other 64-bit platform out there. The
only sort of compatibility worth having is binary compatibility.

Yousuf Khan

On Thu, 29 Jan 2004 01:09:52 -0500, Robert Myers
wrote:

On Wed, 28 Jan 2004 23:15:00 -0600, lyon_wonder
wrote:


IMO, an Intel-branded 64-bit X86 (whether fully-campatible with AMD's
64-bit X86 or not) is more realistic than an average-joe desktop
version of IA64. By the time of the 2008 presidential primaries
(which is a long time in the tech world) IA64 will probably still be
stuck in the high-end server/workstation segment. So much for Intel's
grand vision for IA64 as an successor to X86 Intel just doesn't
want to admit that they'd been beaten by AMD as far as general
widespread acceptable of a 64-bit capable platform. And it would take
a long time for Intel to catch up with then if they wish to go the
IA64 route. So Intel cowers in muted secrecy on what 64-bit X86
support they may be working on.

Intel said, when it rolled out the P4, that the technology would scale
to 10GHz, so I don't think they had any near-term plans to abandon
x86. By the time the P4 *does* reach 10GHz (if, indeed, it makes it),
the world of computers is likely to have changed beyond recognition,
and I don't think Intel ever had IA64 targeted at desktops.

Long before P4 came to fruition, there were "road-maps" which showed x86
reduced to a near-zero role - only STBs IIRC - by 2005/2006. Those
portrayals are no longer available of course so their provenance is
therefore difficult to pin down now but they looked official enough to me
and were published on reputable industry sites like E-Insite. It was
certainly my interpretation at the time that IA64 was targeted to take over
even the desktop market by then and that Intel was the one saying it.

What used
to be desktops will be divided into home/office appliances with
low-power, low-heat, low-noise chips and workstations, where Xeon,
AMD64, Opteron, and Itanium are going to be fighting over a
modest-sized market.

What has happened that _has_ taken Intel by surprise, is that AMD has
successfully invaded a space for corporate servers it thought it
owned. They had expected to have that space nailed down with Itanium
and Xeon with performance out of reach for AMD. My own read is that
IBM's willingness to put the best of its process technology at AMD's
disposal tipped the balance of performance in a way that Intel never
expected.

Yes the AMD servers must be quite a shock to the people at Intel who
thought that AMD would never get more than a nibble at the high ASP sector.
Mind you I haven't seen any firm reports that corporations are biting on
Opteron - AMD *could* do a better job on "visibility".

As for IBM's "willingness" an initial (reported) payment of $46million in
November '02 to fix Cu?/OI for the Opteron (but not for Barton) was surely
a nice incentive.:-)

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

On Wed, 28 Jan 2004 23:09:19 -0800, "David Schwartz"
wrote:


"Yousuf Khan" wrote in message
e.rogers.com...


http://biz.yahoo.com/rc/040128/tech_intel_64bit_1.html


Nothing in the article or elsewhere even remotely suggests that Intel
will use AMD64 or anything that is similar to it in any way other than it
will run 64-bit software (source code compatability in high-level
languages).

The article most certainly does suggest that an analyst read things that
way. Whether Otellini meant that is another matter - Itanium for the
desktop does not "fit" either - an Iteleron??shrug

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

"Yousuf Khan" wrote in message
le.rogers.com...


"David Schwartz" wrote in message
...


"Yousuf Khan" wrote in message
.rogers.com...


http://biz.yahoo.com/rc/040128/tech_intel_64bit_1.html


Nothing in the article or elsewhere even remotely suggests that Intel
will use AMD64 or anything that is similar to it in any way other than it
will run 64-bit software (source code compatability in high-level
languages).

I think it's one of those reading of the political tea leaves sort of
exercises. When Intel says that it's going to have processors ready to
take
advantage of 64-bit software when that software is ready, the only
software
that can be ready at that point is AMD64 software.


Umm, no. He means 64-bit windows software. That is, software that can be
made to run on a 64-bit windows platform of any kind. He doesn't say
anything about binary compatibility and there's no reason to think that's
important.


Obviously, source code compatibility hasn't resulted in a lot of
cross-platform applications coming out, for example between Itanium or
Opteron. Nor between those two and any other 64-bit platform out there.
The
only sort of compatibility worth having is binary compatibility.


That's because there's no 64-bit software market yet. That's Intel's
whole point.

DS

"George Macdonald" wrote in message
...


On Wed, 28 Jan 2004 23:09:19 -0800, "David Schwartz"

wrote:


"Yousuf Khan" wrote in message
le.rogers.com...


http://biz.yahoo.com/rc/040128/tech_intel_64bit_1.html


Nothing in the article or elsewhere even remotely suggests that Intel
will use AMD64 or anything that is similar to it in any way other than it
will run 64-bit software (source code compatability in high-level
languages).

The article most certainly does suggest that an analyst read things that
way. Whether Otellini meant that is another matter - Itanium for the
desktop does not "fit" either - an Iteleron??shrug


There's no evidence that Intel feels that way. Everyone is free to
speculate.

DS