Intel says no to 64-bit until MS Longhorn arrives?

[Black Jack]

http://www.xbitlabs.com/news/cpu/display/20040325151223.html

I wonder if Microsoft would be willing to delay 64-bit Windows that
long to accomodate Intel? Longhorn isn't expected now till something
like 2007, but Microsoft can get 64-bit out with the current Windows
XP.

Yousuf Khan

 

[Gary L.]

http://www.xbitlabs.com/news/cpu/display/20040325151223.html

I wonder if Microsoft would be willing to delay 64-bit Windows that
long to accomodate Intel? Longhorn isn't expected now till something
like 2007, but Microsoft can get 64-bit out with the current Windows
XP.

I saw no mention in the article that 64-bit Intel processors would not
be available until Longhorn shipped. What it said was that the 64-bit
feature would be disabled for general retail desktop CPUs until there
was sufficient support available in terms of software and device
drivers. Presumably when Windows XP-64 ships (perhaps later this
year), the support will be available and PC manufacturers will want
64-bit enabled P4 CPUs.
- -
Gary L.
Reply to the newsgroup only

 

[David Schwartz]

I saw no mention in the article that 64-bit Intel processors would not
be available until Longhorn shipped. What it said was that the 64-bit
feature would be disabled for general retail desktop CPUs until there
was sufficient support available in terms of software and device
drivers. Presumably when Windows XP-64 ships (perhaps later this
year), the support will be available and PC manufacturers will want
64-bit enabled P4 CPUs.

What is it with Intel and crippled chips?

DS

 

[Tony Hill]

What is it with Intel and crippled chips?

Am I the only one that doesn't really understand Intel's reasoning
here?! They don't want to release 64-bit chips to anyone for desktop
use because they're afraid that some customers will have driver
problems? Isn't that an issue for the customers (OEMs) to decide for
themselves? It's not like having a 64-bit P4 *forces* customers to
use a 64-bit operating system!

I can't help but think that this is a DUMB move. It's Intel yet again
deciding what their customers want, despite anything that the
customers may say. Fortunate turn of events for AMD though I suppose.

 

[Black Jack]

Am I the only one that doesn't really understand Intel's reasoning
here?! They don't want to release 64-bit chips to anyone for desktop
use because they're afraid that some customers will have driver
problems? Isn't that an issue for the customers (OEMs) to decide for
themselves? It's not like having a 64-bit P4 *forces* customers to
use a 64-bit operating system!

It's probable that not all features of Intel's AMD64 are working right
yet. So maybe Intel doesn't want to enable it until it's got
compatibility tests running against AMD reference hardware.

Or it's possible that it hasn't gotten enough performance out of
64-bit mode out its own processors to make it comparable to AMD's.
That likely will never come until Intel integrates a memory controller
right into the CPU, as it's rumoured to do.

Yousuf Khan

 

[Ed]

What is it with Intel and crippled chips?

DS

What is it with the Intel buyers?
Ed

 

[Ed]

What is it with Intel and crippled chips?

DS

What is it with people who keep buying Intel's crippled chips? Are they
just brainwashed with no hope of return or what? ;p
Ed

 

[Tony Hill]

It's probable that not all features of Intel's AMD64 are working right
yet. So maybe Intel doesn't want to enable it until it's got
compatibility tests running against AMD reference hardware.

If that were the case then why are they releasing it for *servers*,
where reliability is more of a concern! The Nocona is a 64-bit
enabled CPU that has the same core as the Prescott. It will likely be
the next stepping of that core, so it might fix some minor
compatibility issues, but it will be here later this year. If Intel
plans on having the core working and out in the market this year, why
would they not sell it on the desktop for another 2+ years.

Or it's possible that it hasn't gotten enough performance out of
64-bit mode out its own processors to make it comparable to AMD's.
That likely will never come until Intel integrates a memory controller
right into the CPU, as it's rumoured to do.

Now this is a slightly more plausible explanation if you ask me,
though it seems odd that they would lose much performance, relative to
AMD, when going to 64-bit mode. As it is the Athlon64 and the P4 are
fairly comparable in terms of performance, and 64-bit doesn't seem to
change this much. The larger pointers will put a little bit more
strain on the memory subsystem, but it should be more of a bandwidth
strain than a latency one, hence the integrated memory controller
shouldn't change much.

 

[George Macdonald]

Am I the only one that doesn't really understand Intel's reasoning
here?! They don't want to release 64-bit chips to anyone for desktop
use because they're afraid that some customers will have driver
problems? Isn't that an issue for the customers (OEMs) to decide for
themselves? It's not like having a 64-bit P4 *forces* customers to
use a 64-bit operating system!

No - you're not the only one. I'd also be interested to know what
exactly is involved in enabling/disabling the 64-bit mode on Intel's CPUs.
As I've said before this could be done with a crippled BIOS without
actually crippling the hardware but the feeling I get is that they are
going to take the latter route, which I'd think is an irreversible process
and therefore those chips will never be able to have the BIOS updated to
allow enabling of the 64-bit.

I can't help but think that this is a DUMB move. It's Intel yet again
deciding what their customers want, despite anything that the
customers may say. Fortunate turn of events for AMD though I suppose.

Judging by the noises coming from game developers yup, it's a really dumb
move - could result in a lot of black 'n' blue asses in Intel's marketing
dept. one of those days. Hard to say how much of the game developer
"noise" has been due to AMD's PR but it's still being heard. Given current
pressures on M$ already, I don't see how they can possibly collude with
Intel on holding up WinXP-64 until it's umm, "completey baked" from Intel's
perspective. Surely they're not hoping that the 32-64 transition is going
to take as long as the 16-32 did??

Rgds, George Macdonald

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

 

[chris]

On Sat, 27 Mar 2004 19:50:51 -0500, George Macdonald wrote:

Judging by the noises coming from game developers yup, it's a really
dumb move - could result in a lot of black 'n' blue asses in Intel's
marketing dept. one of those days. Hard to say how much of the game
developer "noise" has been due to AMD's PR but it's still being heard.
Given current pressures on M$ already, I don't see how they can possibly
collude with Intel on holding up WinXP-64 until it's umm, "completey
baked" from Intel's perspective.

I agree, it *must* be Intel being the bottleneck here. After all, nothing
stopped M$ ever from releasing non-finished products... ;-)

Cheers,
Chris

 

[Judd]

On the desktop side... not the server side where the next Xeon will be
64-bit. If you want workstation/server hardware, then you'll get 64-bit.
The masses will have to wait until there is actually support for it. I'm
not sure what the huff is about.

 

[Carlo Razzeto]

On the desktop side... not the server side where the next Xeon will be
64-bit. If you want workstation/server hardware, then you'll get 64-bit.
The masses will have to wait until there is actually support for it. I'm
not sure what the huff is about.

The huff is about Intel stupidly missing out on this market because 64bit
desktop processeors are already out there, and Microsoft is already working
on supporting it. While microsoft does develop at a snails pace this doesn't
mean intel should just sit on this product while they wait for AMD to get
intrenched here. Foolish

Carlo

 

[Eric]

The huff is about Intel stupidly missing out on this market because 64bit
desktop processeors are already out there, and Microsoft is already
working on supporting it. While microsoft does develop at a snails pace
this doesn't mean intel should just sit on this product while they wait
for AMD to get intrenched here. Foolish

Carlo

Not foolish, smart! Do you think Intel is just sitting there? What do you
think the initial costs are in marketing, developing compilers, 64 bit
apps, OS's, drivers etc etc? Whos funding that? Not Intel, AMD! And once
that is all done, Intel releases their "32 bit extended to 64 bit" CPU to
the world (the one (or more)they are testing and refining now), and whos
ahead then? Remember, its about money, thats all, nothing more, just money
and in the end Intel will come out on top in this saga.
Eric

 

[Eric]

If that were the case then why are they releasing it for *servers*,
where reliability is more of a concern! The Nocona is a 64-bit
enabled CPU that has the same core as the Prescott. It will likely be
the next stepping of that core, so it might fix some minor
compatibility issues, but it will be here later this year. If Intel
plans on having the core working and out in the market this year, why
would they not sell it on the desktop for another 2+ years.


Now this is a slightly more plausible explanation if you ask me,
though it seems odd that they would lose much performance, relative to
AMD, when going to 64-bit mode. As it is the Athlon64 and the P4 are
fairly comparable in terms of performance, and 64-bit doesn't seem to
change this much. The larger pointers will put a little bit more
strain on the memory subsystem, but it should be more of a bandwidth
strain than a latency one, hence the integrated memory controller
shouldn't change much.

Dont forget the cache, the double size pointers, and probably more prolific
use of 64 bit ints among other things will effectively reduce the cache
size to some degree and thus performance. So generally speaking, a 64 bit
cpu will require a larger cache, no?
Eric

 

[Eric]

What is it with Intel and crippled chips?

DS

Crippled, how so? Did you not get what was advertised and what you paid for?
All Intel is doing is adding some transistors to a chip, testing them
internally and disabling them before selling the chip on the market, they
never claimed that part to be even available. Taking the costs of spinning
silicon into account, I'd say they have a rather innovative way to save
money, wouldn't you?
Eric

 

[AD.]

Not foolish, smart! Do you think Intel is just sitting there? What do
you think the initial costs are in marketing, developing compilers, 64
bit apps, OS's, drivers etc etc? Whos funding that? Not Intel, AMD! And
once that is all done, Intel releases their "32 bit extended to 64 bit"
CPU to the world (the one (or more)they are testing and refining now),
and whos ahead then? Remember, its about money, thats all, nothing
more, just money and in the end Intel will come out on top in this saga.
Eric

I'm not so sure of that. I get the feeling Intel would rather spend a lot
of money (they aren't short of it) keeping their 'reputation' as the
industry leader and genuine article intact.

Intel has made a lot of money off that reputation, and I'm not sure they'd
want to give it up that softly. Perception is everything. I think they
really have been caught with their pants down because they were too
focused on Itanium.

That reputation probably isn't so important when fighting over the low end
low margin stuff with AMD, but AMD is creeping up towards their higher
margin enterprise customers.

Once AMD is 'legitimised' to enterprise customers, Intel is going to have
to actually compete for that segment in the future.

Cheers
Anton

 

[Black Jack]

Now this is a slightly more plausible explanation if you ask me,
though it seems odd that they would lose much performance, relative to
AMD, when going to 64-bit mode. As it is the Athlon64 and the P4 are
fairly comparable in terms of performance, and 64-bit doesn't seem to
change this much. The larger pointers will put a little bit more
strain on the memory subsystem, but it should be more of a bandwidth
strain than a latency one, hence the integrated memory controller
shouldn't change much.

Well, maybe the P4 won't be very latency-dependent, but maybe if they
do replace the P4 with a Pentium-M core, which is much more
latency-dependent, then an onboard memory controller will be needed.

Yousuf Khan

 

[Black Jack]

On the desktop side... not the server side where the next Xeon will be
64-bit. If you want workstation/server hardware, then you'll get 64-bit.
The masses will have to wait until there is actually support for it. I'm
not sure what the huff is about.

The support for it is already there on the desktop side. Prerelease
Windows, as well as production Linux and BSD of course. And really
even the pre-release Windows is basically finished all they are
waiting for is a few more drivers from ATI and the like; they already
have Nvidia. ATI, as usual, is keeping its driver development a
closely guarded secret, which probably means they are going to be very
buggy as usual at release.

Yousuf Khan

 

[chrisv]

I saw no mention in the article that 64-bit Intel processors would not
be available until Longhorn shipped. What it said was that the 64-bit
feature would be disabled for general retail desktop CPUs until there
was sufficient support available in terms of software and device
drivers. Presumably when Windows XP-64 ships (perhaps later this
year), the support will be available and PC manufacturers will want
64-bit enabled P4 CPUs.

Well what's the point of disabling the feature at the chip level?

 

[David Schwartz]

David Schwartz wrote:

Crippled, how so?

In the sense that you take something that has a capability and you
deliberately disable that capability.

Did you not get what was advertised and what you paid for?

No, I think you're missing the point. I don't buy crippled chips. I want
the best technology that exists, not something that's deliberately had its
capabilities reduced for marketing reasons.

All Intel is doing is adding some transistors to a chip, testing them
internally and disabling them before selling the chip on the market, they
never claimed that part to be even available.

Exactly, in other words, a crippled chip.

Taking the costs of spinning
silicon into account, I'd say they have a rather innovative way to save
money, wouldn't you?

How does crippling a chip save them money? Are you suggesting they can
sell a crippled chip for more than the corresponding uncrippled chip?

DS