Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

http://www.theinquirer.net/?article=19187

Yousuf Khan

--
Humans: contact me at ykhan at rogers dot com
Spambots: just reply to this email address ;-)

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

Yousuf Khan wrote:

> http://www.theinquirer.net/?article=19187

Is Intel's x86 dual-core based on Northwood? Prescott? P6? PM? The
article mentions 3.2 GHz, which, if accurate, rules out P6 and PM.
(Unless 3.2 GHz = two cores at 1.6 GHz... just kidding.)

AMD's dual core is supposed to run slower than their single core. It
looks like Intel hopes they don't have to underclock their dual core?

--
Regards, Grumble

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

Grumble wrote:
> Yousuf Khan wrote:
>
>> http://www.theinquirer.net/?article=19187
>
>
> Is Intel's x86 dual-core based on Northwood? Prescott? P6? PM? The
> article mentions 3.2 GHz, which, if accurate, rules out P6 and PM.
> (Unless 3.2 GHz = two cores at 1.6 GHz... just kidding.)
>
> AMD's dual core is supposed to run slower than their single core. It
> looks like Intel hopes they don't have to underclock their dual core?
>

the dual core clocked at 3.2GHz that Inq lists would be the one they've
been talking about for months--two prescotts stitched together at the
pins. Intel has announced Prescott based, Prescott's successor-based,
Pentium-M-based, and Itanium 2-based dual cores for 2005. That's just
the one most people will care about--the first desktop model.

Alex
--
My words are my own. They represent no other; they belong to no other.
Don't read anything into them or you may be required to compensate me
for violation of copyright. (I do not speak for my employer.)

Archived from groups: comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

Yousuf Khan wrote:

> http://www.theinquirer.net/?article=19187
>
> Yousuf Khan
>

Newsflash: The Inquirer floats down out of geek hyperspace and
acknowledges some everyday reality:

"Intel's aggressive marketing of the kitemark and the Centrino brand has
paid dividends for it. To many folk, wi-fi notebooks and Centrino means
the same."

Staff probably needs a hit of something or other to regain altitude.

RM

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

On Thu, 21 Oct 2004 10:07:55 +0200, Grumble <devnull@kma.eu.org>
wrote:
>
>Yousuf Khan wrote:
>
>> http://www.theinquirer.net/?article=19187
>
>Is Intel's x86 dual-core based on Northwood? Prescott? P6? PM? The
>article mentions 3.2 GHz, which, if accurate, rules out P6 and PM.
>(Unless 3.2 GHz = two cores at 1.6 GHz... just kidding.)
>
>AMD's dual core is supposed to run slower than their single core. It
>looks like Intel hopes they don't have to underclock their dual core?

Err, 3.2GHz is their planned top speed for this dual-core chip when it
arrives in early 2006. Given that they are already at 3.6GHz now and
plan on getting to 3.8GHz before the end of the year, I would say that
they are indeed downclocking their dual-core chips relative to the
single core ones!

-------------
Tony Hill
hilla <underscore> 20 <at> yahoo <dot> ca

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

Tony Hill wrote:
> On Thu, 21 Oct 2004 10:07:55 +0200, Grumble <devnull@kma.eu.org>
> wrote:
>
>>Yousuf Khan wrote:
>>
>>
>>>http://www.theinquirer.net/?article=19187
>>
>>Is Intel's x86 dual-core based on Northwood? Prescott? P6? PM? The
>>article mentions 3.2 GHz, which, if accurate, rules out P6 and PM.
>>(Unless 3.2 GHz = two cores at 1.6 GHz... just kidding.)
>>
>>AMD's dual core is supposed to run slower than their single core. It
>>looks like Intel hopes they don't have to underclock their dual core?
>
>
> Err, 3.2GHz is their planned top speed for this dual-core chip when it
> arrives in early 2006. Given that they are already at 3.6GHz now and
> plan on getting to 3.8GHz before the end of the year, I would say that
> they are indeed downclocking their dual-core chips relative to the
> single core ones!
>

Isn't that supposed to be the whole point of multi-core
for both AMD and Intel ? In other words, to find ways
to continue to improve cpu performance without having to
rely solely on jacking up clock speeds ?


--
Reply to rob.stow.nospam@shaw.ca
Do not remove anything.

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

In article <rGRdd.6870$%%1.5676@pd7tw3no>,
Rob Stow <rob.stow.nospam@shaw.ca> wrote:

>Isn't that supposed to be the whole point of multi-core
>for both AMD and Intel ? In other words, to find ways
>to continue to improve cpu performance without having to
>rely solely on jacking up clock speeds ?

Both AMD and Intel have already been doing lots of things to improve
cpu performance other than only jacking up clock speeds. Multiple cpus
on a die is one of many things they're trying. The reason people are
wondering about how low the clock will be is that they hate
sacrificing too much single-thread performance to get better total
performance. Also, they want to know how much improvement in total
performance that they're going to get.

-- greg

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

Greg Lindahl wrote:
> In article <rGRdd.6870$%%1.5676@pd7tw3no>,
> Rob Stow <rob.stow.nospam@shaw.ca> wrote:
>
>
>>Isn't that supposed to be the whole point of multi-core
>>for both AMD and Intel ? In other words, to find ways
>>to continue to improve cpu performance without having to
>>rely solely on jacking up clock speeds ?
>
>
> Both AMD and Intel have already been doing lots of things to improve
> cpu performance other than only jacking up clock speeds.

Such as ? Take the AMD64 processors, for example.
Multi-core would be the first significant change to
the AMD64 architecture since the Opty 140 and 240 were
released at 1.4 GHz. All we have seen in the meantime
is a steady jacking up of clock speeds and there is
nothing else on the horizon for the next 6 to 9 months.

The situation has been much the same for the P4 since
it first came out. Many small changes have been made
to allow Intel to keep jacking up clock speeds, but
the basic chip design has stayed the same.

You could make an argument for the Pentium M as being
Intel's effort to get performance at lower clocks and
without needing a nuclear reactor in every home, but
since you *still* can't buy a full-fledged ATX
motherboard for Pentium M the point is pretty much moot.

> Multiple cpus
> on a die is one of many things they're trying. The reason people are
> wondering about how low the clock will be is that they hate
> sacrificing too much single-thread performance to get better total
> performance. Also, they want to know how much improvement in total
> performance that they're going to get.
>


--
Reply to rob.stow.nospam@shaw.ca
Do not remove anything.

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

>> Both AMD and Intel have already been doing lots of things to improve
>> cpu performance other than only jacking up clock speeds.
> Such as ?

pipelining, OOO execution, bigger caches, faster caches, higher
associativity in caches, fewer cycles for some operations, on-board memory
controller, wider busses, better compilers, SMT, more registers, new
instructions, more reservation stations, more ALUs, better branch
predictors, ...

> Take the AMD64 processors, for example. Multi-core would be the first
> significant change to the AMD64 architecture since the Opty 140 and 240
> were released at 1.4 GHz. All we have seen in the meantime is a steady
> jacking up of clock speeds and there is nothing else on the horizon for
> the next 6 to 9 months.

Only Ghz can be changed without some significant redesign, so it's no wonder
that within a short time span nothing else than Ghz will change.


Stefan

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

Rob Stow <rob.stow.nospam@shaw.ca> wrote :

> You could make an argument for the Pentium M as being
> Intel's effort to get performance at lower clocks and
> without needing a nuclear reactor in every home, but
> since you *still* can't buy a full-fledged ATX
> motherboard for Pentium M

yes You can, in Japan )

Pozdrawiam.
--
RusH //
http://randki.o2.pl/profil.php?id_r=352019
Like ninjas, true hackers are shrouded in secrecy and mystery.
You may never know -- UNTIL IT'S TOO LATE.

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

RusH wrote:
> Rob Stow <rob.stow.nospam@shaw.ca> wrote :
>
>
>>You could make an argument for the Pentium M as being
>>Intel's effort to get performance at lower clocks and
>>without needing a nuclear reactor in every home, but
>>since you *still* can't buy a full-fledged ATX
>>motherboard for Pentium M
>
>
> yes You can, in Japan )
>
> Pozdrawiam.

Can you get me anything more detailed ?
Brand/Model ? URL ?

Son of a friend of mine is in Japan teaching English
and could perhaps ship something to me.

--
Reply to rob.stow.nospam@shaw.ca
Do not remove anything.

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

Rob Stow wrote:

> RusH wrote:
>
>> Rob Stow <rob.stow.nospam@shaw.ca> wrote :
>>
>>
>>> You could make an argument for the Pentium M as being
>>> Intel's effort to get performance at lower clocks and
>>> without needing a nuclear reactor in every home, but
>>> since you *still* can't buy a full-fledged ATX
>>> motherboard for Pentium M
>>
>>
>>
>> yes You can, in Japan )
>>
>> Pozdrawiam.
>
>
> Can you get me anything more detailed ?
> Brand/Model ? URL ?
>
> Son of a friend of mine is in Japan teaching English
> and could perhaps ship something to me.

Hear, hear. I would love to build an ATX Pentium M system.

--
Mike Smith

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

lindahl@pbm.com (Greg Lindahl) writes:

>In article <rGRdd.6870$%%1.5676@pd7tw3no>,
>Rob Stow <rob.stow.nospam@shaw.ca> wrote:

>>Isn't that supposed to be the whole point of multi-core
>>for both AMD and Intel ? In other words, to find ways
>>to continue to improve cpu performance without having to
>>rely solely on jacking up clock speeds ?

>Both AMD and Intel have already been doing lots of things to improve
>cpu performance other than only jacking up clock speeds. Multiple cpus
>on a die is one of many things they're trying. The reason people are
>wondering about how low the clock will be is that they hate
>sacrificing too much single-thread performance to get better total
>performance. Also, they want to know how much improvement in total
>performance that they're going to get.


Its possible they wouldn't have to. If AMD's highest clocked dual core
in 90nm by A4 2004 is 2.5 GHz, and their highest clocked single core in
90nm reaches 3.5 GHz, it may be possible to overclock your 2.5 GHz dual
core to 3.5 GHz. That's assuming the circuits are the same on the dual
core part -- they might use more lower performance lower power
transistors on the dual core parts so this is not possible. If you
could do it might require more power than some motherboards and power
supplies could handle, and produce enough heat to require a really top
notch heatsink and fan. But if the transistors were the same on the dual
core part I could see this being a favorite activity among overclockers.

Another route would be to support cool'n'quiet on a per core basis, to
allow the OS or BIOS to dynamically manage the power draw and heat on
each core. Each would be capable of 3.5 GHz, but when one went up to
that speed, the other might drop to low power mode of only 1 GHz. It'd
take OS support for this regardless of how its managed since the OS
would need to know if one CPU is 3-4x faster than the other.

Better yet, for the long term, might be to have one high performance
core and a bunch of little ones for less important tasks. A K8 type
core along with four cores that were more on the order of VIA's new C7
in terms of size, power usage and performance.

--
Douglas Siebert dsiebert@excisethis.khamsin.net

"They that can give up essential liberty to obtain a little temporary
safety deserve neither liberty nor safety" -- Thomas Jefferson

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

Mike Smith <mike_UNDERSCORE_smith@acm.DOT.org> wrote :

> Rob Stow wrote:

>> Son of a friend of mine is in Japan teaching English
>> and could perhaps ship something to me.
>
> Hear, hear. I would love to build an ATX Pentium M system.

http://www.google.pl/search?q=%22pentium+m+motherboard
first hit - doh

Pozdrawiam.
--
RusH //
http://randki.o2.pl/profil.php?id_r=352019
Like ninjas, true hackers are shrouded in secrecy and mystery.
You may never know -- UNTIL IT'S TOO LATE.

Archived from groups: comp.arch,comp.sys.ibm.pc.hardware.chips,comp.sys.intel (More info?)

Rob Stow wrote:

> You could make an argument for the Pentium M as being
> Intel's effort to get performance at lower clocks and
> without needing a nuclear reactor in every home, but
> since you *still* can't buy a full-fledged ATX
> motherboard for Pentium M the point is pretty much moot.

Given that the M/B vendors have done other things Intel didn't like,
such as dual Celeron boards and BIOS support for overclock for examples,
I have to think that this is because the M/B vendors don't see the
demand, rather than because there's some conspiracy.

I'm open to another explanation, I just have faith in greed providing me
with toys ;-)

--
-bill davidsen (davidsen@tmr.com)
"The secret to procrastination is to put things off until the
last possible moment - but no longer" -me