Sign in with
Sign up | Sign in
Your question

P4C temperature on a P4P800 ..Is it hot ?

Last response: in Motherboards
Share
Anonymous
a b V Motherboard
June 5, 2004 7:58:59 AM

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

Wondering if 40C / 104F is standard for a P4C 2.8 running default settings.
That's what reports BIOS Hardware monitor right now.

Very fine thread on this matter here:
http://www.intelforums.net/viewtopic.php?t=2778

He reports:
"we tested the Northwood C under the same conditions,
and it was marginally cooler at 104*f (40*C) peak while overclocked,
and 87*f (31*C) at idle. "

My P4 was installed today, I expect the heat-sink to take some time
to settle properly over the CPU as the heat conductive layer should compress a bit,
thus offering a better heat propagation to the HS.

Any thoughts on that matter ?
Regards,
Anonymous
a b V Motherboard
June 5, 2004 9:09:26 AM

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

Just as a quick follow-up..
I just found-out about the AI overclocking that is not affecting voltages,
so I felt safe to make a few quick tests..

Stock 2.8Ghz 40C
+10% ... 3.07Ghz.. No problemo. (Temperature rose to 43C)
+20%.....3.36Ghz...FSB 991Mhz... piece of cake. (Now at 47C)

I just had to stop there.
Felt like Q playing with gravitation constant of the universe.
It's insane.... Plus PAT enabled all the time. Pinch me someone.
(No stability tests yet, but it runs a few programs fine so far)

Makes me laugh thinking back at the good'ol Celeron days. 300Mhz overclocked to 400.
I'm starting to realize why so many are talking about Zalman or such beasty HS.

ASUS.. See what you've done...... I thought I was cured !
:D 
Anonymous
a b V Motherboard
June 5, 2004 4:13:09 PM

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

what type of ram are you running? The AI overclocking reduces ram/cpu
ratio to something like 2\3 in order to maintain stability. You are
getting a higher cpu speed with a lower overall speed.

On my p4p800dlx P42.8c using pc3700 ram running at a ratio of 1:1 I am
safely using a 237 fsb (rather that the standard 200).

At worse if you are using pc3200 ram, use a manual setting of 5:4
Cpu:ram.

At a 1:1 ratio my cpu speed is 3.33ghz

temp is 27 C idle & 50 C while doing the prime 95 torture test

Locust

On Sat, 5 Jun 2004 05:09:26 -0400, "James Bald"
<Rutherford@NOSPAM.ever> wrote:

>Just as a quick follow-up..
>I just found-out about the AI overclocking that is not affecting voltages,
>so I felt safe to make a few quick tests..
>
>Stock 2.8Ghz 40C
>+10% ... 3.07Ghz.. No problemo. (Temperature rose to 43C)
>+20%.....3.36Ghz...FSB 991Mhz... piece of cake. (Now at 47C)
>
>I just had to stop there.
>Felt like Q playing with gravitation constant of the universe.
>It's insane.... Plus PAT enabled all the time. Pinch me someone.
>(No stability tests yet, but it runs a few programs fine so far)
>
>Makes me laugh thinking back at the good'ol Celeron days. 300Mhz overclocked to 400.
>I'm starting to realize why so many are talking about Zalman or such beasty HS.
>
>ASUS.. See what you've done...... I thought I was cured !
>:D 
>
Related resources
Anonymous
a b V Motherboard
June 5, 2004 4:13:10 PM

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

I knew there was a catch.
But I'm barely starting to study the details of this brand new home built system yesterday.
Just started a thread on "Matching DDR" before putting my hands on a set of:
VALUE Kingston KVR400X64C32A/512 MOSEL Assy in China(2)
Onboard chips are MOSEL Vitelic corp V58C2256804SAT5
Here for Paul and alike:
http://developer.intel.com/technology/memory/ddr/specs/...
http://www.validationlabs.com/webreport/PublicModule.as...
Two sticks in Blue mobo's slots. Bought same time; Serial numbers are so close they must come from the same wafer. :) 
One chtick is
I never expected miracles from these rams, and if you say AI overclock reduces the ram/cpu ratio, I'm still qui te happy.
Performance and stability tests Later. SO far so good..

IMHO.. It's value allright cuz the system won't post in Turbo mode (so far), keep in
mind I haven't finished tuning.. This is just an early experiment to determine actual
system temperatures before getting my hands dirty...My mind already is enough :) 

Live long and prosper Lucust,

>end user wrote in message ...
>what type of ram are you running? The AI overclocking reduces ram/cpu

>ratio to something like 2\3 in order to maintain stability. You are
>getting a higher cpu speed with a lower overall speed.
>
>On my p4p800dlx P42.8c using pc3700 ram running at a ratio of 1:1 I am
>safely using a 237 fsb (rather that the standard 200).
>
>At worse if you are using pc3200 ram, use a manual setting of 5:4
>Cpu:ram.
>
>At a 1:1 ratio my cpu speed is 3.33ghz
>
>temp is 27 C idle & 50 C while doing the prime 95 torture test
>
>Locust
>
>On Sat, 5 Jun 2004 05:09:26 -0400, "James Bald"
><Rutherford@NOSPAM.ever> wrote:
>
>>Just as a quick follow-up..
>>I just found-out about the AI overclocking that is not affecting voltages,
>>so I felt safe to make a few quick tests..
>>
>>Stock 2.8Ghz 40C
>>+10% ... 3.07Ghz.. No problemo. (Temperature rose to 43C)
>>+20%.....3.36Ghz...FSB 991Mhz... piece of cake. (Now at 47C)
>>
>>I just had to stop there.
Anonymous
a b V Motherboard
June 5, 2004 7:44:04 PM

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

I have the exact same CPU and it runs at anywhere from 48-50 C on average
use.
"James Bald" <Rutherford@NOSPAM.ever> wrote in message
news:F86dnbZPGdVpe1zdRVn-hw@look.ca...
I knew there was a catch.
But I'm barely starting to study the details of this brand new home built
system yesterday.
Just started a thread on "Matching DDR" before putting my hands on a set of:
VALUE Kingston KVR400X64C32A/512 MOSEL Assy in China(2)
Onboard chips are MOSEL Vitelic corp V58C2256804SAT5
Here for Paul and alike:
http://developer.intel.com/technology/memory/ddr/specs/...
http://www.validationlabs.com/webreport/PublicModule.as...
Two sticks in Blue mobo's slots. Bought same time; Serial numbers are so
close they must come from the same wafer. :) 
One chtick is
I never expected miracles from these rams, and if you say AI overclock
reduces the ram/cpu ratio, I'm still qui te happy.
Performance and stability tests Later. SO far so good..

IMHO.. It's value allright cuz the system won't post in Turbo mode (so far),
keep in
mind I haven't finished tuning.. This is just an early experiment to
determine actual
system temperatures before getting my hands dirty...My mind already is
enough :) 

Live long and prosper Lucust,

>end user wrote in message ...
>what type of ram are you running? The AI overclocking reduces ram/cpu

>ratio to something like 2\3 in order to maintain stability. You are
>getting a higher cpu speed with a lower overall speed.
>
>On my p4p800dlx P42.8c using pc3700 ram running at a ratio of 1:1 I am
>safely using a 237 fsb (rather that the standard 200).
>
>At worse if you are using pc3200 ram, use a manual setting of 5:4
>Cpu:ram.
>
>At a 1:1 ratio my cpu speed is 3.33ghz
>
>temp is 27 C idle & 50 C while doing the prime 95 torture test
>
>Locust
>
>On Sat, 5 Jun 2004 05:09:26 -0400, "James Bald"
><Rutherford@NOSPAM.ever> wrote:
>
>>Just as a quick follow-up..
>>I just found-out about the AI overclocking that is not affecting voltages,
>>so I felt safe to make a few quick tests..
>>
>>Stock 2.8Ghz 40C
>>+10% ... 3.07Ghz.. No problemo. (Temperature rose to 43C)
>>+20%.....3.36Ghz...FSB 991Mhz... piece of cake. (Now at 47C)
>>
>>I just had to stop there.
Anonymous
a b V Motherboard
June 5, 2004 7:44:48 PM

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

Also, not overclocked and never has been.
"Jody Pellerin" <jodypell@NOSPAMnb.sympatico.ca> wrote in message
news:8Tlwc.53676$Np3.2389293@ursa-nb00s0.nbnet.nb.ca...
I have the exact same CPU and it runs at anywhere from 48-50 C on average
use.
"James Bald" <Rutherford@NOSPAM.ever> wrote in message
news:F86dnbZPGdVpe1zdRVn-hw@look.ca...
I knew there was a catch.
But I'm barely starting to study the details of this brand new home built
system yesterday.
Just started a thread on "Matching DDR" before putting my hands on a set of:
VALUE Kingston KVR400X64C32A/512 MOSEL Assy in China(2)
Onboard chips are MOSEL Vitelic corp V58C2256804SAT5
Here for Paul and alike:
http://developer.intel.com/technology/memory/ddr/specs/...
http://www.validationlabs.com/webreport/PublicModule.as...
Two sticks in Blue mobo's slots. Bought same time; Serial numbers are so
close they must come from the same wafer. :) 
One chtick is
I never expected miracles from these rams, and if you say AI overclock
reduces the ram/cpu ratio, I'm still qui te happy.
Performance and stability tests Later. SO far so good..

IMHO.. It's value allright cuz the system won't post in Turbo mode (so far),
keep in
mind I haven't finished tuning.. This is just an early experiment to
determine actual
system temperatures before getting my hands dirty...My mind already is
enough :) 

Live long and prosper Lucust,

>end user wrote in message ...
>what type of ram are you running? The AI overclocking reduces ram/cpu

>ratio to something like 2\3 in order to maintain stability. You are
>getting a higher cpu speed with a lower overall speed.
>
>On my p4p800dlx P42.8c using pc3700 ram running at a ratio of 1:1 I am
>safely using a 237 fsb (rather that the standard 200).
>
>At worse if you are using pc3200 ram, use a manual setting of 5:4
>Cpu:ram.
>
>At a 1:1 ratio my cpu speed is 3.33ghz
>
>temp is 27 C idle & 50 C while doing the prime 95 torture test
>
>Locust
>
>On Sat, 5 Jun 2004 05:09:26 -0400, "James Bald"
><Rutherford@NOSPAM.ever> wrote:
>
>>Just as a quick follow-up..
>>I just found-out about the AI overclocking that is not affecting voltages,
>>so I felt safe to make a few quick tests..
>>
>>Stock 2.8Ghz 40C
>>+10% ... 3.07Ghz.. No problemo. (Temperature rose to 43C)
>>+20%.....3.36Ghz...FSB 991Mhz... piece of cake. (Now at 47C)
>>
>>I just had to stop there.
Anonymous
a b V Motherboard
June 5, 2004 7:44:49 PM

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

Thanks Jody,
Her's for your efforts:
Merlin: Looking at the cake is like looking at the future, until you've tasted it what do you really know? And then, of course,
it's too late. Arthur takes a bite. Too late.
--------------------------------------------------------------------------------
Merlin: Good and evil, there never is one without the other.

regards,

>Jody Pellerin wrote in message ...
>Also, not overclocked and never has been.
>"Jody Pellerin" <jodypell@NOSPAMnb.sympatico.ca> wrote in message
>news:8Tlwc.53676$Np3.2389293@ursa-nb00s0.nbnet.nb.ca...
>I have the exact same CPU and it runs at anywhere from 48-50 C on average
>use.
(SNIP)
June 6, 2004 3:58:03 AM

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

In article <F86dnbZPGdVpe1zdRVn-hw@look.ca>, "James Bald"
<Rutherford@NOSPAM.ever> wrote:

> I knew there was a catch.
> But I'm barely starting to study the details of this brand new home
built system yesterday.
> Just started a thread on "Matching DDR" before putting my hands on a set of:
> VALUE Kingston KVR400X64C32A/512 MOSEL Assy in China(2)
> Onboard chips are MOSEL Vitelic corp V58C2256804SAT5
> Here for Paul and alike:
>
http://developer.intel.com/technology/memory/ddr/specs/...
> http://www.validationlabs.com/webreport/PublicModule.as...
> Two sticks in Blue mobo's slots. Bought same time; Serial numbers are so
close they must come from the same wafer. :) 
> One chtick is
> I never expected miracles from these rams, and if you say AI overclock
reduces the ram/cpu ratio, I'm still qui te happy.
> Performance and stability tests Later. SO far so good..
>
> IMHO.. It's value allright cuz the system won't post in Turbo mode (so
far), keep in
> mind I haven't finished tuning.. This is just an early experiment to
determine actual
> system temperatures before getting my hands dirty...My mind already is
enough :) 
>
> Live long and prosper Lucust,

You have to watch those wordy settings like "turbo". What that
does on some motherboards, is runs the memory with tight timings,
like 2-2-2-5, and it can also automatically boost Vcore. On one
board Vcore is lifted higher than is really reasonable. So, to
start with, you'll need some pretty good RAM to try that setting
and expect it to post. Manual overclocking, by setting the numbers
yourself, is a safer alternative, but even then, you should use a
utility to verify the DRAM settings, in case the BIOS is not really
using the settings you specify. It happens.

Paul

>
> >end user wrote in message ...
> >what type of ram are you running? The AI overclocking reduces ram/cpu
>
> >ratio to something like 2\3 in order to maintain stability. You are
> >getting a higher cpu speed with a lower overall speed.
> >
> >On my p4p800dlx P42.8c using pc3700 ram running at a ratio of 1:1 I am
> >safely using a 237 fsb (rather that the standard 200).
> >
> >At worse if you are using pc3200 ram, use a manual setting of 5:4
> >Cpu:ram.
> >
> >At a 1:1 ratio my cpu speed is 3.33ghz
> >
> >temp is 27 C idle & 50 C while doing the prime 95 torture test
> >
> >Locust
> >
> >On Sat, 5 Jun 2004 05:09:26 -0400, "James Bald"
> ><Rutherford@NOSPAM.ever> wrote:
> >
> >>Just as a quick follow-up..
> >>I just found-out about the AI overclocking that is not affecting voltages,
> >>so I felt safe to make a few quick tests..
> >>
> >>Stock 2.8Ghz 40C
> >>+10% ... 3.07Ghz.. No problemo. (Temperature rose to 43C)
> >>+20%.....3.36Ghz...FSB 991Mhz... piece of cake. (Now at 47C)
> >>
> >>I just had to stop there.
Anonymous
a b V Motherboard
June 6, 2004 4:21:58 AM

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

Does your system run at full load? If yes, 40C is much
cooler than "standard." I've a 100% full loaded Celeron
2.4G, which runs at 50C 24x7. But if I kill the background
job, and at 99.9% idle, the CPU runs at 35C. I use standard
Intel HSF, Intel mainboard and no overclock!

My 2 cents.

Stephen Wong @ Hong Kong

On Sat, 5 Jun 2004, James Bald wrote:

> Wondering if 40C / 104F is standard for a P4C 2.8 running default settings.
> That's what reports BIOS Hardware monitor right now.
>
> Very fine thread on this matter here:
> http://www.intelforums.net/viewtopic.php?t=2778
>
> He reports:
> "we tested the Northwood C under the same conditions,
> and it was marginally cooler at 104*f (40*C) peak while overclocked,
> and 87*f (31*C) at idle. "
>
> My P4 was installed today, I expect the heat-sink to take some time
> to settle properly over the CPU as the heat conductive layer should compress a bit,
> thus offering a better heat propagation to the HS.
>
> Any thoughts on that matter ?
> Regards,
>
>
>
Anonymous
a b V Motherboard
June 6, 2004 6:39:12 AM

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

Paul wrote in message ...
In article <F86dnbZPGdVpe1zdRVn-hw@look.ca>, "James Bald"
><Rutherford@NOSPAM.ever> wrote:
>> I knew there was a catch.
(.....SNIP....)
>> IMHO.. It's value allright cuz the system won't post in Turbo mode (so
>far), keep in
>> mind I haven't finished tuning.. This is just an early experiment to
>determine actual
>> system temperatures before getting my hands dirty...My mind already is
>enough :) 
>>
>> Live long and prosper Lucust,
>
>You have to watch those wordy settings like "turbo". What that
>does on some motherboards, is runs the memory with tight timings,
>like 2-2-2-5, and it can also automatically boost Vcore. On one
>board Vcore is lifted higher than is really reasonable. So, to
>start with, you'll need some pretty good RAM to try that setting
>and expect it to post. Manual overclocking, by setting the numbers
>yourself, is a safer alternative, but even then, you should use a
>utility to verify the DRAM settings, in case the BIOS is not really
>using the settings you specify. It happens.
>
> Paul


Ah, finally.
Even on my older system I could never get Turbo to work and from the manual
thought this was only some sort of burn-in setup. Made me feel 'cheap'.
No longer.

How's your P4C800 doing Paul that Zalman sink so far ?
One thing I noticed about the Zalman's desing is de 'full copper' build.
and perpendicular fins to the propeller.
IMHO, full Cu is bound to hold-up much more heat than copper+Al fins.
Copper over time is subject to oxydation; Over a number of years, My guess would be
that CuO2 layer would cripple the performance over time. Just my guess.
While Al oxide is a much better to heat transfer.
My big gripe with Copper is that Heat tends to STICK to its much higher density
than to aluminum... Hence the widespread use of Al for Heat-sink...nononsense.
I tend to faver Copper core with Al fins a lot more as a reasonnable Cu core will pull
the heat from a device quicker than AL, while Al will dissipate it so much better from
its low density and rather good heat conductivity.

Then again.. I just poked the WWW and found-out that I must have seen an older
Zalman full copper desing cuz this one is really sweet looking.
http://www.techzonez.com/forums/showthread.php?s=07d94f...
While we could learn from the AMD era and check that a full copper solution is good if the air-flow is close to mad using
rather thin fins.....
http://www.techzonez.com/forums/showthread.php?s=07d94f...

Zalman's use of copper an Al should combine very well indeed.
Yet, the airflow could be much improved if it was not so 'odd'.
IMHO.. fins should by bent more perpendicular to propeller surface to increase GREATLY the fow.

Here is the older version I saw..
http://www.casecooler.com/zacnpe4hesi.html I think this is a waste...While Full Cu is easier to manufacture.
Yet I'm not even sure the later AL-Cu version is better than a thick Intel original P4 HS design..
The airflow is much more steady, and AL dissipate plenty fast enough... Best MOD I would think is to better
the junction between CPU and HS...

I've been there pal.. Al is still the best way to get rid of heat... Copper holds to it !
A well designed combo is pinnacle.... Otherwise Full Aluminum does miracle !

Here is a mad desing that works.
http://www.viperlair.com/reviews/case_cool/coolermaster...

Check-out what Zalman does for hot Nvidia GPUs.. LOL
http://www.silentpcreview.com/modules.php?op=modload&na...
Then I read that the ATI rad 9800 HS were designed by Zalman too. I had not idea.

Cool ways to cool a Radeon 9800Pro..
http://forums.sudhian.com/messageview.cfm?catid=74&thre...
Yet Saphire model uses ZM80.

*** A solution for everything: *** (Except price)
http://www4.tomshardware.com/howto/20040115/index.html
http://test.zalman.co.kr/english/product/TNN500A.htm

Forgive me.. I build puters, but also make my own wine and this batch is Excellent;
So is my new ASUS board. :) .. Is this topic HOT or what !
Don't flame me for the OT content.
Good night !
http://www.aavidthermalloy.com/technical/papers/aircool...
June 6, 2004 3:21:03 PM

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

In article <UfadnaaDoPKkIV_dRVn-ug@look.ca>, "James Bald"
<Rutherford@NOSPAM.ever> wrote:

> Paul wrote in message ...
> In article <F86dnbZPGdVpe1zdRVn-hw@look.ca>, "James Bald"
> ><Rutherford@NOSPAM.ever> wrote:
> >> I knew there was a catch.
> (.....SNIP....)
> >> IMHO.. It's value allright cuz the system won't post in Turbo mode (so
> >far), keep in
> >> mind I haven't finished tuning.. This is just an early experiment to
> >determine actual
> >> system temperatures before getting my hands dirty...My mind already is
> >enough :) 
> >>
> >> Live long and prosper Lucust,
> >
> >You have to watch those wordy settings like "turbo". What that
> >does on some motherboards, is runs the memory with tight timings,
> >like 2-2-2-5, and it can also automatically boost Vcore. On one
> >board Vcore is lifted higher than is really reasonable. So, to
> >start with, you'll need some pretty good RAM to try that setting
> >and expect it to post. Manual overclocking, by setting the numbers
> >yourself, is a safer alternative, but even then, you should use a
> >utility to verify the DRAM settings, in case the BIOS is not really
> >using the settings you specify. It happens.
> >
> > Paul
>
>
> Ah, finally.
> Even on my older system I could never get Turbo to work and from the manual
> thought this was only some sort of burn-in setup. Made me feel 'cheap'.
> No longer.
>
> How's your P4C800 doing Paul that Zalman sink so far ?
> One thing I noticed about the Zalman's desing is de 'full copper' build.
> and perpendicular fins to the propeller. IMHO, full Cu is bound to
> hold-up much more heat than copper+Al fins. Copper over time is subject
> to oxydation; Over a number of years, My guess would be that CuO2
> layer would cripple the performance over time. Just my guess. While Al
> oxide is a much better to heat transfer. My big gripe with Copper is
> that Heat tends to STICK to its much higher density than to aluminum...
> Hence the widespread use of Al for Heat-sink...nononsense. I tend to
> faver Copper core with Al fins a lot more as a reasonnable Cu core
> will pull the heat from a device quicker than AL, while Al will
> dissipate it so much better from its low density and rather good
> heat conductivity.

One of the reasons I bought the AlCu, is no review mentioned whether
the all copper one had a protective coating on the fins or not. I've
seen some copper products coated with what looked like lacquer, and
that cannot be good for performance. Yet, without it, the shiny
finish won't last long. A nickel plate might help, but very few
products do that.

>
> Then again.. I just poked the WWW and found-out that I must have
> seen an older Zalman full copper desing cuz this one is really
> sweet looking.
> http://www.techzonez.com/forums/showthread.php?p=56862#...
> While we could learn from the AMD era and check that a full copper
> solution is good if the air-flow is close to mad using rather thin
> fins.....
> http://www.techzonez.com/forums/showthread.php?p=56862#...
>
> Zalman's use of copper an Al should combine very well indeed.
> Yet, the airflow could be much improved if it was not so 'odd'.
> IMHO.. fins should by bent more perpendicular to propeller surface
> to increase GREATLY the fow.

That is a fine observation and something I didn't really think
about too much (because reviews stated the measured noise was
low, I didn't worry about it). So far, I'm really puzzled
by the whole thing, because the fan doesn't really _blow_ the
air. What seems to happen, is there is a warm cloud around the
area outside the heatsink, and yet you don't really feel a well
defined air flow through it. My old heat sink, when you put a hand
near it, you can feel an air flow, and with that flow, you can
imagine the possibility of the CPU fan being used to cool an
adjacent Northbridge for example. I cannot see that happening
with the Zalman, as there is just a cloud of warm air around it.
Even in the Antec Sonata case, with the 120mm fan on the rear
of the case running, and the side off the case, I can still feel
that cloud of warm air. So, I guess my understanding of how it
works will have to change - initially, I thought it was the
large size and the surface area that gave it good cooling properties,
while the large aperture between fins keeps it quiet. Instead,
I think it is relying on turbulence amongst the fins, to transfer
heat from the fin into the air. Thus, bending the fins to reduce
turbulence would ruin it, if that is how it really works. It is the
strangest HSF I've played with. And, I haven't even bothered
to install MBM5 or take any readings yet :-)

One of the downsides of the Zalman, is it is exposed while you work
inside the case, and cables like to snag on it. I like to install
the HSF before putting the motherboard in the case, so I can carefully
inspect the fit. Maybe with this one, it would be better to install
it last.

The other reason I bought the AlCu, is it is lighter than the Cu
one. There are two versions - the 7000 is the original, and the
7000A contains a few extra parts for mounting os S478, S462, and
Athlon64.

>
> Here is the older version I saw..
> http://www.casecooler.com/zacnpe4hesi.html I think this is a waste...
> While Full Cu is easier to manufacture.

If you talk to a machinist, you'll find out that copper is hated
as a material. Copper doesn't cut when you machine it, it smears.
I worked with an Italian machinist, a great guy, and if you say
the word copper, he switches to Italian momentarily, to find some
choice words to say :-)

> Yet I'm not even sure the later AL-Cu version is better than a
> thick Intel original P4 HS design.. The airflow is much more steady,
> and AL dissipate plenty fast enough... Best MOD I would think is
> to better the junction between CPU and HS...
>
> I've been there pal.. Al is still the best way to get rid of heat...
> Copper holds to it ! A well designed combo is pinnacle....
> Otherwise Full Aluminum does miracle !

Copper is good for solving the "spreading angle" problem. If you
want to cool an Athlon, which doesn't have its own heat spreader,
then a copper core heatsink makes sure the heat gets spread to the
outer fins, instead of just the fins in the center. In my opinion,
that is what the copper is best at, and then the only thing the
machinist has to do, is lap the surface to make it smooth.

>
> Here is a mad desing that works.
> http://www.viperlair.com/reviews/case_cool/coolermaster...
>
> Check-out what Zalman does for hot Nvidia GPUs.. LOL
>
http://www.silentpcreview.com/modules.php?op=modload&na...
> Then I read that the ATI rad 9800 HS were designed by Zalman too.
> I had not idea.
>
> Cool ways to cool a Radeon 9800Pro..
> http://forums.sudhian.com/messageview.cfm?catid=74&thre...
> Yet Saphire model uses ZM80.
>
> *** A solution for everything: *** (Except price)
> http://www4.tomshardware.com/howto/20040115/index.html
> http://test.zalman.co.kr/english/product/TNN500A.htm
>
> Forgive me.. I build puters, but also make my own wine and this
> batch is Excellent; So is my new ASUS board. :) .. Is this topic
> HOT or what ! Don't flame me for the OT content.
> Good night !
> http://www.aavidthermalloy.com/technical/papers/aircool...

Some of those designs use heatpipes, and that is another strange
technology. The first heatpipes I saw, were simple copper
pipes, with a liquid sealed inside at one atmosphere. The liquid
would boil at the normal temp (say 67C for ethanol?). Now, to
me, this implies a non-linear response - the boiling/condensation
cycle only really runs when the thing hits 67C, and you would
think there is a potential for catastrophy if the thing gets
too much heat flux pushed through it. The thermal guys I've talked
to, say they have an extremely low thermal resistance, but I'm
not sure the ones we are seeing in computer products are that
effective.

The ones in these products use a copper pipe, but there is a
material inside the pipe with a large surface area. This forms
a capilliary surface, and allows vapor to flow no matter what
physical orientation the heat pipe is in. (May not be true of
ZM80). The fluid in some of them apparently, is not at 1 ATM,
but is under reduced pressure (partial vacuum), which causes
boiling point depression. That way, they can pick any boiling
point they want, when it is designed.

In one review I read, the reviewer cut the copper pipes as
part of his testing. He found very little "juice" inside,
and the most embarrassing part, was the cooling performance
of the product hardly changed at all. So, when I see heat pipes
on products now, I tend to be a little more sceptical of the
whole thing.

You wouldn't think there would be that much money in
making designer heatsinks, but there seems to still be
a good deal of activity.

Paul
Anonymous
a b V Motherboard
June 6, 2004 9:39:31 PM

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

>On Sat, 05 Jun 2004 15:44:04 GMT, "Jody Pellerin" <jodypell@NOSPAMnb.sympatico.ca> wrote:
>I have the exact same CPU and it runs at anywhere from 48-50 C on average
>use.

Mine sits at a comfortable 24C and hits a high of 28C during testing.
Of course I have a Thermaltake Sub Zero4G peltier cooling system on it

8-)

I've never been able to overclock this CPU by much (hea tisn't the
problem, that much I'm SURE of!) and get good stability... have 2X512
Crucial PC3200 RAM.. What settings should I be using on my P4C800
Deluxe?

Thanks!
G. Patricks


>"James Bald" <Rutherford@NOSPAM.ever> wrote in message
>news:F86dnbZPGdVpe1zdRVn-hw@look.ca...
>I knew there was a catch.
>But I'm barely starting to study the details of this brand new home built
>system yesterday.
>Just started a thread on "Matching DDR" before putting my hands on a set of:
>VALUE Kingston KVR400X64C32A/512 MOSEL Assy in China(2)
>Onboard chips are MOSEL Vitelic corp V58C2256804SAT5
>Here for Paul and alike:
>http://developer.intel.com/technology/memory/ddr/specs/...
>http://www.validationlabs.com/webreport/PublicModule.as...
>Two sticks in Blue mobo's slots. Bought same time; Serial numbers are so
>close they must come from the same wafer. :) 
>One chtick is
>I never expected miracles from these rams, and if you say AI overclock
>reduces the ram/cpu ratio, I'm still qui te happy.
>Performance and stability tests Later. SO far so good..
>
>IMHO.. It's value allright cuz the system won't post in Turbo mode (so far),
>keep in
>mind I haven't finished tuning.. This is just an early experiment to
>determine actual
>system temperatures before getting my hands dirty...My mind already is
>enough :) 
>
>Live long and prosper Lucust,
>
>>end user wrote in message ...
>>what type of ram are you running? The AI overclocking reduces ram/cpu
>
>>ratio to something like 2\3 in order to maintain stability. You are
>>getting a higher cpu speed with a lower overall speed.
>>
>>On my p4p800dlx P42.8c using pc3700 ram running at a ratio of 1:1 I am
>>safely using a 237 fsb (rather that the standard 200).
>>
>>At worse if you are using pc3200 ram, use a manual setting of 5:4
>>Cpu:ram.
>>
>>At a 1:1 ratio my cpu speed is 3.33ghz
>>
>>temp is 27 C idle & 50 C while doing the prime 95 torture test
>>
>>Locust
>>
>>On Sat, 5 Jun 2004 05:09:26 -0400, "James Bald"
>><Rutherford@NOSPAM.ever> wrote:
>>
>>>Just as a quick follow-up..
>>>I just found-out about the AI overclocking that is not affecting voltages,
>>>so I felt safe to make a few quick tests..
>>>
>>>Stock 2.8Ghz 40C
>>>+10% ... 3.07Ghz.. No problemo. (Temperature rose to 43C)
>>>+20%.....3.36Ghz...FSB 991Mhz... piece of cake. (Now at 47C)
>>>
>>>I just had to stop there.
>
>
!