Hardware Secrets Forums


Go Back   Hardware Secrets Forums > Miscellaneous > Content Comments



Reply
 
Thread Tools Search this Thread Display Modes
Old 03-01-2011, 01:20 PM   #1
Hardware Secrets Team
Administrator
 
Join Date Nov 2004
Posts: 5,599
Hardware Secrets Team is on a distinguished road

Default Thermal Compound Roundup - March 2011

There has been a new article posted.

Title: Thermal Compound Roundup - March 2011
URL: http://www.hardwaresecrets.com/artic...arch-2011/1207

Here is a snippet:
"Following up our Thermal Compound Roundup - February 2011*review, we are adding another five thermal compounds to our roundup, for a grand total of 10 different models from Cooler Master, Coolink, Dee..."

Comments on this article are welcome.

Best regards,
Hardware Secrets Team
http://www.hardwaresecrets.com
Hardware Secrets Team is offline   Reply With Quote
new Sponsored Links

Old 03-01-2011, 05:40 PM   #2
actionjksn
Senior Member
 
Join Date Dec 2009
Posts: 29
actionjksn is on a distinguished road

Default Curing time

Could you check the curing time myth with Arctic Silver 5 on the next round up? It is one of the brands that supposedly requires the curing time to reach optimal performance. It also just happens to be the one that I use. One of the really thick ceramic products would probably also be a prime candidate for that experiment too. Another fun experiment to try would be to lap the heatsink and maybe even the processor, and see how much temp reduction it's good for. I always lap my heatsink, but I have never done a before and after comparison to see what I'm gaining. I have always heard 3 to 5 degrees which sounds pretty reasonable. Probably depends on how the base looks before it gets lapped, a rough one would probably gain the most. It takes a long time for just a minor gain but I like to do it anyway.
PS It looks like the Thermalright is the worst performer so far.
actionjksn is offline   Reply With Quote
Old 03-01-2011, 05:54 PM   #3
Rafael Coelho
Super Moderator
 
Rafael Coelho's Avatar
 
Join Date Jun 2008
Posts: 182
Rafael Coelho is on a distinguished road

Default

Quote:
Originally Posted by actionjksn View Post
Could you check the curing time myth with Arctic Silver 5 on the next round up? It is one of the brands that supposedly requires the curing time to reach optimal performance. It also just happens to be the one that I use. One of the really thick ceramic products would probably also be a prime candidate for that experiment too. Another fun experiment to try would be to lap the heatsink and maybe even the processor, and see how much temp reduction it's good for. I always lap my heatsink, but I have never done a before and after comparison to see what I'm gaining. I have always heard 3 to 5 degrees which sounds pretty reasonable. Probably depends on how the base looks before it gets lapped, a rough one would probably gain the most. It takes a long time for just a minor gain but I like to do it anyway.
PS It looks like the Thermalright is the worst performer so far.
We will test more compounds for curing time benefits soon, and we will try to include AS5 among them.
Make a "lapping review" is a good idea, we took note of it.

PS: Maybe it is the reason the manufacturer no longer sells this product... They sell the "Chill Factor3" paste nowadays.

Last edited by Gabriel Torres; 03-02-2011 at 07:40 AM. Reason: typo
__________________
There´s nothing as practical as a good theory.
Rafael Coelho is offline   Reply With Quote
Old 03-01-2011, 11:11 PM   #4
Olle P
Senior Member
 
Olle P's Avatar
 
Join Date Oct 2008
Sweden
Posts: 1,736
Olle P is on a distinguished road

Default

Quote:
Originally Posted by actionjksn View Post
... lap the heatsink and maybe even the processor, and see how much temp reduction it's good for. ... Probably depends on how the base looks before it gets lapped, a rough one would probably gain the most.
The cooler used in the test suite should gain just about nothing.
As you note yourself it's rough surfaces that benefit from lapping.

Lapping the CPU heat-spreader shouldn't do nearly as much as removing it completely. With the heatsink directly on top of the CPU dice, removing one layer of TIM and thermal resistance added by the spreader can do wonders.
The provisions are that the cooler base gets in good contact with the dice and that the base is a good heat spreader. No "direct touch" heatpipes allowed!
(The reason for the CPU's IHS to be there in the first place is to make mounting of the cooler less of a precision job.)

Now to the curing issue:
- Noctua state that there's no "Burn-in" time for NT-H1, so testing that one was doomed to provide the given result.
- Spire do claim that "BlueFrost thermal compound will begin to function at its optimum potential, after a gradual increase in heat-dissipation, approximately 168 hours (7days)after initial application.".

Normally the curing isn't just a matter of time, but also about heat. The computer needs to be run at varying loads to get the process going.
Unfortunately Rafael doesn't mention how the computer was used during the curing period.

Anyway, even those compounds that do benefit from curing doesn't gain much as a result. It might make up for an additional degree or two in reduced temperature.

Cheers
Olle
Olle P is offline   Reply With Quote
Old 03-02-2011, 02:38 AM   #5
Rafael Coelho
Super Moderator
 
Rafael Coelho's Avatar
 
Join Date Jun 2008
Posts: 182
Rafael Coelho is on a distinguished road

Default

Olle, the testbed computer was running the Folding@Home SMP Client all the time between the measurements, which is near full-load.
__________________
There´s nothing as practical as a good theory.
Rafael Coelho is offline   Reply With Quote
Old 03-02-2011, 09:56 PM   #6
parsec
Senior Member
 
Join Date Jul 2010
Posts: 50
parsec is on a distinguished road

Default

Thanks for performing a test without any TIM, I have never seen that done before and I was curious about the results. I did not think the difference would be that great.

There are quite a few articles about removing the IHS from CPUs, which is simply not worth it with current, higher-end CPUs, IMO. Those same high performance CPUs have the IHS soldered to the rest of the CPU, with solder even surrounding the CPU die itself apparently (Olle P's post, a translation I think, uses "dice" instead of "die"). Most CPUs do not survive the IHS removal process, I'll pass, thank you.

Also, I would say there is a reason why an Integrated Heat Spreader is used, much more surface area to transfer heat to the heat sink. Still, the IHS-removal club claim lower CPU temperatures without the IHS, so direct contact beats less-direct larger surface area contact. As I recall, older AMD CPUs had naked die's.
parsec is offline   Reply With Quote
Old 03-03-2011, 01:04 AM   #7
Olle P
Senior Member
 
Olle P's Avatar
 
Join Date Oct 2008
Sweden
Posts: 1,736
Olle P is on a distinguished road

Default

Quote:
Originally Posted by parsec View Post
As I recall, older AMD CPUs had naked die's.
All CPUs, no matter what manufacturer, were "naked" before the Pentium4.

The Athlon XP series with its combination of small dies and fairly high power resulted in quite a few overheatings and physically broken CPUs when heatsinks were mounted improperly. That lead to Intel introducing the IHS to allow more slack in mounting the heatsink while at the same time providing some physical protection to the die underneath.

I didn't know that the IHS is hardmounted to the die nowadays. It certainly wasn't on the P4s.

Cheers
Olle
Olle P is offline   Reply With Quote
Old 03-04-2011, 07:21 PM   #8
Azure
Junior Member
 
Azure's Avatar
 
Join Date Mar 2010
Posts: 20
Azure is on a distinguished road

Default

I think it is okay to exceed amount.

I mean, most CPU heatsinks is able to apply enough force to push the thermal compounds.

and exceeded amount will be pushed out and spread around CPUs. It's okay to do this because most thermal compounds are non electric conductive.

I saw some sort of foreign benchmark for how to spread thermal compound, and exceeding amount of thermal compound doesn't hurt temps since you sit CPU heatsink firmly.

I personally prefer exceeding a bit, because lack of thermal compounds lead higher temps, and knowing exact amount of thermal compound is bit pain.
Azure is offline   Reply With Quote
Old 03-04-2011, 09:52 PM   #9
gregzeng
Junior Member
 
Join Date Mar 2011
Australian Capital Territory
Posts: 1
gregzeng is on a distinguished road

Default

NO-NAMES compound, as factory fitted on notebook computers?

My last 2 HP PAVILON notebooks were desktop replacers, using desktop CPUs (years ago). After each crashed, just outside warranty, being used as desktop replacements, 24/7, I opened each.

NO HEATSINK PASTE/ TAPE.

COMPAQ is the down-market version of HP. Is HP the General Motors of the USA? Could u check what is supplied in notebooks, & replacing with a brand-name paste/ tape ...

Australian Capital Territory
gregzeng is offline   Reply With Quote
Old 03-06-2011, 10:36 PM   #10
Olle P
Senior Member
 
Olle P's Avatar
 
Join Date Oct 2008
Sweden
Posts: 1,736
Olle P is on a distinguished road

Default

Quote:
Originally Posted by gregzeng View Post
My last 2 HP PAVILON ... each crashed, just outside warranty, ...
NO HEATSINK PASTE/ TAPE.
That's scary! ... but could explain why we've had so much trouble at work with the HP notebooks used to replace Fujitsu ditto over the past few years.

20 years ago HP was about as good as you could get in terms of computer hardware quality. That doesn't hold true any more...

Cheers
Olle
Olle P is offline   Reply With Quote
new Sponsored Links

Reply

Share This Thread

Bookmarks

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Forum Jump


All times are GMT -8. The time now is 08:38 PM.


vBulletin Security provided by vBSecurity (Lite) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
vBulletin Optimisation provided by vB Optimise (Lite) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd. ()
© 2004-12, Hardware Secrets, LLC. All rights reserved.