Thermal Paste Comparison

pugetsystems · May 31, 2005, 11:25 PM

This article compares the performance of the top rated thermal pastes, so that you can make informed decisions about what will go into your next computer system! Read More,

pugetsystems

Zardon · May 31, 2005, 11:26 PM

Even though its blatant advertising, ill let it slide because you put some effort into it. moved to news.

Lelisevis · May 31, 2005, 11:43 PM

Coolermaster premium thermal compound kit?
Coolermaster Highperformance compound kit?
Akasa AK-450 Silver paste?

Surely these are quite popular pastes too?

CDsDontBurn · May 31, 2005, 11:54 PM

Quote:

Test Results

Idle Load
Arctic Silver Ceramique 25.5C 37.5C
Thermalright Generic 24.5C 38.5C
Arctic Silver 5 25C 39C
Shin-Etsu Micosi 25C 39C
Arctic Alumina 26.5C 41.5C
Nanotherm Silver Xtc 26C 43C
Nanotherm PCM+ (phase change) 26.5C 43.5C

as5 (in bold) is supposed to be the best stuff out there. now anybody who knows, would let it set for at least 3 days before making any temperature readings with the stuff so that i can work properly. The other stuff, i don't know how long it takes for it to set in and properly cool

BWX · Jun 1, 2005, 06:06 AM

I use Ceramique. -- safe and effective.

http://www.arcticsilver.com/ceramique.htm

Quote:

During the system's initial use, the heat from the CPU lowers the viscosity of the compound to enhance the filling of the microscopic valleys and ensure a minimum bond line between the heatsink and the CPU core. [COLOR=Red]Then the compound thickens slightly over the next 100 to 300 hours of use to its final consistency designed [/COLOR] for long-term stability.

mike2h · Jun 1, 2005, 07:30 AM

ceramique user here too. do not know why anybody would not use it.
as soon as my arm heals enough so i can do the lifting that is what i will be using with my new 3.2 prescott.

i agree with cddb, it can take up to a week for some of that stuff to 'cure'.

Zorg · Jun 1, 2005, 10:00 AM

All I can say here - get enough air through your case and the quality of the thermal grease is insignificant. The GPU/CPU really does not care about running at 37.5C or 39C! Both are perfectly fine

CDsDontBurn · Jun 1, 2005, 11:34 AM

Quote:

Originally Posted by Zorg

All I can say here - get enough air through your case and the quality of the thermal grease is insignificant. The GPU/CPU really does not care about running at 37.5C or 39C! Both is perfectly fine

true, but usually when the CPU/GPU run cooler at idle temps, they will also run cooler under load temps. Therefore, you will come across less heating issues with the thermal paste/compount/adhesive that is being used.

bananaman · Jun 1, 2005, 11:45 AM

ceramique here aswell

works fine here, used it for quite a while now having moved from Artic Silver - didnt really notice much difference, but every little helps

mike2h · Jun 1, 2005, 05:34 PM

Quote:

Originally Posted by Zorg

All I can say here - get enough air through your case and the quality of the thermal grease is insignificant. The GPU/CPU really does not care about running at 37.5C or 39C! Both is perfectly fine

while this is basically true it does not apply to people who really want to oc. every couple of degrees helps.

Zardon · Jun 1, 2005, 05:45 PM

Quote:

Originally Posted by mike2h

while this is basically true it does not apply to people who really want to oc. every couple of degrees helps.

Yes, and even more so if you raise the core voltage.

Matth · Jun 1, 2005, 06:09 PM

http://www.dansdata.com/goop.htm

Some interesting results, and some interesting limiting cases... for his test rig, the WORST was 0.66 C/W with no paste
The BEST was 0.41 C/W - with WATER

An average compound seemed to do about 0.5 C/W - this is the COMBINED figure for compound and heatsink. The results for the thermastrate stichs and pads (link on the dansdata article) are also impressive, ther's surely a market for someone buying a roll and breaking it down into enthusiast sized packs, and the squares are even better for people who cannot apply thin/even compound if their life depended on it!

The supreme performance of water suggests that the ideal would actually be a permanently non-drying oil with no fillers.

It would also seem that the performace of the paste would tend to be less critical on larger contact areas, just so long as it does fill the gaps without impairing any metal to metal contact.

The tendency toward metal-capped processors probably makes things a lot easier for the heatsink paste, compared to small, open cores.

BWX · Jun 1, 2005, 07:51 PM

Quote:

Originally Posted by Matth

http://www.dansdata.com/goop.htm

Some interesting results, and some interesting limiting cases... for his test rig, the WORST was 0.66 C/W with no paste
The BEST was 0.41 C/W - with WATER

.

Review date: 13 March 2002.

Quote:

That's right - Toothpaste Superior To Arctic Silver 3! Film At Eleven!

I think that whole review is bunk-

he didn't wait the proper amount of time either. No thermal pad works as good as well applied and cured thermal paste either-
NOT toothpaste..

mike2h · Jun 1, 2005, 08:15 PM

while i agree with the rest of your statement i do not think the whole reveiw was bunk. it does give a baseline for what to generally expect. i would really like to see a side by side done a week later though.

BWX · Jun 1, 2005, 08:28 PM

Well yeah, I mean that is why it is bunk I think- he didn't do the proper "break in" for the thermal paste to properly do it's job..

Matth · Jun 2, 2005, 06:55 PM

The phase change stuff was baked, and improved, but then the phase change material needs to worked - hang about though, if it REALLY cooked, surely all the phase change stuff would disappear, even where you actually wanted it?

I do recall that AS is supposed to improve with a few thermal cycles through it (hence it usually does poorly on a "I switched to AS and it was no better" kind of test).

The burning question is, just how much improvement is there to be had, between generic paste and the high priced alternatives, and could you get just as much improvement for the same price increment in your choice of heatsink?

So the paste will do 10 or more fittings? - but if you only fit the heatsink once, then that's the entire cost of the paste to use it once.

mike2h · Jun 2, 2005, 07:05 PM

i must have missed something?

DogSnake · Jun 2, 2005, 07:25 PM

I have used artic ceramix and Antec silver on the cpu. Currently the silver is on the P4/3.0 cpu with an AC Freezer sink and fan and the ceramix on the ATI 9800 Pro/256 vga. I think a balance of cooling devices is more important to the system operation than the paste. Both have been adequate on the cpu.

mike2h · Jun 2, 2005, 07:45 PM

putting active cooling in the same category of importance as paste is ludicrous. i do not understand why some of you keep doing this. this thread has to do with the the function/cost effectivness of paste. its role is at best minor & is directly proportional to how hard you push your system.
paste is an option, case/cpu/vga cooling is not.

The_Neon_Cowboy · Jun 2, 2005, 09:42 PM

Quote:

Originally Posted by BWX

I use Ceramique. -- safe and effective.

http://www.arcticsilver.com/ceramique.htm

I get it in massive tubes for ubber cheap and it works great and doesn't void your cpu warrently like other silver pastes do. Not to mention it's far less conductive.

BWX · Jun 2, 2005, 10:07 PM

mike2h I agree- the whole idea of comparing thermal paste is for very small gains really- the amount you put on and the way you put it on can have just as much of an effect as the difference of the brands of paste.

However when I had a CPU/mobo combo once with generic past on it installed by someone else- and then changed to AS3 that I applied correctly, I gained 5 to 8 degrees C at full load in the summer time- it was pretty hot..
This was 4 years ago with a Palomino.. It allowed for a much higher OC back then.

Yeah Neon, Ceramique is great stuff. I doubt I will ever use anything else, unless they upgrade the ceramique... Its safe, and works just about as good as AS3, AS5, etc.. I put it on every single CPU, GPU, VPU, I have worked on since Ceramique first came out..

May 31, 2005, 11:25 PM	#1
pugetsystems DriverHeaven Newbie Join Date: Feb 2005 Location: Seattle, WA Posts: 8 Rep Power: 0	Thermal Paste Comparison This article compares the performance of the top rated thermal pastes, so that you can make informed decisions about what will go into your next computer system! Read More, pugetsystems

Jun 1, 2005, 07:30 AM	#6
mike2h DriverHeaven Extreme Member Join Date: Nov 2002 Location: norcal Posts: 6,483 Rep Power: 91 System Specs	ceramique user here too. do not know why anybody would not use it. as soon as my arm heals enough so i can do the lifting that is what i will be using with my new 3.2 prescott. i agree with cddb, it can take up to a week for some of that stuff to 'cure'. __________________

Jun 1, 2005, 10:00 AM	#7
Zorg DriverHeaven Lover Join Date: Nov 2004 Location: Australia Posts: 182 Rep Power: 0 System Specs	All I can say here - get enough air through your case and the quality of the thermal grease is insignificant. The GPU/CPU really does not care about running at 37.5C or 39C! Both are perfectly fine Last edited by Zorg; Jun 2, 2005 at 01:13 AM.

Jun 1, 2005, 11:45 AM	#9
bananaman banana muncher Join Date: Dec 2002 Location: London Posts: 1,254 Rep Power: 64	ceramique here aswell works fine here, used it for quite a while now having moved from Artic Silver - didnt really notice much difference, but every little helps __________________

Jun 1, 2005, 08:15 PM	#14
mike2h DriverHeaven Extreme Member Join Date: Nov 2002 Location: norcal Posts: 6,483 Rep Power: 91 System Specs	while i agree with the rest of your statement i do not think the whole reveiw was bunk. it does give a baseline for what to generally expect. i would really like to see a side by side done a week later though. __________________ Last edited by mike2h; Jun 2, 2005 at 01:02 AM.

May 31, 2005, 11:26 PM	#2
Zardon DriverHeaven Founder Join Date: May 2002 Posts: 32,480 Rep Power: 179	Even though its blatant advertising, ill let it slide because you put some effort into it. moved to news.

May 31, 2005, 11:43 PM	#3
Lelisevis Howlin at the moon Join Date: Sep 2003 Posts: 1,663 Rep Power: 0	Coolermaster premium thermal compound kit? Coolermaster Highperformance compound kit? Akasa AK-450 Silver paste? Surely these are quite popular pastes too?

Jun 1, 2005, 06:09 PM	#12
Matth Flash Banner Hater Join Date: Jun 2002 Location: UK Posts: 3,426 Rep Power: 93 System Specs	http://www.dansdata.com/goop.htm Some interesting results, and some interesting limiting cases... for his test rig, the WORST was 0.66 C/W with no paste The BEST was 0.41 C/W - with WATER An average compound seemed to do about 0.5 C/W - this is the COMBINED figure for compound and heatsink. The results for the thermastrate stichs and pads (link on the dansdata article) are also impressive, ther's surely a market for someone buying a roll and breaking it down into enthusiast sized packs, and the squares are even better for people who cannot apply thin/even compound if their life depended on it! The supreme performance of water suggests that the ideal would actually be a permanently non-drying oil with no fillers. It would also seem that the performace of the paste would tend to be less critical on larger contact areas, just so long as it does fill the gaps without impairing any metal to metal contact. The tendency toward metal-capped processors probably makes things a lot easier for the heatsink paste, compared to small, open cores.

Jun 1, 2005, 08:28 PM	#15
BWX unplugged Join Date: Nov 2002 Location: USNY Posts: 19,669 Rep Power: 110 System Specs	Well yeah, I mean that is why it is bunk I think- he didn't do the proper "break in" for the thermal paste to properly do it's job.. __________________

Jun 2, 2005, 06:55 PM	#16
Matth Flash Banner Hater Join Date: Jun 2002 Location: UK Posts: 3,426 Rep Power: 93 System Specs	The phase change stuff was baked, and improved, but then the phase change material needs to worked - hang about though, if it REALLY cooked, surely all the phase change stuff would disappear, even where you actually wanted it? I do recall that AS is supposed to improve with a few thermal cycles through it (hence it usually does poorly on a "I switched to AS and it was no better" kind of test). The burning question is, just how much improvement is there to be had, between generic paste and the high priced alternatives, and could you get just as much improvement for the same price increment in your choice of heatsink? So the paste will do 10 or more fittings? - but if you only fit the heatsink once, then that's the entire cost of the paste to use it once.

Jun 2, 2005, 07:05 PM	#17
mike2h DriverHeaven Extreme Member Join Date: Nov 2002 Location: norcal Posts: 6,483 Rep Power: 91 System Specs	i must have missed something? __________________

Jun 2, 2005, 07:25 PM	#18
DogSnake DriverHeaven Newbie Join Date: Sep 2003 Posts: 1 Rep Power: 0	I have used artic ceramix and Antec silver on the cpu. Currently the silver is on the P4/3.0 cpu with an AC Freezer sink and fan and the ceramix on the ATI 9800 Pro/256 vga. I think a balance of cooling devices is more important to the system operation than the paste. Both have been adequate on the cpu.

Jun 2, 2005, 07:45 PM	#19
mike2h DriverHeaven Extreme Member Join Date: Nov 2002 Location: norcal Posts: 6,483 Rep Power: 91 System Specs	putting active cooling in the same category of importance as paste is ludicrous. i do not understand why some of you keep doing this. this thread has to do with the the function/cost effectivness of paste. its role is at best minor & is directly proportional to how hard you push your system. paste is an option, case/cpu/vga cooling is not. __________________

Jun 2, 2005, 10:07 PM	#21
BWX unplugged Join Date: Nov 2002 Location: USNY Posts: 19,669 Rep Power: 110 System Specs	mike2h I agree- the whole idea of comparing thermal paste is for very small gains really- the amount you put on and the way you put it on can have just as much of an effect as the difference of the brands of paste. However when I had a CPU/mobo combo once with generic past on it installed by someone else- and then changed to AS3 that I applied correctly, I gained 5 to 8 degrees C at full load in the summer time- it was pretty hot.. This was 4 years ago with a Palomino.. It allowed for a much higher OC back then. Yeah Neon, Ceramique is great stuff. I doubt I will ever use anything else, unless they upgrade the ceramique... Its safe, and works just about as good as AS3, AS5, etc.. I put it on every single CPU, GPU, VPU, I have worked on since Ceramique first came out.. __________________