Question on Thermal Paste

[ConveX]

This might sound a bit silly, but I've been wondering if different brands of thermal paste work better than the stock oones that come with CPU/CPU coolers when it comes to cooling, etc. If it is the case then what brand should I look for? Cheers

 

[silkstone]

Yes.

http://www.hardwaresecrets.com/article/Thermal-Compound-Roundup-January-2012/1468

 

[ogharaei]

I like Prolimatech PK-3 and Arctic MX-4.

 

[BiggieShady]

Yes.

http://www.hardwaresecrets.com/article/Thermal-Compound-Roundup-January-2012/1468

Mayonnaise has pretty respectable results.

 

[LagunaX]

Be careful of some thermal pastes that have a pump out effect - they are great initially but degrade over time such as the Noctua thermal paste.
MX-4 works great for me. IC Diamond wasn't bad either.

 

[newtekie1]

This might sound a bit silly, but I've been wondering if different brands of thermal paste work better than the stock oones that come with CPU/CPU coolers when it comes to cooling, etc. If it is the case then what brand should I look for? Cheers

Yes and no...

The link Silkstone provided is a great resource. It basically tells you that the difference betweem the worst paste and the best is only 7°C, and most pastes fall in a 5°C area. So for anything but an extreme overclock it isn't going to matter. Also, the pastes that come with some coolers, even stock ones from Intel/AMD, is actually very good. I know Intel actually uses Sin-Etsu paste on their coolers, and the paste Corsair uses on their coolers is amazing(not sure what it is though). At this point I keep a tube of Artic Alumina around and use it exclusively when I'm re-pasting something because it is extremely cheap(a 9cc tube is $7.50) and the few ° difference is not going to make any difference.

 

[the54thvoid]

Yes and no...

The link Silkstone provided is a great resource. It basically tells you that the difference betweem the worst paste and the best is only 7°C, and most pastes fall in a 5°C area. So for anything but an extreme overclock it isn't going to matter. Also, the pastes that come with some coolers, even stock ones from Intel/AMD, is actually very good. I know Intel actually uses Sin-Etsu paste on their coolers, and the paste Corsair uses on their coolers is amazing(not sure what it is though). At this point I keep a tube of Artic Alumina around and use it exclusively when I'm re-pasting something because it is extremely cheap(a 9cc tube is $7.50) and the few ° difference is not going to make any difference.

What he said.

But where they can fall down is that they apply the paste pretty poorly sometimes. Sometimes if your temps are wonky, you can take off a cooler and see a paste party going on. In those cases, re-applying the correct amount can do wonders.

 

[newtekie1]

What he said.

But where they can fall down is that they apply the paste pretty poorly sometimes. Sometimes if your temps are wonky, you can take off a cooler and see a paste party going on. In those cases, re-applying the correct amount can do wonders.

That is an extremely excellent point. In fact, I'd say proper application will affect temps far more than the type of paste used.

 

[ConveX]

Thanks for all the replies guys I'll look into some videos on how to "properly" apply the thermal paste.

 

[[Ion]]

Thanks for all the replies guys I'll look into some videos on how to "properly" apply the thermal paste.

The hard part (IMO) is using the right amount--I have a tendency to slather on twice as much as is necessary, which is not only more expensive but it also doesn't work as well.

Amazingly, Lenovo didn't use any on my Thinkpad--I took it apart to clean out dust, and found that it was mostly dust-free, but adding thermal paste dropped the temps over 10C.

 

[ALMOSTunseen]

Here are some of the more popular methods of thermal paste application:
http://benchmarkreviews.com/index.php?option=com_content&task=view&id=170&Itemid=1&limit=1&limitstart=4

 

[newtekie1]

The single dot method has worked well for me for years. Since the CPU core is in the center of the IHS, you really only have to make sure that is covered, the whole silver part doesn't need to be covered to get proper thermal transfer. I know spreading it out is a horrible idea because it creates air bubbles, which is the enemy of thermal transfer.

 

[ur6beersaway]

I wonder if anyone has used these ? Of course is $20 out of control..

http://www.frozencpu.com/cat/l3/g8/...ials-Indigo_Xtreme_ETI-Page1.html?id=ENwg9MBW

 

[silkstone]

The single dot method has worked well for me for years. Since the CPU core is in the center of the IHS, you really only have to make sure that is covered, the whole silver part doesn't need to be covered to get proper thermal transfer. I know spreading it out is a horrible idea because it creates air bubbles, which is the enemy of thermal transfer.

I like to put a dot in the middle and the a smaller dot in each corner linked my thin lines coming from the dot in the center, forming a kind of x-shape.

I doubt it makes much difference from just having the core covered with paste, but i like to have the whole IHS make thermal contact with the HS.

 

[jboydgolfer]

I have ALWAYS just gone to the Local RadioShack in My town, and Grabbed Two tubes of Arctic Silver 5, and keep them on hand for any emergencies that I may encounter. Application For Me, is a method I developed myself, where I cut the head off of a Que tip(cotton Swab), and use the Cotton-less Shaft of the Swab to "peak" the thermal Paste on the GPU/CPU. So that the Entire surface of the Chip is covered with Thermal Paste(kind of looks like a Silver ocean with small Waves). I feel there is NO actual Right way to Do it, But there is definitely WRONG ways, and as far as I'm concerned, My way works BEST for ME. My 6950 Avg's 23c-26c on air, and Maxes @ 57c , running Arma 2 @ high settings. After a few application attempts You'll get it, and You'll determine which method yielded the Lowest temp's For you.
Best of Luck.