is pre-applied thermal paste good enough

[ahhnhnh]

yesterday i bought the corsair hydro h45, and i saw that it has a pre-applied thermal paste.

my question is, do i have to replace the pre-applied one or its good ?

im gonna clean the old one but should i put new after market one or the pre-applied one is ok?

thanks

 

[deathcall666]

Above average. Depends , pre-applied isn't top of the end. After all you have to be made to buy it separately.

 

[LeKeiser]

I prefer to choose my thermal paste and apply it the right way myself. Then I know it's gonna be perfect

 

[Solandri]

There's something of a cult that's developed around thermal paste, not surprisingly encouraged by the companies which sell the paste.

The truth is the paste doesn't make much difference. The job of the paste is to fill in microscopic air gaps, and air is such a great insulator that nearly anything filling in those gaps will result in better heat transfer. Even toothpaste.

http://www.hardwaresecrets.com/thermal-compound-roundup-july-2011/5/

If it makes you feel better, spend the extra $5-$10 for some specialty paste. But there's nothing wrong with using the pre-applied paste. The test results linked above are even more ambiguous if you consider that variation in applying the paste probably accounts for a +/- 2 or 3 degree difference. (Also note that the test system was still usable under full load with no paste.)

 

[LeKeiser]

What the heck is this test?? They test toothpaste with a room temperature of 13° to prove what?? That it can "act" as a thermal paste like the other real ones?? What a bunch of BS! :O
And you're pushing this even more "anything filling those gaps wil result in better heat transfer even toothpaste"??
Dude! Come on! Don't spread that kind of things!
Everyone knows that pre-applied thermal paste will do good but will never beat a I don't care which brand of thermal paste but the one that you will apply correctly over the die. Why do you think that Noctua, for example, don't apply thermal paste on their rad but add the needle in the box?

 

[Susquehannock]

Well ... typical white Toothpaste does have plenty of Aluminum Oxide (alumina). The same thermally conductive ingredient in many thermal pastes. Some products use materials which are better thermal conductors. That said, application method stands for a lot. We can usually do better than an assembly plant robot.

 

[LeKeiser]

It's not just the Aluminium Oxide or whatever they add in a thermal paste, it's also the faculty to deal with heat for a long amount of time, and I really doubt that toothpaste would last that long...

 

[Susquehannock]

^^ agreed. It has been subject in many tests and dried up quickly. Most TIM use oils for that reason.

Rather partial to the test where they used Cheese spread 😀

 

[Solandri]

What the heck is this test?? They test toothpaste with a room temperature of 13° to prove what?? That it can "act" as a thermal paste like the other real ones?? What a bunch of BS! :O

Air has about two orders of magnitude lower heat transfer coefficient than most pasty substances. That is, most pastes conduct heat hundreds of times better than air. That's why there isn't much difference between the different brand pastes or toothpaste. The difference between toothpaste and the more expensive thermal pastes is tiny (about 1-5x) compared to the difference between toothpaste and air (100-500x).

Likewise, metal has about two orders of magnitude higher heat transfer coefficient than most pastes. That's why the CPU is able to operate at full load without any paste. The vast majority of the heat transfer is via metal-on-metal contact (CPU's heat spreader to heat sink). Even with the air gaps, enough heat is transferred out to prevent the CPU from overheating under full load. This is why it's very important not to use too much paste - that would reduce the metal-on-metal contact area.

And BTW, the rate of heat transfer is proportional to temperature differential. So if they had tested at 23 C ambient air temperature instead of 13 C, all other things (e.g. fan speed) being equal it would just add 10 C to the CPU temps they measured across the board.

 

[LeKeiser]

If I'm not mistaken, the latest Intel CPUs have a problem with the thermal past under the cap that's not properly applied, thus resulting on temperature being really really hot, too hot. So I guess if there is some thermal paste underneath the cap, that means that thermal paste is somehow a necessity, correct?
Question Solandri: did you apply some thermal paste on your CPU?