Thermal paste application?

CruelCoin

Allo all,

So, i have purchased all of my components and having finished exams i can now get stuck into building my new pc.

Question is, how do i apply the thermal paste?

Most people seem to be recommending the pea method, but my worry is that this leaves the corners of the CPU uncovered. Is this an issue? Or should i do large central pea with smaller peripheral ones?

Any help/advice in this regards would be much appreciated.

Regards,

CC.

game4it70

There are so many guides and opinions on how to apply paste.
I personally think pea method is way to much.To much paste is as bad as to little.
The object is to have a thin layer to aide with thermal transfer and not a thick layer.

A little over the size of a grain of rice is what i go with.The mounting pressure of the cooler spreads it.

Johnboy1951

I put a very small blob on the CPU, and then with the tip of my finger I first spread it around very thinly and remove any excess, and then 'dab' the surface so that all areas of the CPU are covered, ensuring there is no excess to bulge out when the heatsink is attached.

I have used this method for more than 20 years and it has never failed me.

BTW if you use a small tube of thermal paste do not squeeze any out of the tube, just bring the paste to the surface and dab the top of the tube on the CPU to transfer sufficient paste .... then use your finger tip as described above.

Serephucus

The dot in the centre is the best approach.

Using anything but the cooler itself to spread the paste introduces the possibility of air pockets when the cooler applies pressure.

About a pea size blob (or smaller) is plenty. It doesn't matter about the corners as what you're applying the paste to isn't the CPU, but rather a heatspreader on top of it (the actual CPU is much smaller. See here).

BloodBath

I apply as small an amount as possible and spread it with a plastic bag over my finger so that it covers the surface of the cpu as thinly as possible.

Relying on the heatsink to spread it evenly usually doesn't work. If you remove the heatsink and look at it after using this method it's pretty patchy at best.

Serephucus

That's odd. I've almost always used this method - and checked it - and it's always worked excellently.

BloodBath

It's probably not going to make much of a difference either way anyway so don't worry about it too much.

Serephucus

True that. There'll be more of a difference with cooler placement (how much it gets wiggled about before tightening), how even/tight the screws are, etc. Even with all that, I'd say you'd be lucky to see a 5°C delta between worst and best case.

Johnboy1951

Serephucus wrote: »

The dot in the centre is the best approach.

Using anything but the cooler itself to spread the paste introduces the possibility of air pockets when the cooler applies pressure.

About a pea size blob (or smaller) is plenty. It doesn't matter about the corners as what you're applying the paste to isn't the CPU, but rather a heatspreader on top of it (the actual CPU is much smaller. See here).

Yes it is a heat spreader and is designed to help dissipate the heat into the heatsink, so all of it should receive thermal paste for most efficient dissipation.

.G.

BloodBath wrote: »

I apply as small an amount as possible and spread it with a plastic bag over my finger so that it covers the surface of the cpu as thinly as possible.

Relying on the heatsink to spread it evenly usually doesn't work. If you remove the heatsink and look at it after using this method it's pretty patchy at best.

That's how I do it too. I tried the blob in the middle, took the heatsink straight back off and the spread of the paste was very poor.

marco_polo

Hardwaresecrets did a test once and tiny dot did joint best even though not every last bit of the cpu was covered.

It certainly wins in the performance\effort ratio imo.

But then again even the "just slather it everywhere" method was only three degrees off so not really something the average pc builder need agonise for too long over.

http://www.hardwaresecrets.com/article/What-is-the-Best-Way-to-Apply-Thermal-Grease-Part-1/1303/10

BloodBath

There's too many variables for that test to be even valid.

It's using the white Ceramic paste for a start which is pretty different to the majority of silver based compounds.

Still it seems less is more. Would like to have seen a test with no paste.

marco_polo

BloodBath wrote: »

There's too many variables for that test to be even valid.

It's using the white Ceramic paste for a start which is pretty different to the majority of silver based compounds.

Still it seems less is more. Would like to have seen a test with no paste.

Oh I'd agree I wouldn't take a whole lot from the results themselves as any sort of definitive statement, apart from the fact that it doesn't seem to matter a great deal for most people how the paste is applied once it is there, a no paste test would have been very good though.

I mainly posted it to show there is really no need to worry too much if a tiny fraction of the corners are uncovered using the spot method.(I'm actually a "spreader" myself for the vast majority of my builds )