is this to be trusted?

[Uwillparish]

https://www.pugetsystems.com/labs/articles/Thermal-Paste-Application-Techniques-170/#TestSetup

 

Is the X shape really the best? have i been doing it wrong the whole time?

[Enderman]

no

pea in the middle is best

 

the center of the CPU is what gets the hottest because that is where the chip touches the IHS

having air near the edges or corners is not as temperature-affecting as having air bubbles in the center

 

either way there is not much difference, but the pea method is generally better

[79wjd]

There is no answer to this. Everyone thinks their way is best. 

 

But ultimately, pea vs. X won't yield that much of a difference in temps. 

[LIGISTX]

I mean... everyone does the dot method, and that clearly results in pretty much the same temps... sooooo. lol

 

Is interesting tho

[Toxicable]

https://www.pugetsystems.com/labs/articles/Thermal-Paste-Application-Techniques-170/#TestSetup

 

Is the X shape really the best? have i been doing it wrong the whole time?

I wouldnt not trust this, the difference is very little between them that there would have been some other contributing factor.

If you didn't know, underneath the heat spreader(the top of that CPU that you can see) is the actual die which is not the same size as the heat spreader. So it actually not too important to cover the whole heat spreader.

[WoodenMarker]

I'm not sure if the acrylic they're using bends more than the plate at the bottom of a cpu cooler which might be leading to the air bubbles. Also, the amount of thermal paste used varying might have been the cause of the change in temps. 

[Colt_0pz]

all im gonna say is that, less is better since all you are doing is closing the gap of air between the cpu and the cooler.

[Uwillparish]

ok, well, now to take my system apart to update a motherboards bios, then use the same old pea method 

[AlwaysFSX]

all im gonna say is that, less is better since all you are doing is closing the gap of air between the cpu and the cooler.

Yes, let's put a tiny tab and run in to overheat problems.

[Godlygamer23]

ok, well, now to take my system apart to update a motherboards bios, then use the same old pea method 

From the results they give in their post, I would say it's not a big enough difference to worry about.

[Uwillparish]

From the results they give in their post, I would say it's not a big enough difference to worry about.

then why do we reapply thermal compound on GPU's?

[Godlygamer23]

then why do we reapply thermal compound on GPU's?

I'm talking about the application techniques. Not reapplying thermal paste.

[WoodenMarker]

then why do we reapply thermal compound on GPU's?

Thermal paste may dried out or been badly applied--generally due to the sheer amount more than the shape of application.

[Uwillparish]

Thermal paste may dried out or been badly applied--generally due to the sheer amount more than the shape of application.

.......oh my poor 780........

 

this is a joke, i did re-apply, again when i watercooled it, all is good

 

I'm talking about the application techniques. Not reapplying thermal paste.

and i was talking about temp difference 

[Colt_0pz]

Yes, let's put a tiny tab and run in to overheat problems.

nah. all thermal paste does is close the gap between the cooler and the processor