Seems like X works well

Interesting, I always used the spread method but the X method does seem to be the best.
 
Didn't think about the air pockets if I do the spread method. And Ive been using that method for years! D:

I have never heard of the x method. Looks like it works very well.

Great guide thanks for taking the time to write this.

for me the Happy Face works best
 
good show

very informative man! 

you might want to test out spreading on bigger dies like on the nvidia cards....those are hard to get completely covered.  

Great info here!

my ek water block for a GPU told me to X like the British flag. Is that because the die is so big?

Just stumbled across this.....funny how the method "rice grain" size hasnt changed even though CPU's have gotten wayyyyyyyy bigger!! Its obvious a Rice Grain size is not enough and will result in a fairly substantial part of the Chip being bare of TIM......I used to use pea size on my 1366 and now have gone to the line method on my 2011 Chip!

Somebody had a visual diagram of the difference between intel and amd chips. Ones concave and the other is convex. I believe that would effect tim application as well. ie using more or less based on the chip being used. Its also a slight misconception on the actual spread of our tims on said parts, cpu or gpu. We may not get the same results due to us not being able to apply as much pressure as you did on the two pieces of glass, or not knowing how much is enough to get spread. Sure, you could try the pea method, put your parts together, run it a few days, then pull it back off to check spread then reapply.

You would have figured somebody would have tested which amount is close to perfect without having excess, then just sell that exact amount per chip, cpu or gpu respectively