Originally Posted by Swag
I'm receiving my Asus RMA board on Monday, I've been getting interested in trying out different TIM applications and I think I'm going to use the spread smoothly method this time and see how the temps go. Any thoughts on the matter of TIM application?
I was testing with AS5 because it is cheap, easy to use, and easy to clean. That was helpful in my 30 or so tests of mounting a cooler to an IHS. No two coolers are the same, and no two IHS are the same, therefore there is no "best" way, only what is best for your setup.
I have tested three 3770k, one Antec cooler, and two Corsair coolers. The best way to know is try different combinations and different amounts of TIM. Then there is orientation of the cooling block, and how you tighten it down (not only how tight but which screw in which order). For example, in my current setup the best solution, by far, is the cooler block upside down and screwed down at the top first.
So you try single drop, line, spread, and each one of those can be varied as too how much. Additionally there are usually four positions for the cold plate orientation of your cooler. Each test you check temps and then look at the "fingerprint" that the TIM leaves on both surfaces. You will quickly discover that fingerprint tells you a lot about how effective the cooling will be.
I also found that you can absolutely use too little Ultra on the die. There was so much talk about everyone using too much, I tried multiple times using less and less. Too little is very possible to do, so you just have to test it. This again is chip dependent because the shape of your specific IHS will determine the optimal amount of TIM to be used on both the die and on top of the IHS.
The proper amount on the die is more important than the proper amount on top of the IHS though. But most people don't care about a few degrees either way, so as long as you get the TIM application on the die as close to perfect (for your particular chip) as possible, then you will get the best results. An OK application on top of the IHS vs. a perfect makes a difference, but not as much as getting right with the TIM application on the die. Is that a long enough answer for you?Edited by justanoldman - 2/9/13 at 4:27pm