Loved all the Fujipoly talk today! I used some of that when modding a generic waterblock to fit my Areca raid card.
But all the talk today made me use my mad google skills to better understand the watts per meter kelvin (W/(m-K)) rating that is used on all the TIMs, including Fujipoly. I wanted to understand what it really meant in the real world. My understanding has increased. I'm going to take what I learned and dummy it down so that even I could understand it.
The (W/(m-K)) rating is the thermal conductivity rating. Important, yes....but not the whole story. What you should really ultimately care about is the rate of heat transfer. The formula for the rate of heat transfer is:
Rate of Heat Transfer = Thermal Conductivity * Heat Transfer Area * Temperature Delta between Two Surfaces / Thickness of TIM interface
or...to simplify, I'll call it: RHT = (TC*A*Delta) / Thickness
So, you can see that if you want to maximize your rate of heat transfer RHT (and you do), you can really do it with 3 of the 4 variables in the equation. For simplicity sake, lets just assume that you can't change or control the Heat Transfer Area, or A, in the equation. Imagine that as the size of the IHS touching the bottom of your block. It is what it is. You can't do much to change that (outside of extreme modding).
But, you can definitely impact the other 3 variables. For the TC....just buy the thermal interface material with the highest thermal conductivity you can. Heat transfer will go up. For the Delta....just maximize your radiator space and spin those fans quicker. That will increase the delta between the IHS and the water block....which will increase heat transfer. For the thickness variable....you want as thin a layer of interface material as you can possibly have, while still eliminating or minimizing air between the two surfaces.
I think the lightbulb moment for me was understanding how important the thickness variable was in the heat transfer equation. Some of the Fujipoly pads had incredibly high thermal conductivity ratings. Much higher than you get with a good thermal grease. As an example, let's say you can use a thermal grease that has a thermal conductivity rating of 8 W/(m-k), or a Fujipoly pad that has a thermal conductivity rating of 17 W/(m-K). Which will transfer more heat? It's going to be the thermal grease almost every time...because the thickness of even the thinnest thermal pad is going to be many times the thickness of a properly applied layer of thermal grease.
Of course...that's with everything else being equal. There will be times when you don't have good direct contact between two surfaces...which will necessitate using a thicker thermal pad like Fujipoly. But if you can spread grease and make good contact....choose that option every time. Every time.
If you have to use thermal pads....use the thinnest you can while still making good contact. Fujipoly has some relatively inexpensive pads with a thermal conductivity rating of 6 W/(m-K)....and some super full-retard expensive pads with a thermal conductivity rating of 17 W/(m-K). Guess what. The 0.5mm version of those cheap pads has about the same rate of heat transfer as the 1.5mm version of the full-retard expensive pads. The cheaper pads only have 1/3 the thermal conductivity....but if you use one that is only 1/3 as thick, .....heat transfer is about the same.
I found it all kind of interesting.