i personally agree and kind of disagree w/ RBJTech and think the clamp being on the gpu is a problem and may not possibly be an issue on the cpu heatsink and here is my reasoning:
- the gpu heatsink is not a structurally strong design. i am not a structural engineer but by just looking at it you could see it would be pretty easy to bend down at the areas of the attaching mechanism, the xclamp. and the xclamp does pull down pretty hard. i was surprised when i took mine off how much effort it was puttying on the heatsink.
- the cpu by design is stronger with the manner w/ the ears and the center area and the fact that it does work much better doesn't allow the extreme heat to be built up in it, it is just a much better cooler from a structure standpoint and a heatsink standpoint.
- my reason for thinking the way i do is that i have seen my friends heatsink and it was visually warped. not folded over but definitely warped in a way that be appropriate w/ the xclamps pulling on it at the attaching points. now this is where i feel a lot of the problems come in. the gpu, after time of being heated and cooled over and over becomes weaker with the constant stress of the xclamp on it. this in turn would pull the sides down and the area around the actual core would raise up, even a bit, then there would be no direct contact w/ the gpu core itself, and as we know, air is not that good of a thermal conductor and thus the temperature would rise drastically in the area between the heatsink and the core itself.. i don't feel like the thermal compound ms made the 360 w/ would allow for this movement and would "pop" off and unbind from the 2 units, and my feelings on this again were from when i took my 360 apart and out how the compound was and my heatsink did just that - the heatsink compound was not soft at all but hard.
- also in closing, i really hope that ms new design doesn't make the heatsink weaker and warp easier. i would also like to know if my friends gpu heatsink warping is an isolated incident. with him being a machinist by trade, the warp was very easy for him to see and also very easy for him to mill the heatsink down flat again, i just wonder how many other heatsinks are warped and have been fixed more by AS5 actually giving a good contact between the 2 units. unfortunately i know we will never get real numbers from ms....
my opinions are based of what i have seen w/ my own eyes, and i must reiterate that i have a great deal of respect for RBJTech as i believe he does us all a service by sharing his excellent tutorials with all of us. again, this is my opinion and i may wrong, but again just another opinion
The GPU heatsink is very strong structurally one way (cross fins) and weak inline with the fins. I do not believe the X clamps themselves have the force to bend the heatsink - however, things like thermal expansion can create huge pressures and easily bend things. So maybe because the central bit under the die gets so hot and the outer areas and fins don't - this creates differing expansion rates and the thing bends in extreme cases ...
Anyway - we are getting a little offtopic here ...
I believe the heatpipe will take aware the severe heat and solve the problem..
ps - respect mutual - I admire the work you're doing with the copper heatsink ..
Author
Topic: (Rumor) MS Adding 2nd GPU HeatSink for Better Cooling? (Read 1065 times)