Actual application of the paste (amount & consistency spread over the GPU) are probably more important than the TIM used. Some of my GPU's have owned had shockingly bad retail paste applications. I think that it was my GTX580 where only 3/4 of core had paste on it, the remainder cooled by bare metal. I couldn't understand why it ran hotter than my 480 before I popped the cooler off. At least the paste on the above card seems to have been spread pretty evenly.
Last edited: