The mass of the heatsink is not as significant in notebooks compared to high end coolers on desktops. Because the amount of leverage on a full size tower heatsink is much greater, it is susceptible to motion and shock and can warp the motherboard, rip off of the mount, and potentially damage other components, etc. This is why enthusiasts in that field do not recommend leaving the heatsink on when transporting the tower to a LAN or other destination. Most desktop rigs are vertical setups so the heatsink is perpendicular to gravity. This is a worse scenario compared to our little "upside down" heatsinks since the center of gravity is much farther away from the CPU and the mount. Of course, high tension from the brackets make up for that. But the potential for loss in cooling performance, if not damage, is still greater in desktops than in notebooks if they are transported with the heatsink. I've worked on several notebooks, including the latest Clevo notebooks, from a diverse range of clients all around my local region and each notebook that I have opened has a different story to tell. In none of them have I found any evidence of the thermal paste being rendered less effective from gravity pulling down on the heatsinks with or without the assist of force from transport. Perhaps if a notebook is dropped on a hard surface with enough force then that would reduce the pressure between the component and the heatsink enough to negatively affect the current mount of the paste, but then again I would be more concerned about other damage at that point.
The thermal paste already acts as a form of adhesive between the die and the sink to prevent the two from coming apart during motion. This is what the GPU heatsink on the W8x0CU primarily uses, and with the addition of the adhesive properties of the thermal pads. The bolts on the CPU heatsink are just a bonus.
Between the two extremes of high and low viscosity, the result will still be the same when spread over the whole die. Excess will come out of the sides. As to how much comes out, or if any at all, is to be figured out by trial and error over time. Though with higher viscosity pastes for notebook application, spreading is a must while low viscosity pastes can do with a dot/line method just fine. However, in the end it is all about technique.
http://forum.notebookreview.com/6709478-post173.html
For example, lets take the most extreme case, my modded W870CU. My CPU and GPU heatsinks are the heaviest on this forum weighing almost double if not more than that of a factory W870CU CPU and GPU heatsink. The amount of mass on them is almost to say clumsy and unorthodox as they are pulling against their respective components and mounting screws. Yet, to this day I have still to encounter evidence of the thermal paste being shifted, air bubbles forming, or any kind of anomaly resulting from inadequate or inconsistent pressure from the heatsink that would result in less than optimal cooling performance. The same goes for my old M570TU and that notebook had a boatload of mass added to the existing heatsinks as well. Both of my notebooks were transported frequently on the notorious San Francisco Muni, my occasional sprints to class, in the boot of my C6, and in other medians of transportation that have put my notebooks through the rollercoaster and back.
Problems? See this thread at archive.org.
