April 2, 2009 2:59 PM
Hi Again,
The fact that you have a friend's dead TS to use for check and test is testament to the fact that there are broken machines lying around in people's garages and basements with less than 2 years of use on them. This happens with all manufacturers, and what do you want them to say, "These machines are junk, you should start a class action lawsuit against our company, and make some lawyers rich"? Hah!
No manufacturer will admit that their machines are defective even if they have unusually high failure rates.
With the onset of heat, you begin to have troubles. Maybe check the thermal material under the video card heatsink if so equipped, and also check the cooling fan at the radiator end of the heatpipe to be sure it is not dying or dead. Arctic Silver will help cool the video chipset, but some of the nvidia mxm cards have memory in an L shaped pattern around the chipset that is at a lower installed height than the video chip, and have to use a special 1.5mm thick phase change heat removal pad between the cooling plate of the heatpipe pad and the memory DIMMs themselves on the card. It's obvious that heat is your problem, it's just not yet certain that it can be cooled better due to thermal breakdown of the cooling system, or a malfunction of the memory or chipset. The heatpipe and fan are the first place I'd go, and if you can leave the back cover off to watch the fan operate, do it. The fan could be running down as it heats up from running, slowing to the point that it no longer cools well enough to allow reliable operation. Good Luck, I hope you can save it.
Dave