About a year ago I was having a problem where after playing any graphics-intensive games for any moderate amount of time (about an hour or so), my computer would completely shut off with no warning. I figured it was the fan or the thermal paste going bad, so I ordered some Artic Silver 5, cleaned out the fans and applied it to the GPU. I didn't have any more issues for probably over a year.
About a month or so ago, I've started having the same issue again. I figured my previous fix would work, so I cleaned off the old thermal paste, applied some fresh Artic Silver 5, cleaned out the fans, etc. It appeared to have worked for about a week or so, and now I'm getting the random shut-offs again.
I downloaded a few temperature monitoring programs to see to what extent the GPU was overheating. Both PC Wizard and MSI Afterburner were unable to detect the GPU's temperatures, GPU-Z did work however. I am having difficulty understanding the measures because it is displaying three different temperatures, two of which which idle around 66°C (not great but it's a laptop GPU), and the third idles at 81°C (see picture below).
I attempted to test the temperatures in a game. In Dota2, I reached ~122°C on the "high" measure, and in Far Cry 3 I reached nearly 140°C, which seems exorbitantly high. Is it possible that whatever part of the chip is responsible for monitoring the temperature is bugged and showing higher than accurate temperatures? I ask because while my computer does get warm when gaming, it does not feel quite that hot. I also do not get any graphical artifacts, which I thought were a symptom of an overheating GPU.
I am running Windows 8 (not 8.1) and the GPU is a ATI Mobility Radeon 5870M (mobile card, not desktop). I can provide more information about my system if necessary.
About a month or so ago, I've started having the same issue again. I figured my previous fix would work, so I cleaned off the old thermal paste, applied some fresh Artic Silver 5, cleaned out the fans, etc. It appeared to have worked for about a week or so, and now I'm getting the random shut-offs again.
I downloaded a few temperature monitoring programs to see to what extent the GPU was overheating. Both PC Wizard and MSI Afterburner were unable to detect the GPU's temperatures, GPU-Z did work however. I am having difficulty understanding the measures because it is displaying three different temperatures, two of which which idle around 66°C (not great but it's a laptop GPU), and the third idles at 81°C (see picture below).

I attempted to test the temperatures in a game. In Dota2, I reached ~122°C on the "high" measure, and in Far Cry 3 I reached nearly 140°C, which seems exorbitantly high. Is it possible that whatever part of the chip is responsible for monitoring the temperature is bugged and showing higher than accurate temperatures? I ask because while my computer does get warm when gaming, it does not feel quite that hot. I also do not get any graphical artifacts, which I thought were a symptom of an overheating GPU.
I am running Windows 8 (not 8.1) and the GPU is a ATI Mobility Radeon 5870M (mobile card, not desktop). I can provide more information about my system if necessary.