Dangerous temperatures

I've tried looking up and down this forum and elsewhere for a clear answer, and haven't found it yet.

I've been getting random crashes where the screen washes out with multi-colored lines and have to cold-reboot (and usually have to wait about 10 minutes before trying to boot again or it goes right back to that).

I've been running my m11x R3's gpu at the magic 800/1000 mark, and I suspect (duh) that the OC is what is causing it. I've run Furmark using that OC and seen the temps go up to 95C, which to me is downright horrifying.

I'm running it at 700/900 right now, and Furmark still has it at 91C. If I drop it to stock speeds, Furmark has it at 90C.

I'm inclined to just blame the fuzzy Eye of Sauron, but I'm curious what you all might think. Are these temps normal, or is it possible I got a bad GPU?

My laptop is too new, and I'm not willing to void the warranty at this point by taking it apart to replace the thermal paste.

EDIT: FWIW, running Unigine at the 700/900 speeds, I see the temp holding around 73C. My concern is that certain games will tax the GPU more than Unigine will (more approaching the kind of GPU intensity that Furmark dishes out).

 

You won't void your warranty if you take it apart and change the thermal paste as long as you don't break anything in the process, even in this case warranty will still be active but won't cover your famage unless you have accidental damage plan.
If you don't feel comfortable doing it, call Dell and they'll most likely send a local tech to do it, I'm just not sure of the quality of his job :/

 

95 degrees Celsius is very high. It must be a losuy paste job on Dell's part or something. If it is getting to the point where your comp is crashing, I would definitely do something about it. However, I wouldn't put too much faith in Dell technicians.

I would just repaste and see if that is the problem. Some good pastes include the classic arctic silver, mx4, and noctua. They're all around the same performance, give or take a degree or so.

 

I have heard conflicting reports on what voids the warranty. Some have said that putting on new thermal paste does void it, others say it does not. I suggest calling dell first and finding out for sure before you do anything. And yes 90 or 95c is way too hot. you should not see above 80c max. And temps should range from 50c or so at idle, and the low 70's under load.

I should also mention, alienware/dell does not put the thermal paste on themselves. The mainboard and the heatsink are assembled in china and sold to dell/alienware as one complete piece.

 

The best approach i would see is get a tube of quality thermal paste and call Dell. When the tech pops at your door ask him to use your thermal paste and not his.
Else, you can do it by yourself if you are confident that you won't break anything in the process and without asking yourself where those left-over screws are from at the end of the operation

 

lol i still have 2 screws left over from the first time i took my laptop apart, I have taken it apart 3 times since then and STILL cant figure out where they came from

 

took mine apart 3 times and all the screws found the rightful holes... Most of them are even marked bro.

The R3 (not sure about R1/R2 but i suppose heatsink is pretty much same) clogs with dust REALLY quickly... changed my paste not even a month ago and i already see a bit of a furry dust thing covering it from the inside... (just use a torch from below it and look from behind... can see the dust clearly that way) if theres no dust blocking it you painfully need a re-paste.

Mine was similar when it arrived, would hit mid 80's temps and with a bit dust there it gone all the way to 90++... Thermal paste was applied very poorly with air bubbles etc. so no wonder it didn't do crap... From the looks of it they just take a dense patch of paste stick it onto the copper bit of the heatsink and screw it on... which doesn't exactly transfer the heat very much better than air...