GPU idle temps at 70 C??

So here's what I have, purchased around July 2011.
Sager NP8150
i7-2630QM
Nvidia GTX 485M
8 GB RAM

I've used this computer for gaming at high settings for the past couple of years, but using CPUID, my GPU idle temps have always been around 50 C and can get upwards of 80 C when gaming.

Now, a couple days ago, I noticed everything started to slow down a little. Programs would take a bit longer to open, html links take a second longer to respond. It didn't bother me too much until I started up Battlefield 3 to play the new DLC, and suddenly I have 15 fps. I started to frantically clean up, running CCleaner, virus scans, clearing hd space...but nothing worked. Then I opened CPUID again, and what the heck: my GPU is idling at 70 C.

I remember from before having heating issues, and I know my computer slows down when the GPU overheats. I tried rolling back my Nvidia drivers, installing new ones, closing unnecessary processes (I don't have much running anyway), and nothing works. My laptop is always on a cooler stand, nothing blocking the back, in an air-conditioned room. I haven't downloaded any new programs recently and even tried removing all peripherals. Didn't have much luck on google either.

So yep. I'm stumped. I'm pretty sure 70 C is way too hot for an idle GPU temp. Anyone had this problem before?

 

Your heatsinks may be clogged with dust, it happens.

Blowing hem out with a can of compressed air should fix the problem.

Another possibility may be your GPU needing fresh thermal paste.

 

So my next step should be to open it up and clean out dust/apply thermal paste? I wanted to avoid this but I guess I'll give it a go after work.

 

Make sure your vents are not clogged first. Both for CPU and GPU.. Check your temps after that. If it's still running hot, might be a time to repaste. =)

 

As the others mentioned. But clean the vents before worrying about thermal paste. If cleaning it out lowers the temps call it good unless you're comfortable applying new thermal compound.

 

Thanks for the input guys. So I cleaned out my vents, for both the CPU and GPU, and my GPU is now idling at 52-54 C. But here's the catch: this is with a single screen. The 70 C idle that I listed before was when I had it hooked up to a second monitor, which I've never had problems with.

Now, when I plug in the second monitor, I can watch the GPU temp rapidly rise to 105 C before it shuts itself off. So if anything, it got worse...
Should I try replacing the thermal paste now? Why would cleaning out the dust make it worse?

 

Okay, so I just reapplied thermal paste to my GPU (Arctic Silver 5), and now it's idling at 65 C with two displays. This is just under what I started with, so still hotter than it should be. Tried to look for a system restore point but the only available one was today. Also used LatencyMon to try and find conflicting drivers, no problems there.

Any other ideas would be appreciated.

 

its normal that the 485m runs hotter with two displays attached rather than one. in total it has three different clock states: 2D idle, 2D active/low 3D and 3D active.
when hooking up to an external display the gpu cannot go below the 2D active state anymore, thus heating up more than with the regular 2D idle state before.

what u could do is use nvidia inspector to set the 2D active clocks to the same as the 2D idle clocks, that way ull see the same idle temps as before even when the external display is hooked up

Sent from my Galaxy Nexus using Tapatalk 2

 

I second what jaybee83 says about using Nvidia inspector I've used that with good success but it will make the second monitors graphics performance lessen. If you were going to use the second monitor for basic stuff like internet thats fine but if you game on it it will take a hit in performance.

Depending on how you applied the thermal compound that could be why your temps are higher then you'd like. Too little or too much will give you bad temps. You should use about a pea sized amount right in the middle then let the compression of the heatsink spread it for you. Also dont remove the heatsink once you've put it on.

Make sure your computer is also on a nice flat surface, never carpet or on the bed etc. The vents on the bottom need to be clear so they can circulate air properly.

 

Yep. Just be considerate of a "pea-size" application; people seem to miscalculate this, or they simply don't comprehend the size of a pea (as I always see immoderate deviation). Some recommend a BB-sized application.

See here for detailed info.

 

Really appreciate the suggestions guys. I did follow all those steps when I reapplied, so no worries there. I also always have my laptop on an angled cooler pad so there's plenty of room for intake.
I don't think I want to lose the performance with Nvidia inspector, but I may have to give it a look.

The temps seem to be settling down, it's idling at 64 C now so hopefully as the new thermal paste sets in, it'll go down a little more. It also stays below 80 C when playing Battlefield 3.
Thanks again for everyone's help, I was a little worried for a while there but it seems to be okay now.

 

Hi! What temp monitor are you using? =)

It might be that those temps are the last reading that your monitor got before the card shut off when not being used. I may be wrong in this but I hope I can help.

I have a 7970m. When I exit a game, the last temp which is sometimes in the 70's or 60's is the last registered temp. It will show that even though the card is not in use. I first thought that that was my idle temps. but when I physically checked the exhaust, it was not warm at all.

 

I'm using HWMonitor by CPUID, but I can see it constantly updating, so I'm not sure if that's a problem? Interesting suggestion though.

 

if it keeps updating then I think HWMonitor is reporting temps correctly.. If I were you i'd repaste till I get the desired idle temps.. but that's just me being obsessive.. =)