Help! P150EM 680m throttling at stock clocks?

EDIT: Problem solved, see post #11

Hi!
Looking for a bit of advice. I bought a P150EM a few days ago with the 680m GPU but I'm having issues with it hitting high temperatures and throttling back during games.
I cleaned and reapplied diamond thermal compound twice yesterday; both times I put a small half-pea size blob on the GPU and let it spread when applying the heatsink. Both times I lifted off the heatsink to check I had the correct amount for total but thin coverage, then cleaned and reapplied the same amount.
I know the compound is a good batch as I used the same tube to paste the CPU and also the CPU/GPUs in my desktop PC.

Booting in to windows and launching MSI afterburner, I see the GPU idles around 42-44 degrees C, which is already much better than the factory paste. I fire up far cry 3 and watch the temperatures climb. The fans speed up as expected and there is a stream of hot air from the GPU exhaust point which seems normal. Within about 3-4 minutes the temperature is at 89 degrees and the GPU core clock starts throttling back (and of course the frame rate drops with it). From then on until I close the game, the core clock just constantly cycles up and down as it attempts to keep the temperature at 89 degrees. This is all at stock clock speeds, no overclocking/overvolting etc.

I've checked the fit of the heatsink, there are no blockages/dust etc as this is a brand new machine. I've also re-seated the GPU memory thermal pads just in case some trapped heat from there was causing an issue.
The GPU heatsink must be making good contact as the air being expelled from the vent is hot, I'd be more concerned if it wasn't

Appreciate any advice, or any details of what temperatures your 680m's run at idle and under load.

Thanks!

EDIT: I forgot to mention the ambient room temperature is 20 degrees.

 

Which driver and VBIOS are you running?

310.70 and 80.04.67.00.01 are the latest.

 

Sorry forgot to mention that! I've run just about every driver from the stock clevo 302.77 up to 310.70. My vBIOS according to nvidia inspector is 80.04.29.00.01, is it worth upgrading this to see if it makes a difference?

 

dont think im getting as high as you in far cry 3 but i am in other games where cpu can hit 90c but i think gpu is mid 80c. will check when i get a chance.

could you run intel extreme tuning utility which you should have pre loaded (search in start menu)

this shows my card throttling quite a bit as well.

 

Yeah it's worth it - just PMd you.

 

Thank you both for your help. I'll try updating the vBIOS and see if that solves the issue.
MrDJ - my CPU isn't throttling at all, just the GPU. Will XTU show the GPU throttling state too? I'm currently just going by MSI afterburner readouts, showing the core reaching 89 degrees and throttling the clock speed down.

 

I'd probably try another repaste too.

 

it has the option to pick either cpu or gpu to run the test on but as my gpu was fine i didnt grab a screenshot.

just played far cry 3for about an hour and temps are fine for me with no downclocking. i was mainly getting the problem with B3 and MOH Warfighter which could be something to do with the crud origin platform.

 

All 680M bioses will throttle at 90C (reads as 89 or 90) though I am surprised this is happening at stock clocks/voltage. You may need to check the pasting job on it.

 

already been repasted by mysn on both cpu and gpu. i was getting upto 96c on 3 cores while running extreme tuning with 34% throttle and that was about a month after i got it.

 

Problem solved. Long story short, my heatsink is (was) completely concave. The corners sat MUCH lower than the area that made contact with the gpu. It wasn't obvious to the naked eye and as the bottom of the heatink has screw tags attached its not possible to stand it on a flat surface to see the problem either.
I've done my best to straighten it out, and now after 1 hour of Far Cry 3 my core temperature is 68 degrees.
Terrible quality control from Clevo, this should never have been sent out of the door in such bad shape!! I may contact the place I bought it (pwnpcs, who have been brilliant by the way) to see if I can get a replacement heatsink.

Thank you again to everyone who replied to the thread and made suggestions

Now to investigate a 'real' fix for the horrendous audio output quality, starting with those opamps......

 

glad to hear you got it sorted.
the speakers are a bit pants but with my 5.1 surround plugged into phones/mic/spdif i get great sound.

 

That or get a USB sound card

It's maybe a bit safer than resoldering, but if that's the way you want to go....

In previous posts though the amp used might be built into a chip.

 

ASUS Xonar U3 will fix all your audio woes - awesome little device.