Strange issues with GTX 680m

Hi all,

I have been having some strange issue on my laptop, and I think it's related to my GPU. I am unsure whether this is a hardware fault or if it is a driver issue. Maybe someone can help me out...

For the last 2 months or so, my PC would freeze every once in a while, but I'm really not sure why. Strange part is, this freeze will only happen while my GPU is not active - in other words, only when my PC is idling. When I play games, the PC has never crashed or frozen...

I've recently noticed, thanks to the help of having two screens and use of GPU monitoring gadgets ( Most Wanted Windows Desktop Gadgets - Windows 8, Windows 7 and Windows Vista), that the GPU temperature suddenly "drops" to zero right before the freezing occurs.

I've also downloaded GPU-Z to see whether these readings are correct. But strangely, the monitor crashed before I can actually see any good reading. The program starts off by saying the temperature is around 40C and then suddenly the temp drops to zero and the whole program freezes...


Can anyone help me out with this issue? Or maybe point out what the issue actually is? I'm very confused as to what is going on..

Addition specs:
OS: Windows 8.1 (using addgadget sidebar to monitor GPU temp)
GPU: GTX 680m

(if you think I should put any more specific spec info, please ask)

 

Have you changed drivers recently? What driver are you on?

 

GPU is most likely dying. My GTX680m did the same for a week before dying off.

 

what laptop,what cpu,what hard drive hdd/ssd, what ram. this will all help.
have you ever overclocked. are you running on battery or mains. is it set to high performance or power saver.
all above will help others.

download nvidia inspector and hwinfo64 and core temps from my sig below. great monitoring tools.

 

Thanks for all the replies. I really appreciate the help on my issue

A bit of an update; I uninstalled the software I was using to monitor the GPU temperature (addgadgets) and have not had a crash since. However, GPU-Z crashes every time I try to read the temperature off it when I'm not in game; just says the GPU temperature is "0" (when I open GPU-Z when the GPU is working during a game, it has no problem but then crashes after the GPU is "deactivated")

I think this started happening since the the 320 drivers onwards. When it was on 314 drivers this never happened.
I've had this issue for around a few months now. I don't think it's dying because it still performs extremely well, does not overheat and this issue only occurs outside of playing games. But you could be right....although I hope you're not

Here are all the specs (and the info you asked for @MrDJ):

CPU: I7-3740QM @ 2.7Ghz (3.2Ghz boost)
GPU: GTX 680m, 2Gb
RAM: 8GB @ 1600Mhz (not sure on the model)
Hard Drive (OS and GPU drivers): Sandisk SDSSDH2256G 256Gb
Hard drive 2+3: Hitachi HTS727550A9E364 500Gb (with mSATA samsung SSD PM380 32Gb)

I run on battery when playing games (and...most of the time actually). I use balanced mode.

As for overclocking...I tried doing it to the GPU once or twice possibly using MSI afterburner, but when I only tried it after the GPU started having this problem (or at least, I'm pretty sure)

I've tried core-temp - it's a pretty good software. Nvidia inspector also seems to crash on me, not entirely sure why though.
I've already got hwinfo64 and I can't seem to see any GPU temperature being read at all?? It is meant to be for "GPU[#1]..." right? Because that line is blank

Do let me know if there's any more information you need to know.

Could a GPU flash be a potential for fix this error? And if so, how would I go about doing this?



Thanks Again,
RT

 

Keep in mind that newer drivers turn GPU off completely and by that i mean right after i quit to desktop from a game GPU (680m) turns off and MSI Afterburner starts showing crazy readings for me, but i did not crash
On 320.49 and earlier drivers GPU would only switch to lowest clocks but now it turns off so i believe that your GPU Temparature becoming '0' just before crashing means your PC crashes right after your GPU turns off.

I used to overclock with MSI Afterburner but it does not work anymore due to crazy readings it has with new drivers, now i overclock with NV Inspector

I would suggest turning off all monitoring programs and checking if it still crashes?
I only run MSI Afterburner when playing a game so that is the only time i require monitoring.

 

problem almost solved.
NEVER game on battery as its the easiest way to kill it and sometimes can also damage other hardware due to not having enough power in it.
its advised when gaming to run on high performance.
the battery is mainly for minimal tasks like browsing on the move. whenever on a desk its best if its plugged in. some say when its on a desk for long periods of time removing the battery will give it longer life.
also be cautious as oc'ing it can void your warranty if it is still covered depending on supplier.

download battery care from sig below to see if youve done any permanent damage to it.
my 16 month clevo still has 100% battery health. my almost 7 year old clevo has 6% wear level after almost 7 years.

 

@MrDJ
I must have been very tired when writing that post XD - I don't game on battery on it's own; it's always connected to the mains (otherwise games are virtually unplayable). But thanks for that advice - I'll remove the battery once in a while - should keep it going longer. I also try to drain the battery every once in a while completely (not by playing games though) and it's supposed to increase the battery life.
I downloaded battery care and everything seems to be fine. Thanks for the info

@Captmario
That probably explains quite a lot actually. After not monitoring the temperature I don't have any problems at all, apart from I can't see the temperature, but I guess I'll have to live with that. I'll give NV inspector a go; though last time I tried something with it it crashed almost every time: apparently I need to do some kind of flash to the GPU to actually enable any overclocking with NV inspector. Did you have to do any of that?

EDIT: Just tried out nvidia inspector - works like a charm and actually monitors my temps without issue! Thanks a lot guys to all - I've been having this problem for a while and I'm glad it's fixed

 

well it was 3.20am when you posted it so we will let you off

glad to hear its hopefully sorted.

 

The perks of being a University student XD