Interesting 970m issue

Hey!

So, recently bought an i7 4710QM/970m Sager. The first night I had with it (in a rush to run every game in my Steam library) I noticed that pretty much any demanding game I launched either crashed immediately or had severe visual artifacts when playing. Got plenty of BSODs before adjusting the driver timeout in the registry - now it just crashes now and then. Anyhow, spent the next day trying different combinations of driver configs and making incantations to various gods - but to no avail! I feared for the worst...but then I read a post online about a completely different card going too fast for its own good.

After that hastily downloaded MSI Afterburner and put the core clock Mhz slider down as far as I could go (an underclock of ~135). And what do you know? It worked...but that was on battery. As soon as I plug the machine in and try to start, say, Skyrim, the vicious artifacts and freezes are back. It can't be Optimus, since the 4600 HD couldn't even handle the applications at these settings. Noteworthy is that before the underclock the same thing happened...I can only assume, then, that this is the PowerPlay function kicking in and throttling the already slightly neutered 970m, especially since I never go above or drop below 30 when on battery regardless of settings.

I'm wondering what my options are here. Is it possible to emulate Powerplay's functionality through a more potent overclocking utility that goes further down in clock speed? Is there any way to just force PowerPlay all the time regardless of being plugged in or not?

Thanks. I think this is the right section.

 

Artifacting + BSOD = dying card (experienced it 3 times). *RMA ASAP*

 

which driver have you got loaded. there is a specific 9** only driver i posted a few weeks ago here or on the clevo forum.

artifacts can also be caused when optimus doesnt auto switch and it uses the intel graphics which has affected a few of the latest games.
check each game exe in the nvidia control panel and make sure dedicated card is highlighted.

playing about with the timeout in the registry is not a good idea.

also download who crashed and some temperature software from my link below. post the dump reports from who crashed here and run something like core temps, hwinfo64 and gpu-z or nvidia inspector in the background and post your results.

ive seen 2 posts on new 8** cards last year that turned out to be no thermal paste on either the cpu and gpu and they were getting almost exactly the same problems as you are.

 

Thats why I shy away from resellers that don't offer repasting of thermal paste or say that the manufacturers does a good job and there is no need to repaste but if you want, we can do it for $89+ (such a rip off).