Originally Posted by ramman949
I have almost the exact same ram. But mine is CMX... Not CMZ. Mine has to be set to 1.65 to run. At any rate, I have the EXACT same problem man. BF3 freezes at random on me too. It can take 10 minutes or 2 hours. Or I can go all night with no issues. I have the P8P67 PRO version 1 with a I7 2600k, @ 4.5 all the time. My ram is at 1600 @ 1.65 and my video card is the asus GTX 570.
It seems that my card has issues with being under volted from what I have found on the internet. I used MSI after burner to bump my video cards voltage up some and it seems to have stopped. Others have told me this is a classic sign of a weak PSU. However my PSU has AMP rating of 53 and my card need 38 so Im not so sure its my PSU.
I know Bf3 does NOT like overclocking, so I might run my pc at stock speeds to see if that helps too. But one thing is for sure. You are not alone. It happens to a lot of us. It could very well be a buggy game. I mean, if it didnt before and now it is, either something on your end changed, or the update to the game is causing it.
Im starting to think its BF3 since all my other game run fine.
I had tried so many things, it didn't seem like upping the GPU voltage had helped before - but it may have been a combination of multiple issues, so I might need to try again.
I knew about the GPU voltage typically needing to be bumped up with some cards for DX11 games, particular a system-stressing game like BF3. GPU BIOS updates were issues specifically because of BF3.
I had the voltage set to 1.050V in Afterburner (running a 950/1900/2004 clock), and that was just peachy for what, seven or eight months? Then bam, no fun for you!
I think I had tried that same voltage with reference/stock clocks (822/1645) somewhat recently and Campaign still locked up - but I think I need to try it again in MP.
Currently Afterburner is uninstalled, as I was investigating if any software/processes/drivers were conflicting and causing this issue - but I'd say I basically ruled out most software, still trying to determine if any drivers are responsible but it would be quite swell if it was fixed by a specific voltage needing raised.
I'm kind of bummed that I went through uninstalling and cleaning out all the Realtek audio drivers and going through a headache getting a good driver install for my old PCI X-Fi Fatal1ty card (sitting unused in my last build, I think that's the one still in my forum sig btw
), when it appears my specific hard-lock was NOT related to my onboard Realtek audio (which I had played with just all this while).
I have read that the recent updates to BF3 changed how some data is streamed and whatnot, so I have been curious if a specific voltage like VCCSA or VCCIO needs bumped up, especially since I am running with a SLI config. But I had been testing VCCIO at 1.15V or so, and it doesn't seem right that it would need to be any higher since I've seen many run stable at Auto, 1.05 or 1.10V.
Would 1.20V be too much? And I've read changing the VCCSA voltage isn't a good idea... am I misinformed, and does that need to be raised up?
While I don't plan to make my 2600K last for a decade, I don't want to stress it out to the point it dies in two years either. That, and I never needed to even think about touching those voltages before to remain stable, so I'm leery of raising voltages that don't seem necessary.