Driver Crashes

So this has been going on for a long time, probably 3 months, but I've just been sitting on it hoping driver updates would take care of it. Unfortunately they haven't and I've been starting to worry it's a hardware problem. Wanted to get some other opinions.

While doing anything other than fullscreen gaming, my computer runs without any hitches, no viruses, malware, or really any sort of issue. Fast and reliable. However once I boot up a game, I have to worry that at some point it WILL crash on me.

My symptoms are:
Black screen
I hear the card's fan lose power then rev back up
Screen restores and I have a little nvidia bubble saying something like "Video drivers have crashed and recovered".
Depending on the game, it either breaks the executable or comes right back to where it was. (League of Legends: good. Shadow of Mordor: crash)

Driver v344.11, no OC, used EVGA Precision X to set a fairly high fan curve. Rest of rig is just as the sig says. The power loss makes me think it's a motherboard issue, with the pci-e lane losing power. Does this make pretty good sense? Any of troubleshooting I can try? I've found nothing on Google.

Thanks in advance

You said you have an aggressive fan profile, but what are your temps at on load for the GPU? CPU?
You also mention a driver, 344.11, did you try others?
What power supply do you have?
Do you have another GPU to test with?
I doubt its the mobo PCIe slot... you can test by switching slots though...
Have you tried reseating the card in the slot? The power plugs?

You said you have an aggressive fan profile, but what are your temps at on load for the GPU? CPU?I was just playing Alien: Isolation and when I backed out I was around 60C gpu, which I know is well within reasonable limits. As for CPU, I'm not currently running an OC (although I safely ran +.6GHz) with H2O, so I really don't think that's an issue.

You also mention a driver, 344.11, did you try others?
I've tried every driver released over the last few months as well as some older, more reliable ones (314.xx).

What power supply do you have?
XFX 850 Pro I think. Definitely XFX 850W

Do you have another GPU to test with?
Unfortunately no

Have you tried reseating the card in the slot? The power plugs?
Yeah, I completely rebuilt the build in a new case in August. Problem existed before and after.

Did you try switching PCIe slots?

Tried underclocking the GPU or giving it a bit more voltage?

PCI slots yes, I've tried a couple. Same issues. As for underclocking, no I haven't but I will make that my next test. I have to go to class soon, so I'll drop the clocks and run Heaven while I'm gone. I'll keep GPU-z running as well to keep track of maxes.

The crash error sounds similar to those I've experienced with my card overclocked while gaming. I'd try ATM's suggestion. Is it possible that Precision X is undervolting it? Do you get the crashes when PrecisionX isn't running?

I guess I've never noticed, as I have it start at log-in and never really mess with it all that much. I'll get back to this once I've burned it for a couple hours, as for now I'm off. Thanks already guys

Yeah, I'd also test with PrecisionX out of the picture.
Will probably have to remove it from startup, rather than just exit.

Edit: Also, what version of PrecisionX?

Zephos said:

I hear the card's fan lose power then rev back up

Click to expand...

One other thing to point out. The fans likely aren't losing power. I think your power supply is fine. More likely the GPU core temp is just suddenly dropping (when the game crashes, the GPU doesn't have to render anything), so the fans spin down quickly. This will be even more noticeable in your case because you have the fans spinning higher than normal with your profile.

Let us know how it goes and have fun at class

ninjacore said:

One other thing to point out. The fans likely aren't losing power. I think your power supply is fine. More likely the GPU core temp is just suddenly dropping (when the game crashes, the GPU doesn't have to render anything), so the fans spin down quickly. This will be even more noticeable in your case because you have the fans spinning higher than normal with your profile.

Let us know how it goes and have fun at class

Click to expand...

Gpu-z told me that at default clocks and voltage I maxed at 87C, 108%TDP, and 1.212V after roughly 2.5hrs of Heaven. The overage in TDP seems a little bothersome, further prompting me that I need to drop the voltage. That's what I'll do for the rest of the night as I play games and stuff.

I just remembered that I think I may have flashed a custom BIOS to the card because the higher range of fan speeds was actually disabled by default. That's all it changed, but could that custom BIOS be the cause of this issue? I didn't notice it right after doing that though.

My PrecisionX I just updated this morning so it's 5.2.3.

Also here's my gpu-z status just if it helps

Yes, a custom BIOS could absolutely be the issue.

Alright then if undervolting doesn't solve my issues I will flash the original bios I found on TechPowerUp just now. Idk why that never crossed my mind, I guess because the problem didn't crop up right away.

UnderCLOCK or overVOLT.

oh duh *facepalm*

That was just a slip, I'll work on that now. Thanks everybody, I'll be back if nothing works out.

Be sure to update us with what does work as well.

Will do, thanks again

A slight overvolt seems to have corrected my issues, as by now I'm sure I would've had at least a few crashes. If it comes back I'll be sure to try here again. I always get the best results