Update: I received an RMA'd GTX 460 and am still having the same problems. I'm going to RMA the AX750. The funny thing is, while I was waiting for the GTX 460, I managed to secure a GT 520 and I had no problems with it. However, the GT 520 hardly uses any power (no PCI-E connectors on the card) so I'm pretty sure that's a bad card to use as a comparative test.
I tried a slew of other things too - disabling DWM, clean installs of different versions of NVIDIA drivers, overvolting and undervolting the card, etc. etc. Heck, now I'm getting PINK screens with this card!
I really hope it's my PSU that's the culprit. Please let me know if anyone has any ideas.
This is my current build:
EVGA X58 (first generation) motherboard Intel i7 965 @ stock clocks 3x 2GB DDR3-1600 Corsair RAM at stock timings and voltages Corsair AX750 80 Plus Gold PSU 1 Optical Drive 1 Seagate 7200.10 500 GB drive 2x Western Digital Caviar Black 1 TB drives OCZ Vertex 1 60 GB EVGA GTX 460 Antec 1200 case HT-Omega Striker 7.1 Sound Card Windows 7 32-bit Professional (PAE Enabled)
My graphics card started artifacting while I was playing a game. It artifacted, the display blinked, then I got an NVIDIA driver has crashed and recovered message. Kept going, more artifacts, another crash, but this time my display blanked out and I couldn't do anything.
Restarted my computer - artifacting is in the BIOS - got to Windows 7 but it BSOD'd before I could even log in.
I restarted the computer again - artifacts cleared themselves out and I managed to get to Win7, but it soon started blinking in and out and artifacting again. Checked the card temps and they're well within range. (50 idle, 70 full load) The ambient temperature here is about 80-85F with high humidity.
Tried Safe Mode and it still froze up/BSOD'd.
Already tried the following to fix this problem:
- Reseat the graphics card and swapped in a different slot.
- Removed cover on card and sprayed with compressed air to clean it out.
- Swapped around memory and/or went with only using one stick at a time.
- Underclocked card
I called EVGA Tech Support and they said that the voltage on the 12V rail of my Corsair AX750 PSU was on the high end of the "acceptable" range (12.4V, highest within acceptable range is 12.6V - optimal is obviously 12V). They gave me an RMA number anyway, but I want to get a second opinion from you all before I send this thing off, as shipping from where I live to EVGA is kind of pricey. This PSU is only 6 months old.
So that I don't have to play RMA tag, which case would be most likely? I'm strapped for cash at the moment so I want to reduce the amount of RMAs I have to do since shipping is expensive here. Is there any surefire way to test to see if it's really the graphics card or the PSU?
I tried unplugging any devices that were connected to the 12V rail (except for my SSD and graphics card) as I do have 3 mechanical hard drives, but the voltage for that rail didn't drop (it remained at 12.4V).
I'm fairly sure it isn't the drivers since I'm getting the artifacting at the BIOS too. Right now I'm back in Windows 7 but I don't know for how long until it messes up again.
Any ideas? Any way to try to buy time for this card without voiding the warranty (I am not throwing this thing in the oven).