OC Issue -- Need suggestions/help

[Midian_Falls]

Hello Folks,

I seem to be having an issue with my once rock-solid OC on the system listed in my signature. Over the past month or so I've experienced a small number of system lockups while gaming (4-5 total). I generally get some pixel garbage on screen then the system locks up requiring a hard reset to correct the issue. The only change to this system during this time was my upgrading to a pair of 8800GT graphics cards which is where my initial suspicion was laid.

After a crash last night I rebooted the machine clean and ran Orthos blended test to stress both CPU and Memory. The test bombed out at 1 hour 5 minutes and this machine had been 8+ hours solid before. This is leading me to believe I’ve got a memory or CPU issue instead of the once suspected GPUs.

I brought the machine into work with me, downloaded Memtest86+ (V1.70) and ran the standard tests for 5 passes successfully, then solely test 5 for going on 15 runs now successfully. Memory is set at the recommended voltage (as per the OCZ site) and not Auto in the bios.

Temperatures have been slightly warmer than usual, but nothing way out of the ordinary. I could back off the OC some and see if the issue persists, but the system was solid, no errors, issues prior to this past month and I’m having a hard time believing it just won’t work anymore unless something is failing.

I’m starting to run out of ideas of where to look or what to try next so I’m looking for anything you guys might have to offer.

Thanks

Tentative Update:
Beginning to suspect a bum BFG 8800GT OC now. I did a moderate software OC on both cards (Factory 625/900, OC 675/953) via the Nvidia control panel, both cards passed 3dmark06, but one card pixelated (very similar to what I've seen previously) and BSOD'ed the computer during the Nvidia stress test. So it appears it could very well be one of the video cards at this point causing system instability, the other card takes the OC and stress testing in stride. I don't have the time left here at work to run Orthos long enough to see if that was causing the issue, but I'll give it a go when I get home.

Conclusion:
It *is* the video card, I just ran an 8 hour 30 minute Orthos session overnight with the questionable card removed from the system. It must be a thermal or manufacturing defect on the card that only presents itself when stressed hard enough causing system instability. Now I have to convince BFG that something is amiss with this card which should be...interesting.

Final Addendum:
I maybe in luck, the card pair was purchased Nov. 11, 2007 from Bestbuy and thus covered by their gift policy of returns and exchanges by Jan 31, 2008. Checking availability and such now.
Thanks to all those who read and pondered this, whether you replied or not.

 

[Oline61]

Have you tried backing off on the OC and then running tests?

 

[Midian_Falls]

Not yet, no. I've been running all sorts of other tests to try and isolate what maybe going on. I downloaded, installed and ran SiSandra Lite, but the burn in test caused no errors nor any lockups.

I've been running single video cards alternately trying to discern if it's a GPU issue using Nvidia's Control Panel to temp overclock both cards and push them thru 3d06 and the Nvidia Stability test. So far both appear to be working fine and I can't recreate the lockup.

What gets me is that nothing on this system changed except the GPUs and the drivers for said GPUs. Why all of a sudden am I randomly locking up?

ARGH!

But if I cannot figure out what's going on, I will back the OC down a bit to see what happens. It's just frustrating is all.

Thanks,

 

[Oline61]

If single video cards don't cause any crashes but dual video cards do then SLI may be at fault.