6970m crossfire issue.......P180HM

OK, got an issue that is becoming more and more annoying. Regardless if I install the beta 12.1 driver, or the 11.10 or the original 11.6 driver, i get the CCC installed, it detects both cards, boots up and works just fine. Now every once in awhile regardless of what game I am currently playing the system locks up, I get the typical BSOD screen, and then a reboot. The problem is when the system reboots it only detects one card now. No matter what I do, driver or catalyst uninstall etc, with a few reboots even, it still only shows one card under the display drivers.

THe only way I can get 2 to show back up is to do a fresh install of Windows 7 x64. W T F is goin on here? Suggestions and help please.

 

It sounds like you might possibly have a faulty card. Driver issues could cause the BSOD and reboot, but a card failing to show up or be detected seems to indicate a failing card and/or motherboard slot. However, it's strange that it suddenly starts working again with a fresh Windows install.

 

Hmmmm, well I just updated the bios to 1.01.03 and the EC to the1.00.05 and restarted and now the second card has shown back up along with the missing controls in CCC to enable crossfire................any thoughts?

 

Hmmm, jusr ran the latest update of FurMark in crossfire mode...........when one of these video cards gets up to temp(one runs about 5 degrees hotter celcius?)the screen freezes, then a BSOD appears. And it seems to crash when the temp hits right around 90 degrees Celcius? Upon immediete reboot again only one video card shows up in device driver. But if I shut down the laptop and wait a half hour to let the system cool and then boot back up, both cards are being detected it seems. Think its a garenteed defective video card or? Opinions and recommendations please.

 

Bung card for sure, send it back to where you got your machine from so they can replace it?

 

Well here is the thing, Im gonna run FURMARK some more, I know at least one card is fine running on its own. So Im gonna swap the card positions to make sure the problem follows. I assume one of the MXM slots on the P180HM is the primary display slot? Or does the system just autodetect whichever one is occupied?

 

Yeah, decided to test these cards individually. The original(one Master slot installed 6970m)main display card can run FURMARK all day long. But when I swap in the second 6970m card into the "master" MXM slot and run it solo on FURMARK, the screen goes black(no BSOD, but I think thats because the video card outright fails?)after about 7-8 minutes and when it hits the 90c temp mark. When I shut down the computer though manually and then restart, it does boot back up though immedietly(Doesnt need really anytime to cool down). But it just cant seem to handle FURMARK for longer than 5-10 minutes. THoughts? Definetly card and not motherboard right?

 

Sounds like card issue if one card works fine in the same slot and the other doesn't.

Not having a dual GPu system do both GPU's run the same vbios or does one have to run slightly different due to it being secondary card? (sorry for stupid question).

 

both have same vbios revision.

 

Yeah, it sounds like the card is shot, as I'd thought previously. Testing them both in the same slot seems to confirm that as you continue to have pretty serious issues, but with only the one card. I'd see if you could get that one card replaced under warranty. Considering it's almost certainly a card issue, you may even be able to just get a replacement card sent to you to swap out without having to mail your whole machine in (though it depends on your retailer).

 

It was bought through Eurocom.com through Ebay(And I didnt leave feedback yet, hehe) originally about 2 weeks back. Recieved it last week. Was kind of doing the "there is no way it could have been sent to you defective because we are ISO 9001 certified". BUt after a few more emails I think they have decided to actually believe that what Im saying is spot on. I did do alot of testing on various fresh window 7 reinstalls, had a few extra HD's on hand, even extra 4gig memory sticks(they were trying to blame a potential faulty HD or even memory, but I ran memtest86 v 4.0A and everything was a clean error free pass).

I just dont see any reason why it could be anything else other than a faulty 6970m card. If one can pass and run FURMARK all day long and run in the 95c range without a single hiccup, then the other one should be able to as well.