Hello, long time reader on here but this is the first time i have posted. I'v got 2 sapphire 6970's in crossfire and they are about 9 months old. few months after I got them I put a ARCTIC COOLING Accelero XTREME Plus II on each one, they stay pretty cool, in crossfire the top card gets about 72x and the bottom 60c for a long time, about 5 months or so I'v ran a mild OC core was 910/880 and mem 1400/1375. when the cards were new they could do this stock but then I got BF3 and i had to bump the voltage just a bit, stock v 1174 and i bumped to 1181, very minimal bump but it worked. about a month ago they just started freezing in any gpu intensive situation, now they wont even run at stock, i run 860core 1350mem and i have to turn the volts up to 1189, even then if everything hits just right in BF3 it locks up, i can't believe my cards would degrade like this over 9 months with the temps always being so low and the OC i ran was so mild, but i guess they have, thought i would run it by you guys and see what you think
forgot to add, each card works just fine by itself, but in crossfire i have to have them in the order they are in, if i move the bottom card to the top and the top to the bottom and run crossfire, they are really unstable, its odd they both run fine by their self which is why im not 100% sure they r going bad, but everything has been fine for 9 months, also I have tried different drivers all the way back to 11.11
honestly i havent OCed them by their selves, ill try that tomorrow, the psu is also 9 months old, 1000w bronze 80, all the parts were from a new build. the thought about the psu had crossed my mind, not really sure how to test it, i realy dont wanna pul all my wiring out, i do have another 1000w but its in another computer and id hate to tear apart 2 computers just to test one
what brand is your PSU? 1000w sounds solid, but most quality brands only make 1000w PSUs at 80+ gold or at least 80+ silver. some of the less reliable PSU manufacturers test their PSU outputs at room temperature, rather than at 50C or so (the temperature the PSU will be running at). so I still see it as a potential problem...
I understand the wiring issue, but an easier test would be to try running crossfire in your other rig. I understand there may be potential driver conflict issues etc, but I'd try that before pulling out all your wires.
also, a slightly easier test, especially if you have a spare harddrive laying around: try booting your PC from a fresh install of windows. just install the bare minimum and some benchmark software (including BF3). and test a few drivers, especially the newest beta drivers fantastik suggested.
EDIT: Just had another thought, have you tried thoroughly cleaning your two cards? 9 month is certainly enough for dust build-up. your temps suggest your cards are fine, but it's always a healthy thing to do, and dust accumulation caused heat issues would certainly limit your overclocking
just as an update i finished testing one card, OCs like a dream by itself and stays perfectly stable and cool, on to testing card number 2
@vmem this is a fresh install of windows 7, like yesterday fresh. cause of the arctic cooling fans they wont fit in my other case . but if i have to i think i can set them next to each other and just run the wires from the other case but still gotta test the second card first, also there is virtually no dust on them, got a hafx case with all kinds of dust filtering
EDIT: just got my answer, the second card is degrading, it had been a month or so since i ran each on their own but never OCed on their own, soon as i loaded a bf3 map up with OCed settings on the second card it crashed, i always get "display driver stopped working" set to stock settings and crashed again a few min in, this card has gone down hill since i last tested it by itself a month ago, now im stuck wondering, do i get another 6970 to replace that one? do i go ahead and get a 7970 and sell my good 6970? decisions decisions, thanx for the tips guys