Sign in with
Sign up | Sign in
Your question

overheating sli issue

Last response: in Graphics & Displays
Share
June 4, 2008 4:19:23 AM

ok i've tried a few things before posting but here's the issue.

EXPLINATION
i recently installed a second 8800gt from inno3d, (both are the overclocked version)

i have an antec 900 case and a evga 750i mb

whats happened with this setup is the antec 900's psu bay is at the bottom of the case instead of the top (you know just to be different)

and EVGA have done a fantastic job in spreading the 2 pcie slots as far away as practical.

this means that the space inbetween my cards is fine but the bottom card is only 1 maybe 2 cm from the psu.

this causes my bottom card to be approximately 10c warmer then the top card, after about 30min playing cod4 the driver fails, check the temps card 1 is at 60 and card 2 is at 65-70.

What ive tried to fix it:


swapped cards around to see if one cards heatsink did a better job [fail]
installed the extra fan in the side cover [fail] maybe a 5c drop if that

what do you think about these possible solutions
possible Solutions:
does anyone know of a good 2slot 8800gt fan i can buy to vent the air out the back.

case modding to move the psu somewhere else?

New case like a thermaltek armor


until i can solve it im going to have to unplug a card

More about : overheating sli issue

June 4, 2008 4:45:56 AM

i could be really tired and sleepy but isn't that supposed to be jus the opposite effect of having he psu at the bottom? in my cm 690 case, where psu is at the bottom, the ocz psu sucks the air inside and blows it outside, providing an extra fan in my system. so shouldn't this extra cooling be cooling the card, instead of heating? i have the ocz gamextreme 700watts psu, which one do you have?
June 4, 2008 4:56:10 AM

silverstone 650w thier is no draw in fan in it only the rear exhaust fan
Related resources
June 4, 2008 4:57:32 AM

Your vid cards are rated at ove 90c before issues happen... look elswhere for the problem.

A second gpu may overwhelm your psu (12v rail) and cause the problem.

What psu, cpu, and ram are you using?

The other culprit is ram voltage, needs to be at least 2.05 for that board.
June 5, 2008 1:12:28 AM

i've tried using GPU1, by itself then GPU2 by itself same issues, so its not insuffiecient power, its not heat, i tried using the 169 nvidia drivers, and the 175 with each card.

i've upped the ddr2 voltage to 2.05 testing now

EDIT:: i think it might be the PCIE bus speedn Currently set to 3150, supposed to be 3000, cant believe i missed that, i think it was changed last time i played with nvidia's auto system tuner.

not very useful btw
a b Î Nvidia
June 5, 2008 12:43:12 PM

Yeah I agree, I don't think Temps are your issue here. I would look at a power issue or a defective card perhaps.
June 5, 2008 11:39:29 PM

ok i found out whats going on not sure why yet,

temps ARE an issue but only when i plug my cards in a certain order, thus thats fixed. (when incorrect one card hit 90c in 10 mins whilst installing crysis. the other was 50c)

during load the nvidia ntune service for some reason oc's my pcie speed from 100mhz to 120mhz and thus the cards are unstable. does anyone know why it does this, i have never told it to, i unintsalled and reinstalled the ntune program to wipe any config data it had telling it too do that but didnt work.

!