Sammitch

I've been fiddling around with my system the last few days, trying to tweak some fan profiles to keep things cool, and while running a simple 10 minute OCCT one of my GPUs seemed to quit at about 85 degrees C. All seemed well, nothing seemed to be getting hot enough to be dangerous, but once GPU2 [the hotter of the 2] hit mid 80s the screen went this color and froze.

After the reboot GPU2 did not come back online, and rebooting a second time did not resurrect it. It was late, so I shut the machine down and hoped a night would cool it off and bring it back. Luckily this morning I booted up and it came back, but I'm quite worried that this happened at such a relatively low temperature. Looking around the internets the heat limit is rated around 100C, am I correct?

Part of the blame for the hotter temps I would put on the Asus M3N72-D having its lower PCI-E slot for SLI way too far down, leaving very little clearance for air circulation on the lower card. Not to mention obscuring the air channel further with the cables for the case ports and such. I've tried to keep the fan port clear, but I can only do so much.

The card is still well within its warranty period, I'm wondering if I should be considering an RMA because of this, or is this a normal thing to happen? I've also been tossing around the idea of watercooling these cards as they both tend to run hot even though I'm not pushing them as hard in normal operation as they will be in the future. Thoughts?

TIA

7hm

Get evga precision and crank that fan speed to 80%
http://downloads.guru3d.com/EVGA-Pre...load-2251.html

__________________

JKBenchmarks

You can use EVGA precision on BFG cards?

__________________

7hm

Yes you can use EVGA precision on ANY nvidia card aslong as its supported.

__________________

manolith

it will just bit a little loud =b

__________________
We Are Gr33n Machine

Sammitch

Well, I've got RivaTuner and have launcher tasks set to set the fans to 40/60/80/100, but I'm lazy so I had been tweaking the automatic fan profiles to ramp up the fan speeds more aggressively. That was actually the main reason I was running OCCT is to make sure that the fans would spin up properly when the cards are under load. I'm just troubled that one of my cards would conk out before hitting the nVidia specified temperature.