Bitcoin Forum

Hi,
I have three cards running at 100% fan speed at the moment and the temps are 68 max. I still haven't overclocked them.

Any info greatly appreciated - thanks.

You don't really have to run them 100% all the time. Just keep temps below 80C or so. (6990's run hot... Seems like you're generally not hurting anything till you go above 90C)

Yea I have 2 x 6950's and have set fan speeds using Afterburner. One seems to run at a higher temp than the other so atm 1 runs @ 61c @ 71% fan speed and the other @ 71c @ 91% fan speed. I could lower the temps further but then you have the noise issue to contend with. I think this setup of not quite full fan speed and < 71c temp to be a happy medium. At these settings my cards arent being hammered and they can run all day like this (they could run @ even higher temps safely) w/o fear of component failure.

i'm running 12 x 6990 at the absolute edge in overclocking mode (power, temp) - they are all running at 100% 24/7 - unfortunately i have to clock down if the ambient temp is rising, but i use a script for that - a few pic: http://imgur.com/a/usqDL

I love this pic https://i.imgur.com/fevHc.jpg

Would you care to share your script(s) or how you show temperatures and the last picture where it shows V (is that volts?)?

Sorted.

Yes, run the fans at 100%. You'll get more lifetime from the cards with lower temperatures, but the fans might wear out sooner.

sorry for the OT

the bash script for the autom. clock control (based on temp) is from the forum here, but i don't have the thread in mind, you may try the forum search function

the lower terminal windows are the output of lm-sensors (took me 2 days to fix all the kernel and driver stuff :-( ) - important for me is that temp sensor, it give me some estimate about the ambient (server room) temperature - the A/C is a old lady already  :-\

the V - yes that are Volts, since there is nowhere a hwmon for the 12V, i use the 3.3 V reference as a guessing - in idle it is around 3.3, below 3.10 the motherboard or the cards crashing frequently, so it is a good guess ;-)

I think my card's temp is an issue. But I have no idea how to fix it. I currently only have 1 6970 and when it's mining it's at 90-92 C range. But when I stop mining, it gradually drops to low 50s. Reading this thread, it seems like 61 C and 71 C is the range 6990 should be in. 6970 is even less hash power than 6990; under that logic, should the temps be lower?

But how do I fix the issue? For Win system, people use MSI Afterburner to control the GPU fan; but what does Ubuntu (Linux) use? I looked at MSI Afterburner and it doesn't look like they have a version for Ubuntu. So I youtube and found out people use LM-SENSORS. I haven't tried setting that up; but the whole clip looks weird.

Anyways, enough of my ranting. As usual, I welcome any and every advise.

Why are you running 12 6990s may I ask? Got a need for speed?