Two 6990 and GPU0 high temp

[ErMurazor]

Hi everyone my first post here.

I have run a mining rig now for month with two 6990 working fine on Ubuntu 11.04. Now I see a problem with GPU0 with high temp and unstable hash rates but GPU1,2,3 is stable and not running over 100C temp. And also when GPU0 reach 100C clock speed drops to 500Mhz and GPU loads flips between 0 and 100%
Is possible I have overheated one GPU on the card if so how do check this? Or any other tip?

Code:

Adapter 0 - AMD Radeon HD 6990
            Sensor 0: Temperature - 103.00 C

Adapter 1 - AMD Radeon HD 6990
            Sensor 0: Temperature - 69.00 C

Adapter 2 - AMD Radeon HD 6990
            Sensor 0: Temperature - 74.50 C

Adapter 3 - AMD Radeon HD 6990
            Sensor 0: Temperature - 71.00 C

Adapter 0 - AMD Radeon HD 6990
                            Core (MHz)    Memory (MHz)
           Current Clocks :    500           1250
             Current Peak :    830           1250
  Configurable Peak Range : [500-1200]     [1250-1500]
                 GPU load :    100%

Adapter 1 - AMD Radeon HD 6990
                            Core (MHz)    Memory (MHz)
           Current Clocks :    830           1250
             Current Peak :    830           1250
  Configurable Peak Range : [500-1200]     [1250-1500]
                 GPU load :    99%

Adapter 2 - AMD Radeon HD 6990
                            Core (MHz)    Memory (MHz)
           Current Clocks :    830           1250
             Current Peak :    830           1250
  Configurable Peak Range : [500-1200]     [1250-1500]
                 GPU load :    98%

Adapter 3 - AMD Radeon HD 6990
                            Core (MHz)    Memory (MHz)
           Current Clocks :    830           1250
             Current Peak :    830           1250
  Configurable Peak Range : [500-1200]     [1250-1500]
                 GPU load :    98%

/ErMurazor

[1715058306]

1715058306

[1715058306]

1715058306

[Acidka]

just add few fans to blow away hot air from cards

[ErMurazor]

Just added one more fan still the same problem. Should the fan blow to the cards or away from the cards?

[RyanWebber]

Something is wrong with that core IMO...there's no way it should be running that much hotter then the others. My 6990 that had one core that was about 10 celcius hotter then the other core. But 30 celcius higher? Looks bad to me.

[lordbuggie]

What are the fan speeds?

[ghxost]

Just added one more fan still the same problem. Should the fan blow to the cards or away from the cards?

Depends on your setup and case.

Example: I have 3 fans blowing out, two on the top of the case, one on the top of the back plate. I have 1 intake on the front of the case blowing across my drives. I have passive intake holes on the side of my case right where my card is. So, passive air is sucked in due to the air flow in my case, takes the hot air from the card up to the fans blowing out. Easy setup since hot air rises.

[mizike29]

Seems to be sensor is incorrect, theres no way that core is that freaking much higher then the rest of them, seems pretty damn inaccurate,  I know the early bios versions would downclock at a 100 degrees which would explain your fluctuation after its hitting a 100, I would take it apart, and check it out, make sure its got the grease on that bad boy and hasnt dried up.  Not like these freaking cards were made to run this hot 24/7 anyways, I mean its like doing a furmark stress test foreverrrrrrr

[Maxim Gladkov]

I think you must somehow put Cards further from each other. It will help air to flow between them freely and cool down your card.