Reference R9 290 Question

[Voids]

Premise

 

I have had this card for a good number of years and over time I can only imagine that components will start to degrade. However, in this instance, it's the temperature at this time I am looking at the sensor monitoring in GPU-Z and the Temperature is 81c at 0% load and with a Core clock of 300MHz. This is completely outlandish.

 

What I think the issue is

 

I have re-applied thermal paste 4 times now, twice with generic compound and twice with the Grizzly Kryonuat. With each application, I applied a different amount and a different method. Each application gives me results ~1/2c difference. So based on this experiment I can tell that the paste is not the issue. 

 

However, I can see that and from what I can only describe is the thermal pads and their decay. My terms decay stems from the fact that I very lightly moved a piece of the thermal pad and it just came apart. Ther force was so little that the rest of the thermal pad did not move except the tiny bit that came away.

 

I would just like to know if this absurd temperature from the GPU could solely be based on the fact that the thermal pads are just too old and ineffective and if they are the cause of such high temperatures.

 

 

[Sauron]

It could be the thermal pads though the gpu itself should still be relatively cool if it's making good contact with the heatsink. Check that the fan is spinning and that there is no dust buildup clogging the fins.

[Daniel Z.]

Bad thermal pads shouldn’t make the core so hot. I would consider a new card anyways. The 570 would give similar/slightly better performance and much better power draw/thermals.

[Voids]

5 minutes ago, _StrikE_ said:

Wait you changed the thermal compound but didn't change the thermal pads ? It is possible that it could be a problem for the excess heat, since those cards are power hungry it puts strains on the VRM's which if they heat high enough can affect the temps of the other components. Good thermal pads aren't expensive so its worth a shot, if that doesn't fix it though its probably better to get a new one like Daniel said.

I did a 2-minute stress test, took the card out, opened it up and lo and behold, the VRM's were quite noticeably hotter to the touch (I know very stupid but I knew the risk) compared to the actual graphics chip.