Idling at 45c....

[opio]

I have 2 msi GTX 760 ITX Editions in SLI, both of them are liquid cooled by Corsair H55's using an NZXT Kraken G10, I can't remember exactly but my top GPU used to idle in the 30's, now it's idling at around 43c - 45c while my bottom GPU idles at 27c. I just reformatted everything and reinstalled Windows 7... plus I just got brand new fans that have more static pressure for the radiators, anybody have an idea with what may be wrong? Is one of my H55's messed up?

 

[plaintuts]

Thermal paste

Reapplying thermal paste might solve the problem

 

[opio]

Well I also just realized that the top card has 1075mV going to it and the bottom card has 1065mV... which makes no sense because I haven't changed any voltage or clock settings in msi afterburner

 

[opio]

Anybody know why the top GPU is getting more voltage?? And how I can change it down to 1065mV like the bottom card is getting?

 

[Deehomes]

Anybody know why the top GPU is getting more voltage?? And how I can change it down to 1065mV like the bottom card is getting?

I have two 760's in Sli aswell. My bottom card actually sucks more juice and runs about 5c warmer than my top card. They both have the Evga ACX coolers on them.

 

[Kari]

Anybody know why the top GPU is getting more voltage?? And how I can change it down to 1065mV like the bottom card is getting?

During manufacturing the chips are tested to find the most suitable operating voltage for each chip, in your case they are actually quite close to each other. CPUs are like this as well, some of them have higher stock voltage than others. Besides 10mV difference wont cause such drastic temp difference.

Have you checked if the top card is actually idling properly? Low clocks and very little or no load? Might be a bit silly question but there are malware (bitcoin miners) that'll use the gpu while it is supposed to be idling...

Other than that check the thermal paste and mounting as suggested earlier.

 

[opio]

Anybody know why the top GPU is getting more voltage?? And how I can change it down to 1065mV like the bottom card is getting?

During manufacturing the chips are tested to find the most suitable operating voltage for each chip, in your case they are actually quite close to each other. CPUs are like this as well, some of them have higher stock voltage than others. Besides 10mV difference wont cause such drastic temp difference.

Have you checked if the top card is actually idling properly? Low clocks and very little or no load? Might be a bit silly question but there are malware (bitcoin miners) that'll use the gpu while it is supposed to be idling...

Other than that check the thermal paste and mounting as suggested earlier.

I actually thought some sort of bitcoin or litecoin malware/adware could've been the culprit so I reformatted and reinstalled Windows 7 OEM, but alas, top GPU still idles at the same temperature.

I mean both GPU's and my CPU are cooled by Corsair H55's (Yes if I could go back I would've chosen different closed loop coolers), but ya, the CPU and bottom GPU are fine it's just the top GPU, I guess I'll have to take off the waterblock and reapply some thermal paste, hopefully that'll fix it. Thanks for the explanation though, I didn't know that each CPU/GPU die has its own stock voltage, makes sense though.

My next upgrade for sure is a custom water loop.

Anybody out there who is looking into closed loop liquid cooling, Corsair is fine, but just know that basically anything below an H80i is not for overclocking and not really "high performance" (I suppose the H75 could be to an extent), and don't get anything with a radiator that is 120mm by 120mm or 140mm by 140mm that is thinner than 36mm.