Hello guys and girls, like the title says...My gfx card's idle temperatures are waaaay too high, I'm talking about 55 to 60s Celsius. I didn't even dare to put it under load when just a simple 2k youtube video brought it in the mid 70's range. It wasn't always like this, so strap yourselves for a little story. 
 
When I first bought my hybrid in february 2016, it ran like a dream. OOB boost to 1389 and an 80 or 81% asic. It idled at about 21 celsius and got up to 55 max under load. I didn't OC it because I'm too lazy to do the stability benchmarks and all that stuff. 
 
Now the trouble starts , 1 week ago I boot up my PC after coming back from work and I hear the blower fan going batsh** crazy all of a sudden (I have a pretty aggressive fan curve). I looked at the temp in preX and I saw 60 Celsius straight off the bat.
Pushing the PC's power button, 10 seconds 'till everything loads (thank the Gods for SSD's) then BAM 60' Celsius. I kinda instantly knew that meant  pump failure. 
 
Opened up the case, touched the radiator, noticed it was still cold, closed the case. Did some crims crams to get the card to 70 for a bit, checked the radiator again. Still cold. Packed the card, completed a RMA form from my retailer and RMA'd it the next day. 
 
Today I get an email from them saying that the pump is working perfectly and there are no issues with the card. At this point I really don't know that the issue is. My case has a metric f***ton of airflow, I was using a push/pull config on the radiator, I installed it above the card, all the things that optimize cooling were already there. One day it was performing perfectly and the next day it runs hotter than Satan's a** hole.
 
Any ideas on what could be the cause? Low cooling liquid? Obstructions in the cooling hoses? Mars is retrograde? 
 
Also, do you guys think the RMA dudes did their job or just booted up the card, saw that it worked and then shipped it back. (dunno if i can say the name of the retailer here, but it's from Germany and pretty well known.)
 
TL:DR
Booted up the card, temp got to 60's Celsius while idling, checked if waterpump is ded after a bit of light usage.
Radiator cold, pump dead. RMA'd it describing the issue.
RMA guys say it's all working fine and they'll ship the card back.
Should I just RMA it back to EVGA Germany and tell the retailer RMA bros to do their jobs right or?
I'm pretty sad.  
 
PS: rolled back to a older driver, same issue. 
 
 

EVGATech_JaesonW
Hi and welcome to the forums meshoo93!
 
It sounds like something may be putting a load on the card when it's supposed to be at idle. Check the clock speeds at a time when it's supposed to be idle, and see if they're higher than they should be as well. If so, something is running in the background and using the card. It could be something such as Battle.net or another game launcher, or it could be malware or a bitcoin miner.

Hi and thanks for the suggestion, but I already did that(and forgot to mention). I checked the clock speeds and all the background processes and they were all ok. The clock speed was in the low 300s I think ( can't say for sure since I haven't received my card from RMA yet). I'm currently using a 650TI and it idles at 32 Celsius and @324 Mhz. So bitcoin miners and game launchers are out of the question. 
 
edit: grammarz

arestavo
Yeah, that's the reason why I didn't go for a hybrid - I've had 4 AIO coolers die on me, all pump failures and all right outside their warranty windows.
 
Sounds like a pump failure, and nothing to be done except RMA or find a different cooler for it. 
 
EDIT: I suppose you could tap on the pump, if it starts then stops shortly after the pump is dying/dead/clogged. You could also verify that the power connector for the pump didn't come unplugged (although that is extremely unlikely).

Yea...I kinda knew that I was getting myself into when I got a AIO, but failing after 3 months? that's unexpected. What bothers me is the fact that the RMA dudes told me everything is fine. 

Check you TIM and GPU contact. If the pump is good and the rad is cold that means the heat is not being transferred to the water/rad from the GPU and could indicate poor contact.

meshooo93, it sounds like you RMA'd it through the place of purchase, correct? I highly recommend at this point you RMA it with us. If you RMA it directly with us, we can offer you a replacement instead of testing and sending back the same unit to you.