It's been a good 2 weeks - OP, did you fix it or did you need to RMA?

Hello.
I just went through this whole riggamorale with my Titan X (not water cooled though).
I did all of the stuff, + re-applied thermal paste (arctic silver 5). Well, gaming temps went up by 2-5c. Yes, up, except idle temps went down 2c (although AS5 paste is known to have a break-in period of a month).
My games still were all wonky, and desktop slow as slow can be.
I was fortunate enough to have another card to throw in to test (I didnt even uninstall the drivers, left them on).
Well, with the old card, everything was great again, desktop, games, etc, worked great again.
I rma'ed after that, and evga said no problem.
I think my issue was with the memory or something though, as the temps did get a bit higher than usual (but only a few c higher than normal).
So, if you have tried everything here, or dont want to waste 2 months (like I did) fiddling with it to see if it gets better with newer drivers, or using a oc tool to watch temps and fps in games and note them all down, then retry with different drivers etc, and up the fan speed via tools (which I eventually had to do to make the card even remotely playable), then just do the rma.
Seems the 900 series cards qa (or the boards themselves) are just hit and miss (I have been with nvidia since the 90's, and never heard so many problems as these have had, besides the 440o series fans being louder than vaccum cleaners, lol).

cant say. like I said in this post no artic temps in this room [lower Alabama] with some ac on I will run a room box fan  to help circulate  the air in this room , can start to feel pretty hot like I think the AC quit working at times when the sun hits the wall just right or something ?? .   so far max  I seen with it  has been 48c [that's with a OC ]  I expected  a cool card   with nothing running it over 60c   so far the temps I get from it is better then  I expected overall ..   my asic is only 74.5 [??]   I guess like you said a bit of hit and miss  and a little luck of the draw  ?

Exidrial
Update: Yeah it's been a while sorry. I RMA'd the card. Temps are better.

Glad to hear that you got it all sorted out!

Dr.Death
me too ,  and that you got a ''new'' 980ti hybrid  in exchange as well not a refreb  . I was getting suspect to that  or if they would try to send you something like a gtx 1060 in exchange  ???.   I was getting ready for that let down  if I ever had to rma mine  .. that does give me a bit of peace of mind
 
thanks

 I think it was a new one. But why would they send me a 1060? lol Temps are nowhere near the magical 50°c that a lot of people seem to have. My new 980Ti never exceeds 70°c but rather hovers at around 57-65°c most of the time when being pushed hard by games.
Well yeah but I guess I'll have to RMA it again. http://forums.evga.com/An...-m2630327.aspx#2630443

Dr.Death
that's why  .    I mean they don't even offer the 900 series pumps anymore for some time and figured all you would get is a referb RMA [someones elses junk]  card in the hopes its better then the one you sent in  cause they don't have any new direct replacements  cards or new parts to warrant or service  them  , so in exchange they offer you ''the next best of equal ''performance ''  [LOL ]  like a 1060  air cooled card to satisfied you  .   in the end they cant service what they sell   anymore   under warrantee or not  .   I posted this concern a few times here and around  .  its my only regret on getting this card  .   no parts offered  like the pump and radiator  to fix myself   and doubt evga is now able to do so  .
 
just be sure to keep informing us how it goes with your 2ed RMA  ?   heck you may get a air cooled 1070  offered   in its place  now ?? 

Will do. Getting a working card can't be too much to ask for...

Ambient temps play a role, so unless you know the ambient temp(s) of the reference systems you cannot directly compare.

Exidrial

arestavo
Ambient temps play a role, so unless you know the ambient temp(s) of the reference systems you cannot directly compare.

They are room temperature most of the time.
It's room temperature here so my results should be comparable unless the reference system was in a 7°c room or something (yeah I know it's not linear growth)

Plus fan speeds, radiator and shroud.

Seeing as how both my old 980 Ti Classified and 980 Ti G1 Gaming (clocked around 1450 to 1500MHz) both sat at around a max of 55C under load with a H55 with two Noiseblocker B12-3 71CFM high static pressure fans in push/pull at 100% fanspeed, I'd say 57C with a stock EVGA AIO (one fan only, right?) isn't bad at all.

In fact, that's a good temp. Especially so if you don't manually set the fan speed to 100% (can you even do that with the stock EVGA AIO's radiator fan? I read that you can't unless you move the fan power to a motherboard header or fan controller).

Exidrial

arestavo

Exidrial

arestavo
Ambient temps play a role, so unless you know the ambient temp(s) of the reference systems you cannot directly compare.

They are room temperature most of the time.
It's room temperature here so my results should be comparable unless the reference system was in a 7°c room or something (yeah I know it's not linear growth)

Plus fan speeds, radiator and shroud.

Seeing as how both my old 980 Ti Classified and 980 Ti G1 Gaming (clocked around 1450 to 1500MHz) both sat at around a max of 55C under load with a H55 with two Noiseblocker B12-3 71CFM high static pressure fans in push/pull at 100% fanspeed, I'd say 57C with a stock EVGA AIO (one fan only, right?) isn't bad at all.

In fact, that's a good temp. Especially so if you don't manually set the fan speed to 100% (can you even do that with the stock EVGA AIO's radiator fan? I read that you can't unless you move the fan power to a motherboard header or fan controller).

When placing the radiator in the front infront of my SP120 fans (Push/Pull) and putting the side panel on temps rise up to around 60°c
BTW the GPU-Z window up there that shows 57°c is without any overclock. It's completely stock settings. How come others (and reviewers) get temps of 45-52°c MAX?

Edit: you put the radiator, with the stock fan still on it, in front of two other fans set up as push pull? If that is the case, those SP120s are likely causing air cavitations which can starve sections of the Hybrid radiator of airflow which would increase temperatures. Either put the two SP120s onto the Hybrid radiator after removing the stock fan, put the Hybrid radiator elsewhere, or remove those SP120 fans.
 
 
 
 
Manually set the fan speed higher, run it on an open air setup, cool the ambient temperatures to 20C or lower, win the silicon lottery (less voltage leakage, less voltage needed, lower temps), put two of the same fans in push/pull as intake and not exhaust, make a blood sacrifice to the God(s) of your choice.
 
Take your pick. 
 
Getting the exact same results as a completely different setup with a different card (as in, it isn't your personal card being tested by those reviewers) is darn near impossible due to so many different variations (including those that I didn't mention).
 
Buy a pair of quiet, high static pressure fans and put them in a push/pull configuration as intake. Replace the stock TIM with Kyronaut Thermal Grizzly. Set your radiator fan speeds to 100%.
 
That WILL lower temperatures a good deal. How much? Can't say - every card is not the same.

replacing thermal paste is best choice

Exidrial
 
I didn't spend hundreds of Euros to get a product that people suggest I fix myself... twice.

So you just want to whine here instead of trying to find a solution to your issues? Got it.

firestrike stress test  [so it says 20 loops ? ]
http://imgur.com/Ak24Pbi
 
after 3 loops  of valley and gpu-z render test at the same time .
 
http://imgur.com/cjRaHUk
 
so far during the hottest summer day in here   this card has only got to 48c  max   and that was before I set the radiator as I showed earlier here
 
http://imgur.com/UbwqG6X
 
I guess I can feel your pain  on that card you got ??  I do know and seen  any slight blockage of the air flow through the radiator  don't help  I got mine as open air as possible   even a case fan vent opening  may block enught air flow to increase the temps  a bit  [4-8c ]  easy  the radiator does need unobstructed air flow  to work at its full efficiency 
 
I played around with this card a lot for over a week trying all kinds of things out to see what worked the overall best   and some slight changes  like mounting placements and all makes a good temp differences . it don't take much to add stress to the coolers overall  efficiency . the way I got mine now has given me a 7c drop in overall everyday max temps 
 like they say what works well for me and my card may not for anyone else in there conditions 
 
good luck

not bad  compared to what you were getting .  the max I ever got tis card was 48c and that was when I was playing around with radiator placements  ..  like I say this card has cooled a bit better then I expected  to start with   [luck of the draw ??? ]    and in the heat of summer it can get pretty warm in this room when the sun hits just right on it   ..  like who turned off the A/C ???
 
looks like you now can enjoy your card  instead of worrying over it
 
good luck