I think my video card is going out.
I was getting all kinds of hitchy-glitchy gameplay going on, terrible pop-in, shaders/textures not filling in for like 5-8 seconds, etc
I went and got the precisionX software, and ran the fan at 95-100%. All of a sudden my issues kind of went away, and everything was good ( thread here: http://forums.evga.com/Can-PrecisionX-OSD-position-be-moved-m2499421.aspx#2499421 )
So, 1 week later, and even with the fan at 100% is still steadily going back to being hitchy and stuttery like before I used precisionX to turn the fan always on 95-100%. (in even simple games like Borderlands 2, Quake 3, etc, although Mirror's Edge Catalyst seems to run fine, but keeps going between 3gb and 7.5gb, so I think there is a problem with how much memory is actually shown to programs. I have had issues where most programs see the card as a 3gb card, and some see it at full 12gb).
It really seems like the Titan X is overheating on me, but its not. :(
 
The room its in never gets above 78F (25.5c).
The PSU is new (last year) EVGA 1300w single rail. (the whole machine is new last year, including the evga Titan X, which has NEVER been oc'ed).
The cpu is water cooled. I have checked both gpu temp and cpu temp, and both are really good. GPU never goes above 70c (I think the most it got was 68c), and the cpu never above 51c.
The case has excellent airflow, and no dust (HAF-X case), and all SSD/HDD's are air cooled.
The monitor is 2560x1440 144hz with Gsync (although I dont use Gsync), and the most strenuos game I have played (before this stuttering) only made it go down to like 39 fps? Most of the time it was always in the 89-110 fps range.
 
So, does anyone know things I can check for (besides unplugging video card and reseating it, which I did. and no dust bunnnies in the machine. checking memory. checking cpu (which has also never been oc'ed before).
I am using windows 8.1.
 
Thanks for any help you can give me.

I DONT want to send in my 12gb video memory card and get only an 8gb card back (Mirror's Edge Catalyst can use up to 7.5GB video memory with enhanced visuals on. But I want this machine to last at least 5 years, which is why I went with the Titan X and the 12GB memory on it).
:(
 
Dave

Scarlet-Tech
If you have an EVGA Titan X, I suggest reapplying the thermal material on the die. There can be a bubble or gap in the TIM, causing the issues.

You may not see temperature increases if the TIM is covering the portion of the die that monitors Temps, so the first (and easiest) thing to do is check and replace the thermal material.

P.s, if you send your Titan X back for RMA, you will get a Titan x back from rma. I don't know where you are getting the idea that you would get a 1080 back, but get that out of your thought process.

+1 try a new coating of the thermal compound. Then retest.

Dave3d
 
And taking apart the Titan wont void the warrenty?
 

As long as you do not physically damage the card and can return it to the original state in the event of an RMA, eVGA allows you to remove the shroud.  It is one of the wonderful things about eVGA.

Wasn't worth the price imo anyway

Make sure you clean the GPU with isopropyl alcohol or a paste remover (if you have it) with q-tips and coffee filters, and let it dry. It should shine like a mirror when all the paste is off, if you can't see your nose hairs, keep cleaning lol. For the application, I've always just done the bead in the middle, and smooth over the entire face of the GPU with a cleaned credit card. Others may argue, but it's worked for me for years. Use a coffee filter to clean the creditcard between passes to get a real smooth, even, thin layer. 
 
As for the memory pads, just use a tweezers and stick them back where they belong. It's usually pretty obvious where they were looking at the heatsink, like they leave an imprint where they were.
 
Hope this helps. I've paste changed pretty much ever GPU I've ever owned. Not that I don't trust the stock application, just that I know that I can have more faith that it was done right if I do it myself.

Dave3d
Yeah, I have done lots of cpu re-pastes.
I have only done 2 gpu's though, and those dang, dumb, non-sticky memory pads never go back right, they are too pressed from the beggining, and the memory heats up over time.
I have tried arctic ceramic pasted before, but the gap was too big, and I had to clean all of it off and use the pads. :(
Yes, I have a huge bottle of 98% isopropal alcohol, and lots of TIM laying around, so I am good there (although I think the last time I used any of the TIM was a few years ago, so it may be time to order some new stuff).
I wish they would sell those memory pads somewhere. :(

The best places to get them is pretty much any site where watercooling blocks and the like is sold...
 
Just one example...
http://www.frozencpu.com/cat/l2/g8/c487/list/p1/Thermal_Interface-Thermal_Pads_Tape.html
 
I'm sure if you call EVGA, a tech may be able to tell you the thickness of pad you need.

 
These might help you
 
https://www.youtube.com/watch?v=1tqRHfWkbMw
 
https://www.ekwb.com/shop...K-IM-3831109830550.pdf
 

 

I wish I still had my 1/2 inch water cooling setup.
This card would have been the perfect card to water cool.
Oh well.....

86c is hot. Max allowed temp for a titan x is 91c.

Well, JUST around to doing this.
I wish I would have said what the idle temps were (they were either 43 or 53c).
Anyway, I took it apart and applied some artic silver 5, and now the idle temp is 39 (still slowly going down also. I know AS5 will get better and stay there within a month).
Going to do some game testing, and see what temps I get, then will do the 3dmark advanced Fire Strike Ultra Stress Test and see what that is.
:)
 
Edit 1:
OK.
Game temps not so good.
Warhammer Vermintide always was 63c.
Now, its 65-66c. (I know, 1 month burn in for AS5 to settle).
And, the problems have not gone away.
:(
 
Edit 2:
OK, did the 3dmark ultra stress test. But, I cant compare it to the old one above, since I set a more agressive fan speed.
The fan speed ran at 94%, and reached a high of 72C. It also gave a stability of 97.6 % in fps.
At 100% fan speed, it reached 71c 1 time in the whole 20 runs of the test, so basically 70c. It had an fps stability of 98.5.
 
So, if the fan were to be at 60% like last time, I think it would have run HOTTER than before putting the paste on.
:(
 
I have 1 more thing to try, which is use another display port to the monitor, and see if that works better or not.
If not, I am afraid this card needs to go back.
:(

Read the edits above.
I didnt even test anything yet.
Now I have done gaming and 3dmark ultra stress test.
Not good results.
:(

Temp seems fine for 100% fan speed inside a case with a unknown ambient temp.
 

Dave3d
I was getting all kinds of hitchy-glitchy gameplay going on, terrible pop-in, shaders/textures not filling in for like 5-8 seconds, etc

Try doing #1 #2 & #5 listed here to see if helps resolve your problems. You don't have to do #5 if you didn't optimize your games with geforce experience. You should also try driver 347.88 if the newest drivers aren't helping.

Make sure you register your card if you already haven't: https://www.evga.com/support/register.asp 
 
GPU warranty info: http://www.evga.com/support/warranty/graphics-cards/
 
EVGA contact info (can call or email to start an RMA): http://www.evga.com/about/contactus/ 

Cant register.
It says invalid PN with the SN.
Stupid crap.
:(
 
Also, tried contacting them, and all they have is tech support right now.
Have to call tomorrow to get someone on the line to do an rma.
 
Thanks though.

Got it to register.
I accidentally had 1 too many 0's.
And, no, their number said that rma was closed for the night, only tech support was available.
I didnt try to do an rma through them though.