Hello everybody,
I have been running my 1080 Ti SC Black Edition GAMING for a couple of months now and only have I used the custom fan setting in XOC.
Now since a couple of days I got a bit into overclocking, checked out a couple of video's.
 
But whenever I up my power target% above 100% I get graphical glitches and crashes without even touching GPU clock and MEM clock.
I tried testing with Kombuster and Uniengine Superposition, both have issues. Also running fans at 90% while testing and not getting above 78 celsius.
 
However when I reduce my GPU clock to the lowest (-188MHz) and MEM clock to around (-200MHz), the graphical issues no longer appear and runs stable.
Albeit under lower performance of course.
 
 
Is this normal or this an issue with my card?
 
 
Specs:
 
WIN10x64
mobo:  Asrock z370 k6
proc:    Intel i5 8600K
psu:     Seasonic focus gold 650w
ram:     2x Corsair 8 GB DIMM DDR4-2666
gra:      Evga 1080 Ti SC Black Edition GAMING
 
 
 

Alright thanks for your detailed response.
I do use 1 single cable connecting into the psu and from there it splits into 2 cables to the gpu ( I believe 6 and 8 pin ).
 
Is it or could it be a defective card then?
If so I do believe I still have warranty on the card.
 
I don't mind overclocking at the moment but I might want to in the future.

GGTV-Jon
 
 With the link above you will need to switch the drop down on the right from the 1000w to your 650w to see the correct cabling chart
You should have 2 PCIe (VGA) cables, run them both. EVGA has stopped supplying that splitter cable for good reason.

Ah you sir are a hero, I thought the one cable with 2 ends was specific for the 1080ti.
So now it comes from 2 power outputs from the psu correct?
 
I've tried it and currently testing, it looks good, I can run on 120% target power!
Tried setting the core clock a little higher but crashed after gradually pushing it to +100MHz.
 
Also temps seem to rise a lot more with 120% 0+ clock 0+ mem, stress test for 7 mins and got upwards to 88 celsius.
 
 

With GPU-Z open and on the sensor tab what is the perfcap reason and the temps - you will probably need to figure out how to get your GPU temps down to push it any further
 
Once your post count gets higher you can post link / add add attachments

When doing your OC don't get too focused on the +xxx number - look at what the actual clock is. Temps and power are what will be holding you back. Also that is just a 2 fan setup, there will be a limit to how hard you can push it. That card shipped with a base clock of 1556mhz and a boost clock of 1670mhz, anything you add will add to the boost. It also goes in steps of 13, so fist step is +13, then +26 onto +39 ect.
 
To keep things in perspective my 1080ti FTW3 on water and never going over 40c under full load maxes out at +65. I leave it +52 and only use the +65 when doing benchmarks and all radiator fans are on max. I am fine with this as it performs fine and that is just the silicon lottery for this chip. Core speeds are 2025-2037mhz to a peak of 2050
 
You being on air will always be fighting temps and how well your case flow is for fresh cool air
 
Look at Sajin's thread for OC instructions - https://forums.evga.com/FindPost/2820280

I just took this screen shot of my AB settings - idle
 

That is the default base clock speed for that card
 
Since I made the above screen shot just for the hell of it I took a screen shot of about 15 min of Heaven looping
 
 
 
That is what good cooling does - I do admit though there was a window open pretty close to the front of the case though
 
You could look into a hybrid cooler for yours if you had room in your case

The reason I started overclocking now is because I upgraded some case fans to artic f12's, nothing special I know but seem to push some nice air.
 
I have liquid cooled cpu with double fan and radiator in the top of my case acting as an exhaust.
1 corsair stock fan as back exhaust (above psu)
1 f12 on the left side window as intake
1 f12 blowing over the gpu, sitting as front intake

Test starts around 1 minute mark and end at 4 minute mark.
 
This is what I am getting back at
power limit: 105%
temp limit: 87c
Core: +39
Mem: +130
 
Uniengine superposition @ 2560*1440 High and High settings (+motion blur and DoF)

Just thought I'd step back in here and kinda refocus on the issues(s). Apparently the first issue of stability is resolved by using two PCIe power cables from the PSU, which was in my original list of suggestions. Your second obvious issue is temps.  You need to get your temps under load lower. The first question is: what is your case model and the set up of your fans? I've a basic idea that you have some input air, you're using the CPU AIO as exhaust on the top, which is usually good. With adequate airflow, you should be able to keep your GPU temps in the mid 70's with low noise.
Alternatively you might a a bad TIM application. It won't be that the material is inferior, they use Shin Etsu which is quality TIM.  That your temp rises gradually over time and does not shoot up rapidly to 80C+ tells me the TIM is probably okay. Have you tried taking a fan with the side of the case off and blowing air directly into the case to see if your temps decline?