Same here, bought a GTX1080Ti FTW3 Elite in 2017 and at the beginning of 2020 the card startet crashing even at default settings in games and in benchmarks....EVGA sent me a "new" card, which has the same problem after 6 months.
Crashing at games with default settings, manage to do benchmarks just fine. Not as bad as your card, but can only make the card work by dropping powerlimit to 85-90%. Both cards have new thermal paste from the beginning and are at 70-75C while gaming.
I'am realy pissed of, Precision X1 is probably the problem. I guess both cards where damaged by this software....even at default settings, the GPU hits the V-core limit of 1,07V...this can't work out well over time.

MitchWh
Same here, bought a GTX1080Ti FTW3 Elite in 2017 and at the beginning of 2020 the card startet crashing even at default settings in games and in benchmarks....EVGA sent me a "new" card, which has the same problem after 6 months.
Crashing at games with default settings, manage to do benchmarks just fine. Not as bad as your card, but can only make the card work by dropping powerlimit to 85-90%. Both cards have new thermal paste from the beginning and are at 70-75C while gaming.
I'am realy pissed of, Precision X1 is probably the problem. I guess both cards where damaged by this software....even at default settings, the GPU hits the V-core limit of 1,07V...this can't work out well over time.

huh. now that i think about it, i reckon it started having issues around the time X1 came out. I can't RMA because its past its warranty now. is there a specific problem with X1 that you could narrow it down to?

the_Scarlet_one
Precision isn’t damaging your hardware.

Which bios have you been running the card on?

i havent updated gpu bios before. Only nvidia drivers. Is that an issue?

This happens to me too

time to change 3000 series cards

i recorded the values of my gpu when it crashed, this is what i got.
GPU Clock: 1999.5 Mhz
GPU Memory Clock: 1377 Mhz
GPU Temp: 56 degrees celcius
GPU Load: 100%
GPU Voltage: 1.06V

Renkei
Take it apart and reapply thermal paste, i had a friend who got a kingpin card that kept tripping over temp due to a really shoddy pasting job from factory. they could of overlooked it when they tested it on return.

Are you sure? my temps are fine so I dont think that would be a solution

I also have just have this problem with my K|ngP|n, 11G-P4-6798-KR!.. Fortunately seems mine still has some time in warranty. But it is going to be a pain to ship it and the wait... And to survive some days with my old MSI TwinFrozr 670 that has been lying in its original box for years now. Maybe I just stick to the onboard videocard until it returns?.. Well I didn't even send it. It works well for desktop display still and I just opened a ticket. I wonder what they're going to suggest me to try before assuming it is defective... :(
 
In my case, I purchased Fallout 4 in December and played it for several hours before it started to artifact + CTD. As I was still fiddling with tweaks and fps unlocks, I assumed this could be an issue exclusive to FO4. Then I pulled power down to 86% and it worked well for some weeks ... with some occasional artifacting.
 
Before that, I have played Skyrim SE at 110FPS since I purchased the VGA, around September 2018. Plus my 1440p (2560x1440) display, I am pretty sure it's been pushing the card to its limit for a long time. Always been in "green" (normal) BIOS switch. I currently have 566 hours on Skyrim SE; give or take, I'd say 400 of these were spent in the K|ngP|n card without a single hitch. Actually, around Riften, if I came from Helgen, in the trees was the only place I noticed it falling from 110 FPS down to 85-95.
 
Today, as Fallout4 crashed even at 86% "target power" underclock on Precision X1, I turned it back to stock (100% power, no underclock, no overclock other than the "stock settings"), and within 5 minutes in Skyrim, the game artifacted and froze. So sad, this is something that never happened with me on Skyrim! :( I guess it is really busted this time. Skyrim froze even though I was on a place (outdoors) where it could keep at 110fps without issues.

I start having sudden hard crash/reboot problems with my 1080 Ti FTW3 about 2 weeks ago.  The PSU was tested and clean.
 
When I toggle the card into debug mode, I am able to limp by but the performance is severely hamstrung.
 
I'm about 6m out of warranty so I suppose I could check the thermal paste.  Can I use regular thermal paste that you'd use on  CPU?

I was really looking at tips on good thermal paste. I ended up with a evercool one I had lying about here.
As I said, I was out of warranty, so I just went hands-on. Took apart back plate, cleaned up all the messy melted thermal ribbons, then went front...
Voila.



Am I mistaken I see a big bubble around the left-right, going down and then at the bottom parts of the silicon? If thermal paste was present there, it *must* have been stuck to either the heatsink or the silicon, no?
It is funny how it looks like there's excess, round applied around the heatsink. Heatsink paste was solidified, and the one in the die was still greasy.
I then cleaned al that, including most of the excess grease that was (likely) heating up the poor die "cache" around the silicon (it would be cache memory in ordinary CPUs). I also used isopropyl alcohol to wipe clean clean the contact surfaces. I noticed a small line of dried out thermal paste in the die that didn't go just by softly rubbing around. And then I very carefully scratched it out with a metallic thin-tip tweezer. Quite dangerous!
 
Following the meticulous cleanup carefully reapplied the evercool white paste:

No grease on the heatsink in advance. Instead when I applied the heatsink above, I have gently moved it to seat all paste evenly. Has been worked very well on CPUs. I went the other way to what HeavyHemi suggested and placed -just- around the die, and didn't actually spread around much. I wanted to maximize contact between the silicon and the copper only. But I was towards HeavyHemi idea in the sense that I was more generous in the thermal compound layer I pored in. I used a cut zip-tie end to spread the paste evenly in the die. And motion in hopes it would adhere well to the copper heatsink.
I focused in avoiding the paste to spread heat to the poor suffered "cache" components around, but there is probably some mess I didn't remove the heatsink to look up after placing it. This would just increase the chance to create air bubbles between the compound and copper. :)
 
Well... Then reassembled everything, every screw in place, back in its dearly PCI-X16 slot, booted up like a charm (there were some problems with moody hard disks, you know the drill, always something "loses contact" when we mess up around the case).
Up to before I crashed for good, I could only play the game underclocking in Precision X1 to 86% Target power. I was in a save point in Fallout4 where I just needed to "trade items" with a nearby NPC that the game would artifact and crash at the spot (well, around 5-10 seconds in the exchange interface).
 
So then I opened Precision X1 (PX1 hereon) and all my monitoring stuff I always leave on while the computer is on, but left PX1 at 100%, all stock. Opened Fallout4 to the said scene... Hanged out for a while searching closely for the colored deadly spots that sprout on screen ... 
 
No deadly big color circles, no crash!!! \o/

The funny thing is how didn't this happen for so long with the thermal paste applied like that. That's copper dissipation I guess!
I still have lots to test but undeniably the thermal paste review improved things in the least! So I may still have light in the end of the tunnel!
 
So for you all also having artifacting issues and wonder if your case may be the same as mine, notice this is not VRAM artifacting. It (I believe) was GPU artifacting. How can I tell them apart: VRAM artifacting usually is persistent and not always crashes the game; when severe it shows crazy stuff on screen. My specific case was just huge "dots" (circles) in different colors, sometimes just white, that popped in random screen locations under GPU average load.
For instance, once this started I could just look down to the ground (GPU load would go down) and the game won't crash. The game would only crash if I kept looking to the graphics intensive scene while the circles rained gradually more on screen. This could happen in an interval of 2-3 seconds. If when I noticed the first "drops" I looked down, then it would "recover" and I could make the process restart if I looked back at the scenery.
 
This specific behavior was actual under the Fallout 4 game. In Skyrim SE (it puts a strain in my 1080 cause I play at 1440p, 110FPS), the crash was inevitable once it started, and it consisted of only white circles. In Skyrim also, the game froze all the times, while Fallout4 almost every time crashed to desktop. Only once I had to completely reboot. The first time Fallout failed after my 86% underclock (at that item exchange with NPC I mentioned above), the screens went standby (no signal) and I heard the "USB disconnected" sound from windows, the keyboard was still responsive but I simply had no clue how to shutdown and what damage could have been done.
Other than that, if I hit the horse with FurMark, it crashed without artifacting. I had to set it up at least with antialiasing in order to load the card enough for it to crash.
 
Hope this helps other people out of warranty with similar issue with their K|ngP|n 1080Ti!

avengerx
I no longer think the thermal paste job improved the problem anymore.
It fixed the problem for me. Now gonna order a better thermal compound just in case. :)

"I no longer think the thermal paste job improved the problem anymore."

Not sure I follow your post here, and as far  I could tell you did follow my advice about spreading the TIM evenly across the die instead of just putting a pea size dollop in the center.
I've had to redo a paste job more than once. Also what was the condition of the thermal pads on the memory and VRM? Lastly that area around the die is not 'cache'. Those are all filter caps for power. There are no active components on the GPU substrate external to the die. You don't want a lot of excess squeeze out, but some, is normal and won't cause any issues.

avengerx
I no longer think it improved the issue because it simply solved it. So I am now sure I addressed it with the thermal paste application.
 
 
 
Sorry I can't make succinct posts :P

I can . I am glad that replacing the TIM rescued your GPU.