Hello, 

I own a 1080 Ti SC Black Edition since a couple of weeks and I am facing a disturbing issue where its usage falls to 0% while on full load! The card is running as is out of the box (no manual overclock).
 
It happens basically on all kind of recent games (FPS, racing, open-world ...) in which at some point I can witness frames to freeze for 1 to 3 seconds. Then it goes back to a normal highly fluid frame rate. The drop seems to occur every 10-15 min but I haven't measured it precisely yet. Please have a look at the attached monitoring charts from Precision X and Afterburner.
 
These two screenshots represent two distinct drops but with equivalent results:
     - GPU usage drops to 0%
     - GPU power drops to ~33%
     - framerate/frametime respectively collapses/rises like crazy (hard to tell in Precision X for some reason ...)
 
At first, I thought to a CPU bottleneck but by looking at the monitoring I've got mixed thought about it : in a gaming session, CPU core usage does not exceed 75%, temperature is reasonable but in the meantime what's bothering me is that in the picture, we can also see that GPU is reaching the "no load" limit. I should mention the CPU was steady overclocked to 4.5GHz. I went down to 4.3GHz and now back to its default 4.0GHz (frequency used in both pictures) but issue was/is still here no matter what frequency it's running to.
 
Below is what I tried (without success):
     - Following steps from the Basic troubleshooting thread (except #1 as I can't get Steam/Origin to work on Clean Boot from Windows to test games)
     - Running alternatively without Precision X, without Afterburner and then without both (to exclude monitoring tool interference).
     - Running aggressive fan curves for both GPU fans and CPU liquid cooler fans (to exclude any overheating; GPU/CPU temp below 65°C)
 
I'm not sure if this due to the GPU, the CPU or even the 750W power supply which could start to get weak (only 7 months old !?) or simply too short on wattage?
If there's one thing I'd rather not change, it's definitely the CPU.
 

Quick specs summary (if signature not visible) :
Intel i7-6700k (+ Corsair H75 cooler), EVGA 1080 Ti SC Black Edition, Asus Z170 Pro Gaming,  CoolerMaster G750M Bronze (Semi modular), 1440p @144Hz G-Sync monitor, Windows 10 Creators Update

 
Any help will be greatly appreciated :)
 
EDIT: The issue is gone after plugging the PSU on a distinct outlet (was previously on a 8 outlets powerstrip)

Goobers
Do you have the HDD LED plugged up to your case? Sometimes that can give a hint if the issue is related to paging to the SSD/HDD. While I personally would no longer build a system with "only" 16 GB of RAM, it should be sufficient unless there's a bugged app running in the background. You don't have RAM showing on the second image and the first is just plain too small for me to see anything clearly.

My RAM usage was at most around 10GB, leaving almost 6 free. I've monitored the pagefile with the built-in perfmon.exe tool from Windows then with "pagefile usage" from Afterburner but everything seems correct, even if I see it is always being used by Windows ... apparently by design. SSD/HDD operations are not an issue neither.
 

Goobers
The best way to check if the PSU is the issue is to swap it out with another "known working" PSU.

Unfortunately, I don't have a spare PSU to test with. Howewer I just realized it is plugged on a eight-inputs power-strip and seven of them are currently used. 
I'm gonna plug the PC on a dedicated outlet and double check cables inside the PC to see if this brings any change.
 

Goobers
Going a bit from this... don't use multiple monitoring programs at the same time. They can be installed, but having multiple RUNNING at the same time can cause conflicts.

I guess my post was a bit confusing but this is something I indeed never do. I just wanted to compare alternatively with different tools.
 
 
Anyway I'm gonna keep an eye on the RAM and pagefile (maybe temporary disable it to test) and also see what I can do for the PSU (too bad renting is not available on these things).
 

Alright so after additional testing and plugging the PSU on a distinct outlet, it seems the powerstrip was the key of the issue. After playing 40 min, I didn't get any stutter with a usage/power drop whereas it would have occurred at least 3 times on the powerstrip.
 
Looks like I will have to rearrange my desk cable management and invest in some material.
I will also keep a close look to disk access for some days to make it doesn't go wrong.
 
Thanks Goobers for your time and your suggestions :)

Strange, I am getting hard 1-3 second freezes in games along with the GPU usage cutting out and then recovering but I have a solid PSU plugged into a solid 900W sinewave UPS...  Even more strange is that this is not consistent across every one of my games.