PC Restarts When Gaming (No BSOD/Dump files being created)

May 8, 2018
3
0
10
I built this PC a little over 2 years ago and around the end of January (2018) I started to experience these restart crashes while playing games. I have yet to experience one of these restarts while just browsing/idle and every single time it has been while playing a game, so it seems to be directly related to stress/power consumption.

My specs:
CPU: i7 6700k skylake, stock clock, no OCs at all
GPU: GTX 980ti
SSD: Samsung EVO 850 pro 1tb (OS is on this), then I have another EVO 250G for some more storage.
Motherboard: ASUS ROG MAXIMUS VIII HERO
PSU: Corsair Hxi 1000i (Platinum)
Memory: 2 Corsair Vengeance 8g

Yes, I have restart on crashes disabled, and yes I have minidumps set up for crashes, and yet I am not getting any BSOD messages or dump files to look at.

Things I have tried:
Updated BIOS
Replaced PSU with brand new, exact same model
Replaced surge protector power strip
Updated drivers (GPU drivers also up to date)
Reset BIOS settings to default
Cleaned dust out (what little there was)
Ran Memtest86 for 8 passes with 0 errors
reseated memory
Ran a realbench test multiple times without crashing, all components working perfectly fine according to scores

Temps while under stress tests did not exceed 61c for cpu and did not exceed 72c for GPU and are not getting anywhere close to those temps while gaming and where I am experiencing these restarts, so I doubt very much it is temperature related.

After researching this for months, and seeing most problems being solved by replacing the PSU, I was pretty disappointed after experiencing another restart after replacing mine with a brand new one.

Are there any other known issues that could cause this? I'm on the verge of just replacing the mobo at this point, but even that is difficult because there is no visible damage, no blown caps, nor have there been any performance drops in any areas and the temps are about the same as when I first built the rig.

The problem can be very hard to replicate also because I have gone a couple weeks at a time without a single restart, and sometimes will have multiple restarts a night. But as I said 100% of the times this has happened, I'v had a game of some sort running.

Could I possibly not be drawing enough power from the wall? I have noticed other devices (router and other PC) have these 'restarts' at times also, but that could be typical behavior of a router and the other pc is like 7 years old, so that could be coincidence. If this is even a possibility, how would I even go about checking something like that?
 
May 8, 2018
3
0
10
I did some more testing last night and seems as though I can get it to auto restart pretty reliably now from loading the GPU to 100% for about 10-20 minutes. I cannot however get it to restart from CPU stress testing for over 5 hours. During these restarts, my GPU never went over 73C, so again, it is not an overheating issue. My mobo voltages appear to be perfectly fine in both bios and hw monitor, even when stress testing.

For the first time tonight, instead of restarting one time, my screen went black as if the video driver failed, I still had audio in game. I restarted my pc manually and the on board graphics driver loaded instead of my Nvidia ones. I did a clean reinstall/wipe of video drivers with guru3 display driver remover in safe mode, but I'm still getting the restarts, same as before.

I can play wow (a really non intensive GPU game) for hours without any problems but I usually play with capped framerate so the GPU is hardly doing any work, but not long after playing a more graphically intensive game with uncapped framerate, I get a restart. This problem seems to be directly related to the GPU and it either drawing way more power than it should (GPU went bad), GPU not being able to draw enough power from a 1000w tier 1 platinum power supply (the old psu AND new psu are bad), or the mobo is having some sort of problem with power delivery (mobo went bad).

So I either need a new mobo, both my psu's are bad, my gpu is failing in a weird way, or my RAM passed 8 passes of Memtest86 with 0 errors and is somehow failing. I'm just so lost at this point as to what the most likely solution is, as I've said, I've had 0 other performance issues (same boot times, no fps drops, same scores on realbench as when I built it)

I'd hate to go through replacing another psu, then the mobo, then the ram, to find out it was the GPU and the very last thing I tried to replace.