System restarts under heavy load... PSU?

darkmore

Prominent
Dec 1, 2017
6
0
510
Two days ago I was playing Ark Survival Evolved when after about 20 minutes played my computer restarted itself. No bluescreen or error message, just the restart. I loaded up Ark again and the same thing happened, after about 20 minutes play time the system restarted itself. I've had the computer for just under a year, and had almost 300 hours played in Ark before the problem started.

This led me to believe it was either overheating or the PSU was not supplying enough power. I do believe I have thoroughly tested temperatures and they all seem to be well within tolerance. But I don't know anything about power supply units or voltages.

My system:
MSI H110M Gaming
I5-6500 3.2 GHz
MSI RX 480 ARMOR 4G OC
2x 8GB DDR4 2133
SanDisk SSD Plus 480GB
EVGA Supernova 550 G2
Windows 10 64 bit

I used HWMonitor to monitor temps during tests and CAM to monitor temps during gaming.

My testing:
1. I ran Prime95. CPU load 100% Max temp 69°C. No restart, no error message.
2. I ran Heaven benchmark. GPU load 100% Max temp 72°C. No restart, no error message.
3. I ran Furmark. GPU load 100% Max temp 82°C. The system restarted after about 20 minutes.
4. I ran windows memory diagnostic. No errors found.
5. I ran Heaven benchmark and also started Prime95 shortly after. Running both together- CPU and GPU load 100%. CPU Max temp 68°C GPU max temp 72°C. After about 15 minutes of both
together, the system restarted.
6. I ran Ark Survival Evolved. CPU load 50-60%. GPU load 100%. CPU Max temp 68°C. GPU Max temp 72°C. System restarted after about 20 Minutes.

The temps don't look bad to me, and it's only under heavy load that the system restarts. I played a couple of hours of Skyrim SE today without problem and everything else seems to work just fine. I don't have any other games as resource hungry as Ark though.
Is there anything else I can try? I'm currently stuck on thinking it's the PSU.

Edit- I just realized I had updated the video driver only a few days before. Could that be the problem? Maybe I'll try an earlier driver just to be sure.
 
Solution
Temps aren't horrible - and the PSU certainly shouldn't be an issue.

Looking at the benchmarks run, you only appear to get a reboot while stressing the GPU heavily (Heaven & Furmark)
It *could* be the PSU, but the G2 is a quality unit and more than capable of running that system at 100% load sustained.

I'd suspect you might want to try a GPU driver removal with DDU and a clean install from AMD's website.
http://www.guru3d.com/files-details/display-driver-uninstaller-download.html

Barty1884

Retired Moderator
Temps aren't horrible - and the PSU certainly shouldn't be an issue.

Looking at the benchmarks run, you only appear to get a reboot while stressing the GPU heavily (Heaven & Furmark)
It *could* be the PSU, but the G2 is a quality unit and more than capable of running that system at 100% load sustained.

I'd suspect you might want to try a GPU driver removal with DDU and a clean install from AMD's website.
http://www.guru3d.com/files-details/display-driver-uninstaller-download.html
 
Solution
Temps look OK, looks like the PSU might be overheating since it seems to take awhile before the system restarts.

What happens if you take off the side of case?

Also if it's a bottom mounted PSU make sure the filter is clear if there is one.

 

ajakusev

Prominent
Dec 1, 2017
1
0
510
It could be your motherboard. I just had mine replaced that had a very similar problem. Try removing the gpu and running some cpu-intensive tests, Handbrake video-encoding could do. Also, I think there is an option somewhere in windows to disable restart on critical errors, which should produce a BSOD for you to examine, there's also an option to generate minidumps and full ram dumps on BSODs which clever tech support guys can look into for you on some forums. In my case though the motherboard power-system failure BSOD was referring to the wlan driver as the culprit and with that removed wouldn't name any process, while the system was still as unstable as before, so just run some high cpu load scenarios and see what happens.

One more thing. If you will indeed keep getting these restarts even with the gpu removed, try doing a cpu voltage offset in bios and see if that increases the time your system can run without an error. If it does, then that's almost certainly a motherboard cpu power supply failing, which is not surprising if you're using a tower cooler in a poorly ventilated case.
 

darkmore

Prominent
Dec 1, 2017
6
0
510
It's a top mounted PSU and unobstructed, though it does have a filter which I clean regularly. I'll try a clean driver install next. We'll see what happens with that.
 

darkmore

Prominent
Dec 1, 2017
6
0
510
Update: My problem seems to have been solved.

I reinstalled a previous video driver (and used DDU, thanks) which seems to have solved the problem. I was able to run Ark for 2 hours without incident.

Thanks to all posters, I do appreciate the help.