PC suddenly shutting down and having difficulties booting up again

saure_teigtasche · July 11, 2023

So my last PC actually had similar problems towards the end of its life and it was resolved by purchasing a new PSU, because the old PSU was already out of warranty, but I still wanted to check if my suspicions are correct. If everyone here agrees that it's a PSU problem I'll contact my PSU manufacturer (be quiet!) since my PSU should still be under warranty.

Description of the problem: Since a few days ago, my PC has started shutting down at rather random times, but a bit of a higher load was present, i.e. one game (mainly Genshin Impact or Overwatch) was open with a few other programs (browser, discord, etc.). After the shutdown it does not properly boot up again. Sometimes it straight up will not boot up for a short amount of time. Then the fans spin and it can light up, but nothing gets outputted to my monitors. Only after I hold down the power button and turn it on again it does it boot properly.

The problems did start occuring after the most recent heatwave in southern Germany, but looking at my temps my GPU and CPU generally do not exceed 75C and for such a catastrophic failure to occur instead of just throttling a much higher temperature is necessary as far as I know. I think the problems also started after heavier thunderstorms but I do not know if this has anything to do with it.

Specs:

PSU: be quiet! Pure Power 11 500W

CPU: AMD Ryzen 5 5600 mounted with Coolermaster Hyper Evo

GPU: Nvidia RTX 2060 (6GB)

RAM: 2 X 8GB

Storage: 1 X SATA HDD, 1 X SATA SSD, 1 X M2 SSD

If any other important specs are missing you can request them.

According to the be quiet! power calculator this PSU should be able to handle this load

It even shows up in one of the recommendations.

So yeah, let me know if you have any other ideas than contacting the PSU manufacturer.

Whatisthis · July 12, 2023

Bad motherboard is my first hunch here. I’ve had a similar issue in the past that was fixed by a new board.

To test, run cinebench r23. If it is anything like the issue I faced, the high cpu load will cause it to crash. That’s the mobo power delivery not working right.

if that checks out, run furmark to put a new 100% gpu load on the system.if that passes too, run both at the same time to test your psu.

saure_teigtasche · July 12, 2023

Okay, I ran cinebench r23 for windows 64-bit systems. It ran by without a hitch, with max temperatures ~69C (nice).

I then ran the furmark 1080p preset and it also went well. Max temps around ~71C.

Lastly, both at the same time. While cinebench was running I ran furmark 1080p preset twice until I realised there's a stress test and just ran that until cinebench was done. The max temperatures for the CPU at ~74C and for the GPU at ~78C. And... it didn't crash? According to HWMonitor the power draw was at the about expected ~75W and ~160W; similar to the TDP.

I should mention this is a new day, so temperatures have fallen quite a bit. The thermometer in my clock says it's only 2 degrees cooler for now though. Maybe it's the combined load from the GPU, CPU and the less power intensive components that causes the crash?