Question PC Shutting Down After Some Use

[K5]

I have a Sapphire NITRO+ RX 580 8GB card on an ASUS ASRock Z97 Pro3 with a Corsair CX 600w PSU and three HDDs (1TB 860 EVO, 1TB Seagate, 4TB Seagate), LG DVD Writer, Memory Card Reader, USB 3.0 PCI-E x1 card and a Sound Blaster X-Fi PCI sound card.
For the most part the PC was working totally fine but in some specific cases it began turning off and booting with a black screen. I have also noticed the GPU fans to be turned off. After some time, the system (including the gfx card) powers up normally and I may not see this issue again for months. In some cases I have also noticed the GPU fans try to spin up but the system shuts down shortly after (in some cases the system even POSTed successfully after this).

Power supply tester readings are fine. OCCT PSU stress test does cause an immediately power off though. I'd appreciate it if someone could help me troubleshoot this further.

 

[Arkaign]

So you've got three main possibilities of any note, and they can't be fully explored without special tools or replacing components to test and rule them out.

I'd say probably :

80% probable PSU faulty. This can happen for any number of reasons, even like if you're at the grocery store or whatever and a transformer in the area has a rolling surge that rips through and bulges a cap in there just a bit, or otherwise reduces tolerances.

10% probable GPU faulty. Could be bad or flaky VRM on there causing it to sometimes pull too much or too little power. (You can partially test this by running onboard video from the Z97 Mobo, but that might also lower load enough on a flaky PSU to mask the issue if it turned out it was not the GPU causing issues).

5% probable Mobo faulty. Same reasoning as GPU : flaky/damaged VRM and power delivery circuitry and ICs.

5% other gremlins. I've even seen situations where a cardboard crushing machine in the stockroom of a big box store caused intermittent crashing and flaky behavior in the managers PC across the building through some bizarre wiring and electrical weirdness on the 110 line and wiring. Back in VGA days I knew someone who was killing GPUs about once a month for a while, turned out his ViewSonic 19" CRT was somehow returning some bad voodoo down the cable and killing the cards, lol. You'll never live long enough to see every weird potential problem with PCs that exists.

All that said, most likely #1, still definitely possible #2, less likely but possible #3. Got any spares around or a buddy who could lend you a spare PSU to test with? Intermittent failures are incredibly aggravating to deal with. In major IT support situations we don't even diagnose them in place, the uptime is too valuable and the downtime too risky for high value units and users, so we just outright replace the entire stack and test the individual components later on, or recycle them if its not worth it.

 

[VirtualLarry]

OCCT PSU stress test does cause an immediately power off though.

PSU or GPU, almost certainly. Remote case that it's related to mobo VRMs not being able to handle the load as well.

Is this system overclocked?