[SOLVED] System crashes randomly, Overclock present

[Zackyy]

Specs:
RTX 2070SUPER (Safe OC +100/+500 by JayzTwoCents [blame him])
i5-9600K (OC to 4.8GHz at 1.33V)
850W Golden PSU (new)
700W old PSU (old)
2 Kingston A400 SSD, 1TB HDD
2x16GB DDR4 3000MHz RAM

So, when I bought the CPU+GPU combo, I realised I get a little bottleneck. To overcome this problem, I tried overclocking my CPU. I heard people doing 5GHz at 1.35V or something, but I knew I had to do step by step instead of pumping it that way.
I started overclocking slowly, raising my ratio by 1 every step and stress testing for a minute or two just to see the temperatures. However it's not the temperatures that are problem, but rather a stable (as I thought) setup at 4.8GHz. While playing, i.e. Battlefield V, 10 minutes to 2 hours it's all fine, and then suddenly boom just turns off on me and turns back on again.
I thought it's the PSU that couldn't handle the CPU power, so I got a new PSU for christmas, a 850W Gold FOCUS Seasonic. Plugged all in, and turned my overclocks on. +100/+500 on GPU and 4.8GHz profile on CPU. Got into Witcher and.. well, GPU failed. The game crashed. Nothing much, but still weird.
Then I was watching a video and my pc started doing BRRRRRRRRR and it turned off again. Even without load.

Any other ideas what might be the problem?

 

[scout_03]

did you check your temp on the oveclock on both cpu and gpu .

 

[cherry blossoms]

Overclocks are dependent on silicon lottery. No guarantees you'll be stable at any particular setting or combo of settings.

Unsure if you mean total system blackout, then reboot, or crash to desktop. Both seem to be mentioned.

Motherboard wasn't mentioned. VRMs or other system component may be at play.

Other software may be at play. Recommend a clean boot config to test that.

Check your error logs.

Try another outlet.

 

[Zackyy]

did you check your temp on the oveclock on both cpu and gpu .

Tempa are quite low. GPU never gets hotter than 70° and with my Be Quiet! Dark Rock Pro 4 CPU is not hotter than 80°

 

[Zackyy]

Overclocks are dependent on silicon lottery. No guarantees you'll be stable at any particular setting or combo of settings.

Unsure if you mean total system blackout, then reboot, or crash to desktop. Both seem to be mentioned.

Motherboard wasn't mentioned. VRMs or other system component may be at play.

Other software may be at play. Recommend a clean boot config to test that.

Check your error logs.

Try another outlet.

Outlet is a problem, since I have one for every thing PC related. MB is Asus Z-390A Prime, which is decent. Usually the restart is sudden and no lags happen beforehand.

I usually turn off everything during games.
Crash to desktop happened only in Witcher 3 with GPU overclock and only with the new PSU.

 

[cherry blossoms]

First rule of stability: Is it rock solid stable with all stock clocks?

Testing "for a minute or two" tells you little about stability

No Kingston modules listed on the memory QVL at 3000Mhz, may not be stable at XMP settings

Remove any Asus Utilities
Disable any unnecessary LEDs

Are the onboard graphics disabled if possible?

Again, do you see anything unusual in the event viewer? This would probably be more relevant for witcher as a crash to desktop, than a sudden power loss/reboot scenario.

Can try additional cooling if a motherboard component is overheating.

 

[scout_03]

then use occt and stress test new psu for about 10 min also take not of voltages before and on test report any abnormal readings .