Question System Crashing during data mining (High CPU use)

Dec 28, 2020
6
0
10
Hello,
I am hoping y'all might be of some assistance as I'm stuck.

I have the following set up:
ASUS ROG B350F MoBo
Ryzen 7 1700 (OC to 3.75)
2x 16GB Corsair Vengeance LPX (2400 OC to 2800)
Aorus RX580 8GB

My issue is the the System crashes during data mining use (Searching for trading strategies) CPU is running at full spec.
The system will crash about 5-10 mins into the program.
When it does, the screen goes black, the Light up Aorus logo on the GC flashes pretty consistently, but the system doesn't reboot unless you hit the power button.

I've updated the bios, downclocked from 3.8 increased DRAM voltage, increase switching frequency, increased CPU and SOC load line calibration... I'm kinda at a loss as to what is killing it?

Any suggestions would be awesome! Thanks

Andrew.
 
Dec 28, 2020
6
0
10
Remove any overclock. It's obviously not stable.
this seems like the obvious solution, however it's not been an issue in the past, but it has presented moreso lately, and if I'm being honest I feel a little defeated doing it, even if it is the most obvious solution. I'd like to understand why/how it is causing problems before rolling back.
 
Elevated RAM clocks were not quite as a sure thing for first gen Ryzens and mainboards...; lower those as well to experiment...

(No one can give you a firm 'this is the reason' without first experimenting with assorted settings, any one of which can cause instability)
 
Dec 28, 2020
6
0
10
Possible degradation of CPU. What voltage?
I have set the Voltage to 1.3 (from 1.187)>

Elevated RAM clocks were not quite as a sure thing for first gen Ryzens and mainboards...; lower those as well to experiment...

(No one can give you a firm 'this is the reason' without first experimenting with assorted settings, any one of which can cause instability)
Yes I was aware of this also, The RAM appears stable after running MEMTEST 86 again.

I tried to return the system to stock settings, but it wouldn't POST afterward. So I reset the CMOS and started again.

So clocked to 3.6 (CPU voltage back to 1.3), RAM to 2800 (1.35 v)
However I read somewhere along the way that the Radeon GPU software and drivers can cause these kinds of issues, I rolled it back to version 19 (was at v20) and it seems to have fixed the issues for now. I haven't given the system a really hard bench but it appears significantly more stable. I'll give it a proper bench when I have the time to do it. Having said all that I don't think I'll extend the OC beyond these parameters even if they do prove stable.

Thanks for your help everyone. I'll report back after a proper bench and stress test.
 
Dec 28, 2020
6
0
10
OK,
I've run the data mining for over an hour and no crashes, I also ran Kombustor for a good while and the system is stable (even if the scores werem't great).
Perhaps the culprit here is actually the display driver.
Reverting Radeon to v19 from v20 seems to have been the problem, or at least part of it.