9600KF 5ghz Internal Parity Error

theodordaniel4

theodordaniel4

theodordaniel4

Associate
Joined
21 Oct 2020
Posts
5
Hi,

Recently I overclocked my cpu to 5ghz, I'm kinda new to this topic, from what I read the parity error may occur because of the voltage being too low or too high...
Mostly I used the settings from this video: https://www.youtube.com/watch?v=RegtSFGiqxM
My question is, how should I determine what voltage I should set so no WHEA errors show.
I tested the system with prime95 avx on and the temps are fine.
The parity error shows quite rarely, I would say one time per month, randomly, I'm not sure if this appears when I'm playing a game or normal use.

Rig:
i5 9600KF
MSI z390-a pro
HyperX Predator 16GB 3000mhz
psu nJoy Titan 600
cpu cooler Noctua NH-D15 Chromax.Black

bios settings:
cpu ratio 50 with dynamic mode
ring ratio 47
cpu core voltage mode: override mode
cpu core voltage auto: With vcore on auto the voltage is set to 1.28V
cpu loadline calibration control auto
intel c-state disabled
 
1.28v seems low I would up that to 1.3v at first keeping an eye on temps, Most chips need between 1.3v to 1.35v to keep 5ghz stable.
 
Never had that error before. No idea what it relates to.

but when overclocking generally speaking if you start to have errors in CPU that means it’s unstable (kind of obvious). So you need to ramp up those volts.

typically you go balls to the wall with the volts or max safe volts on overclocks. Then test stability. If that is not stable you dial back the frequencies. Until stable then you see if you can get away with a little less volts and so on so forth.

going in blind with a volt and frequency is a bit unusual. Also you need to stress test your system. OCCT and AIDA are the tool I use for CPU stress test (for cooling and CPU stability) and overall system stability. If it falls apart at those tests it means something isn’t right and that issue won’t manifest itself frequently but surely and eventually it will. Worst case scenario is that the OS is corrupted overtime.
 
Tried 1.3V with OCCT and got freeze after 2 seconds, this is not good.
So, I go with 1.43V (supposedly max, based on the msi video) which is quite high by me but it runs OCCT without crashing.
The next step would be to decrease the voltage until that sweet spot, no?
 
theodordaniel4 said:
I managed to drop it to 1.33V, temps looks good and OCCT runs without crashing, I’ll try 1.32V but first I’ll make sure it’s stable at 1.33V.
Click to expand...
That looks like a decent volt. Although it looks like the intel 14nm can take upto 1.4v as stable vcore.

anyway turn on your step-speed or c-states and go into windows change the power plan min. CPU state to something like 25%. That way the cpu will lower the vcore when it isn’t doing much etc
 
I turned c-state on but the voltage doesn't drop below 1.33V, it may be because cpu core voltage mode is on override mode? From the MSI website: "Override Mode keeps core voltage fixed so either the CPU is idle or under load. Adaptive Mode will change the voltage depending on whether it’s idle or under load."

Should I change the voltage mode to adaptive maybe?
 
I also see an option named "C1E Support" which is disabled by default (I see that it reduces the voltage and multiplier to save some energy), I think I can leave this disabled and only activate C-State.
 
theodordaniel4 said:
I turned c-state on but the voltage doesn't drop below 1.33V, it may be because cpu core voltage mode is on override mode? From the MSI website: "Override Mode keeps core voltage fixed so either the CPU is idle or under load. Adaptive Mode will change the voltage depending on whether it’s idle or under load."

Should I change the voltage mode to adaptive maybe?
Click to expand...
Ya I think you need adaptive mode. Although I can’t remember how to set the intel voltages exactly not. I think under adaptive mode, u still can set the vcore voltage. Which will be your maximum vcore and then it will ramp up and down
 
You must log in or register to reply here.
Back
Top Bottom