WHEA Errors when overclocking i5 3750k

Associate
Joined
26 Jul 2011
Posts
503
Location
Harlow
WHEA Errors when overclocking i5 3750k

Hi 8 Pack,

When overclocking my i5 3750k i can get this prime stable and even run games but then i check the event viewer and i have pages of WHEA errors regarding CPU core parity checks failed.

Are these to be worried about and are they a sign of a failed OC, i no longer get the WHEA errors at a much lower OC 4.2 Ghz. What would you suggest to stop the errors and could they damage the CPU? Would more Vcore be the answer?

Thanks
 
It is a sign of instability some where within the system yes. Usually with overclocked CPU, RAM or IMC / VTT.

I would first Run memtest without CPU overclock but at the XMP to rule this out.
 
A related question.

When people are using Prime95 for stress testing are they enabling 'Round off checking' and/or 'SUM(INPUTS) error checking'?

When I was testing without them enabled the overclock appeared to be stable, but I was getting WHEA errors. With them enabled Prime95 was flagging errors which usually coincided with WHEA entries in the logs.

It isn't something I've seen mentioned. General advice seems to be just test and see if something really obvious crops up (application crashes, BSOD, etc.).

How many of the people claiming to get Ivybridge processors up beyond 4.5GHz at their first attempt, with really low volts, are fooling themselves?
 
It used to be that Round Off Checking etc were enabled by default, but I've seen some versions come with them turned off, so it's best to always check because as you say it masks instability.
 
It used to be that Round Off Checking etc were enabled by default, but I've seen some versions come with them turned off, so it's best to always check because as you say it masks instability.

I thought they used to on by default back in the PII/PIII days. The versions I downloaded recently had them off.

Maybe people should start mentioning it when Prime95 is recommended. It would probably be worth adding a note to the download link that's in the sticky at the top of this forum as well.
 
A related question.

When people are using Prime95 for stress testing are they enabling 'Round off checking' and/or 'SUM(INPUTS) error checking'?

When I was testing without them enabled the overclock appeared to be stable, but I was getting WHEA errors. With them enabled Prime95 was flagging errors which usually coincided with WHEA entries in the logs.

It isn't something I've seen mentioned. General advice seems to be just test and see if something really obvious crops up (application crashes, BSOD, etc.).

How many of the people claiming to get Ivybridge processors up beyond 4.5GHz at their first attempt, with really low volts, are fooling themselves?

you can use low volts just depends on your cpu,

sxj7yb.jpg
 
you can use low volts just depends on your cpu,

I never claimed you couldn't. I just think that some people will be getting a false impression of how stable their machines are because the error checking in Prime95 defaults to off. With the checking off you can get thousands of WHEA errors without any indication of a problem from Prime95. You basically have to push things until the machine falls off a cliff.
 
Last edited:
I never claimed you couldn't. I just think that some people will be getting a false impression of how stable their machines are because the error checking in Prime95 defaults to off. With the checking off you can get thousands of WHEA errors without any indication of an problem from Prime95. You basically have to push things until the machine falls off a cliff.

thats true,but if its gonna fail it usually will,gaming is a good test for stability as that is the thing that craps out first,either cpu v or imc/dram v not enough,same goes for video editing/encoding
 
^^^ Vccio / VTT. Vcore or Tune PLL (Down to near 1.65 or up to near 1.9). Also as the signal toteh DRAM becomes weaker when you near the top clock you can acheive add a touch of DRAm voltage too if need be.
 
Back
Top Bottom