cancel
Showing results for 
Search instead for 
Did you mean: 

Maximus XI Hero + 9900K WHEA CORRECTABLE ERROR - Need advice

FRBYTE
Level 7
Hello I just returned my Aorus master motherboard for several issues it was giving me game stuttering, coil whine etc and purchased a MAXIMUS XI HERO WIFI. It was working perfectly fine for 3-4 days but today I noticed the first WHEA-LOGGER error in windows event log. I suppose I was playing BF5 or Blackops4 when it happened. The description of the error is as follows

A corrected hardware error has occurred.

Reported by component: Processor Core
Error Source: Corrected Machine Check
Error Type: Internal parity error
Processor APIC ID: 4

Can anyone please help me what causes this error? (XMP is not enabled + My RAM is memtest tested 0 erros, drivers are latest, bios is 06xx stock it shipped with, all BIOS settings are stock nothing changed there) is my CPU dying or did I get a bad CPU because why am I getting this error everything stock? I have ran several UNI-ENGINE benchmark runs, PRIME 96 (BLEND TEST FOR 30-40 Minutes) and no errors. Please help me what to do? and there are no other errors of the same kind just 1 in 4 days.

One thing I did notice is while running prime 95 my Vcore initially was 1.27 for the first few minutes of test and after the cpu heats up to 80c after 20 or so minutes vcore drops down to 2.43 or even 1.22? while clocks are running the same frequency. Looking forward to help from the community help a brother out. 🙂
5,409 Views
16 REPLIES 16

Falkentyne
Level 12
Did you change the internal AC/DC loadlines to 0.01 or anything?
I believe one setting on the Asus board will set these to 0.01 automatically.
That error means your CPU voltage is too low, but first try this:

Set your CPU voltage MANUALLY to 1.20v, at 4.7 ghz, and set your cache ratio to 4.4 ghz (this is the usual recommended default voltage set at 4.7 ghz).
Then set your loadline calibration (LLC) to LLC5 or LLC6.

Now test your games and see if you get any more WHEA's.

Internal parity error occurs when the voltage is too low and a game uses AVX instructions. Stuff like Realbench and cinebench and prime95 are far more likely to generate cache hierarchy errors (CPU L0), which is a corrupt instruction register, rather than parity).

Falkentyne wrote:
Did you change the internal AC/DC loadlines to 0.01 or anything?
I believe one setting on the Asus board will set these to 0.01 automatically.
That error means your CPU voltage is too low, but first try this:

Set your CPU voltage MANUALLY to 1.20v, at 4.7 ghz, and set your cache ratio to 4.4 ghz (this is the usual recommended default voltage set at 4.7 ghz).
Then set your loadline calibration (LLC) to LLC5 or LLC6.

Now test your games and see if you get any more WHEA's.

Internal parity error occurs when the voltage is too low and a game uses AVX instructions. Stuff like Realbench and cinebench and prime95 are far more likely to generate cache hierarchy errors (CPU L0), which is a corrupt instruction register, rather than parity).


I haven't changed anything in bios. I just loaded bios defaults and started using the pc, not even XMP was enabled my cpu uses 1.27 vcore at default settings and it drops down to 1.22 or 1.23 after a few minutes while using prime95 blend test I thought maybe 1.22 is too low for stock settings (as i think MCE is enabled by default) but your comment has me worried because if 1.2 is normal for 4.7 Ghz then the voltage my chip is taking is nowhere near too low as it never dips below 1.22 maybe my processor has degraded too much? because my Aorus board was running it at 1.45 - 1.5 volts due to a bug i found that out and fixed it but maybe damage is done but it never overheated......Should I RMA the processor as it should not be throwing WHEA errors at stock without me tinkering with anything to make the errors go away. What do you suggest I should do.

Did you do a fresh Windows installation? Check BIOS/hardware drivers.

https://answers.microsoft.com/en-us/windows/forum/windows_10-performance-winpc/whea-logger/897a3ec3-...

pndiode wrote:
Did you do a fresh Windows installation? Check BIOS/hardware drivers.

https://answers.microsoft.com/en-us/windows/forum/windows_10-performance-winpc/whea-logger/897a3ec3-...


well it has only happened once and yes It is on a fresh windows install (installed 3-4 days ago when i purchased this motherboard) now I just cant get it to produce that error again but I am worried thinking I should RMA the cpu because it gave me this error when everything was stock which indicates it requires more than "normal" voltage which the motherboard was providing it with to run stock frequencies.

Might be a faulty chip! RMA if it still under warranty. Though your device works well now, it might cause problems in the long run. INTEL will give you one of two answers for a RMA: “YES or NO”.

FRBYTE wrote:
I haven't changed anything in bios. I just loaded bios defaults and started using the pc, not even XMP was enabled my cpu uses 1.27 vcore at default settings and it drops down to 1.22 or 1.23 after a few minutes while using prime95 blend test I thought maybe 1.22 is too low for stock settings (as i think MCE is enabled by default) but your comment has me worried because if 1.2 is normal for 4.7 Ghz then the voltage my chip is taking is nowhere near too low as it never dips below 1.22 maybe my processor has degraded too much? because my Aorus board was running it at 1.45 - 1.5 volts due to a bug i found that out and fixed it but maybe damage is done but it never overheated......Should I RMA the processor as it should not be throwing WHEA errors at stock without me tinkering with anything to make the errors go away. What do you suggest I should do.


You're complaining a lot yet you're refusing to do what I asked you to do.
That's not helping anyone. It's unhelpful if you won't cooperate with people asking you do specific tests.
Set your bios voltage to 1.25v, Set core to 4.7 ghz. Set cache to 4.4 ghz. Set Loadline calibration to LLC6. Don't forget to max out all current and power limits. Test prime blend with these settings or whatever you tested that gave you a WHEA correctable error.

Every CPU should pass this. If you pass, drop the voltage 10mv at a time until you start getting problems, then post where you have issues at.
Note: you said battlefield 5 uses AVX right?
Prime95 won't generate "CPU parity errors". Only games will, that use a "low load" AVX. Apex legends loves to do this on people who have otherwise stable overclocks. So you'll have to also test your games too. However what you CAN test if you can keep the chip cool enough, is prime95 29.7 build 1, and test "AVX fixed 15K FFT's". This is a good test because if you pass this, no games will crash.

If you want to save a lot of time, set your bios voltage to 1.20v first, LLC6, then test prime95 29.7 build 1 (check their forums for the FTP site).
test custom in place AVX 15K-15K fft size.
If you crash, increase voltage 10mv until you are stable. Then post back with results.

Falkentyne wrote:
You're complaining a lot yet you're refusing to do what I asked you to do.
That's not helping anyone. It's unhelpful if you won't cooperate with people asking you do specific tests.
Set your bios voltage to 1.25v, Set core to 4.7 ghz. Set cache to 4.4 ghz. Set Loadline calibration to LLC6. Don't forget to max out all current and power limits. Test prime blend with these settings or whatever you tested that gave you a WHEA correctable error.

Every CPU should pass this. If you pass, drop the voltage 10mv at a time until you start getting problems, then post where you have issues at.
Note: you said battlefield 5 uses AVX right?
Prime95 won't generate "CPU parity errors". Only games will, that use a "low load" AVX. Apex legends loves to do this on people who have otherwise stable overclocks. So you'll have to also test your games too. However what you CAN test if you can keep the chip cool enough, is prime95 29.7 build 1, and test "AVX fixed 15K FFT's". This is a good test because if you pass this, no games will crash.

If you want to save a lot of time, set your bios voltage to 1.20v first, LLC6, then test prime95 29.7 build 1 (check their forums for the FTP site).
test custom in place AVX 15K-15K fft size.
If you crash, increase voltage 10mv until you are stable. Then post back with results.




Today I returned my previous setup and purchased a new 9900k + Aorus master mobo and..........its pulling 1.35 volts at stock during prime 95 blend and games, my previous processor was pulling 1.27 or lower at stock. I have tried updating bios to latest also tried with optimized defaults and MCE turned off. Does this mean that my new processor is worse than my previous one because its pulling 1.35 under the same settings as my previous one which was pulling much lower voltage on auto...??

I think it has to be that my new processor is much weaker because AFAIK motherboards reads required values off svid table and the requirements for this one are much higher.....????

MrAgapiGC
Level 13
Before doing these. Install BIOS/ime/chipset from the board website. Skip the MEUpdate for now

The Bios flash it with 0903 USE Another computer to download the bios. use the included app to change the name of the bios
Chipset Version 10.1.17903.8106
MEI Version 1828.12.0.1152

If you use the included CRATE just find these on the motherboard website and update these manualy. and let see.

Work only with default settings of bios
Learn, Play Enjoy!

Itzycharles wrote:
Before doing these. Install BIOS/ime/chipset from the board website. Skip the MEUpdate for now

The Bios flash it with 0903 USE Another computer to download the bios. use the included app to change the name of the bios
Chipset Version 10.1.17903.8106
MEI Version 1828.12.0.1152

If you use the included CRATE just find these on the motherboard website and update these manualy. and let see.

Work only with default settings of bios


It might take ages for the error to appear again, but i have kind of given up on this build now....for months I was going from one Aorus board to another trying to get rid of coil whine then finally purchased a Maximus xi only to find out I am getting micro stutter in games ... in the middle of trying different fixes (win 10 1607 standby list cleaner do this do that) this WHEA error pops up.

Thanks for all your replies but I guess you really can't fight presistent bad luck I am going to return everything and might just buy a console. Back in the days when I built my first PC (first gen i7 x58 Sabertooth) things were not so complicated, with time manufacturers are forced to keep adding new features and in the process reliability suffers. I mean this motherboard for example has nvme slots, frickin on board RGB water pump headers, USB C, built in WIFI old rigs didnt have all of these .... AMD forced intel to come up with a 8 core chip which can't be cooled easily. PCs used to be simple and reliable my last aorus board fed my cpu 1.5 volts due to a bug when system awakes from sleep stuff like this was unheared of during the "BIOS" era, now everyone wants unreliable 5ghz *RGB puke boxes which cant even run stock specs.