[SOLVED] WHEA_UNCORRECTABLE_ERROR

[Inverse93]

I get this error about once a day on average. The PC is a new build from about a week ago. I'm worried my CPU may be the issue because I was missing the second 4-pin connector for the motherboard's CPU socket. I turned on the Mobo's overclocking feature and this led to my first crash. I don't know if it's for sure the problem but I don't know how to read a debug file. Thank you in advance for your help.

I9-9900k
Z390 Gaming Pro Carbon
32gb RAM 3200mhz
RTX 2070

Here's what the windbg gave me:

WHEA_UNCORRECTABLE_ERROR (124)
A fatal hardware error has occurred. Parameter 1 identifies the type of error
source that reported the error. Parameter 2 holds the address of the
WHEA_ERROR_RECORD structure that describes the error conditon.
Arguments:
Arg1: 0000000000000000, Machine Check Exception
Arg2: ffffbc8e69757028, Address of the WHEA_ERROR_RECORD structure.
Arg3: 00000000b2000000, High order 32-bits of the MCi_STATUS value.
Arg4: 0000000000030005, Low order 32-bits of the MCi_STATUS value.

Debugging Details:
------------------


KEY_VALUES_STRING: 1

Key : Analysis.CPU.Sec
Value: 1

Key : Analysis.Elapsed.Sec
Value: 7

Key : Analysis.Memory.CommitPeak.Mb
Value: 74


PROCESSES_ANALYSIS: 1

SERVICE_ANALYSIS: 1

STACKHASH_ANALYSIS: 1

TIMELINE_ANALYSIS: 1


DUMP_CLASS: 1

DUMP_QUALIFIER: 400

BUILD_VERSION_STRING: 18362.1.amd64fre.19h1_release.190318-1202

DUMP_TYPE: 2

BUGCHECK_P1: 0

BUGCHECK_P2: ffffbc8e69757028

BUGCHECK_P3: b2000000

BUGCHECK_P4: 30005

BUGCHECK_STR: 0x124_GenuineIntel

CPU_COUNT: 10

CPU_MHZ: e10

CPU_VENDOR: GenuineIntel

CPU_FAMILY: 6

CPU_MODEL: 9e

CPU_STEPPING: c

CUSTOMER_CRASH_COUNT: 1

DEFAULT_BUCKET_ID: WIN8_DRIVER_FAULT

PROCESS_NAME: System

CURRENT_IRQL: f

ANALYSIS_SESSION_HOST: DESKTOP-473RKDK

ANALYSIS_SESSION_TIME: 07-30-2019 00:40:38.0537

ANALYSIS_VERSION: 10.0.18914.1001 amd64fre

STACK_TEXT:
ffff8d012fce5d78 fffff8056aef97e8 : 0000000000000124 0000000000000000 ffffbc8e69757028 00000000b2000000 : nt!KeBugCheckEx
ffff8d012fce5d80 fffff8056c181920 : ffffbc8e67cde5f0 0000000000000000 ffffbc8e69757028 0000000000000000 : hal!HalBugCheckSystem+0xd8
ffff8d012fce5dc0 fffff8056a73d7ca : ffffbc8e67cde5f0 ffff8d012fce5e49 0000000000000000 ffffbc8e69757028 : PSHED!PshedBugCheckSystem+0x10
ffff8d012fce5df0 fffff8056aefb146 : ffff8d012fce5f10 0000000000000005 ffffbc8e67cde640 ffffbc8e67cde5f0 : nt!WheaReportHwError+0x37a
ffff8d012fce5eb0 fffff8056aefb5da : 0000000000000010 ffffbc8e67cde640 ffff8d012fce6068 ffff8d012fce62b0 : hal!HalpMcaReportError+0x72
ffff8d012fce6010 fffff8056aefb4b4 : ffffbc8e67ae37c0 0000000000000001 0000000000000000 0000000000000000 : hal!HalpMceHandlerCore+0xf2
ffff8d012fce6060 fffff8056aefb720 : 0000000000000010 0000000000000001 0000000000000000 0000000000000000 : hal!HalpMceHandler+0xe0
ffff8d012fce60a0 fffff8056aefa7c8 : 0000000000000000 ffff8d012fce6330 0000000000000000 0000000000000000 : hal!HalpMceHandlerWithRendezvous+0xd4
ffff8d012fce60d0 fffff8056aefb9a7 : ffffbc8e67ae37c0 0000000000000000 0000000000000000 0000000000000000 : hal!HalpHandleMachineCheck+0x5c
ffff8d012fce6100 fffff8056a6a1e60 : 0000000000000000 0000000000000000 0000000000000000 0000000000000000 : hal!HalHandleMcheck+0x37
ffff8d012fce6130 fffff8056a5cb8ba : 0000000000000000 0000000000000000 0000000000000000 0000000000000000 : nt!KiHandleMcheck+0x10
ffff8d012fce6160 fffff8056a5cb571 : ffff8d0100000000 fffff8056a5cb4ae ffff8d012fc40180 0000000000000000 : nt!KxMcheckAbort+0x7a
ffff8d012fce62a0 fffff805828d138f : 0000000000000000 0000000000000000 0000000000000000 0000000000000000 : nt!KiMcheckAbort+0x271
ffffe0037fe6f6a8 0000000000000000 : 0000000000000000 0000000000000000 0000000000000000 0000000000000000 : intelppm!MWaitIdle+0x1f


THREAD_SHA1_HASH_MOD_FUNC: 74d6f842f394d0bf91f311b36b5441a27b8e18a2

THREAD_SHA1_HASH_MOD_FUNC_OFFSET: 6dff1a4b633a01ddd73cc761379d01277eaa6747

THREAD_SHA1_HASH_MOD: 48245875d4c5535ecbacf57c10ad759e5c3df6ae

FOLLOWUP_NAME: MachineOwner

MODULE_NAME: GenuineIntel

IMAGE_NAME: GenuineIntel

DEBUG_FLR_IMAGE_TIMESTAMP: 0

STACK_COMMAND: .thread ; .cxr ; kb

FAILURE_BUCKET_ID: 0x124_GenuineIntel_PROCESSOR_MAE_UNKNOWN

BUCKET_ID: 0x124_GenuineIntel_PROCESSOR_MAE_UNKNOWN

PRIMARY_PROBLEM_CLASS: 0x124_GenuineIntel_PROCESSOR_MAE_UNKNOWN

TARGET_TIME: 2019-07-30T07:27:54.000Z

OSBUILD: 18362

OSSERVICEPACK: 239

SERVICEPACK_NUMBER: 0

OS_REVISION: 0

SUITE_MASK: 272

PRODUCT_TYPE: 1

OSPLATFORM_TYPE: x64

OSNAME: Windows 10

OSEDITION: Windows 10 WinNt TerminalServer SingleUserTS

OS_LOCALE:

USER_LCID: 0

OSBUILD_TIMESTAMP: 1972-02-02 00:33:06

BUILDDATESTAMP_STR: 190318-1202

BUILDLAB_STR: 19h1_release

BUILDOSVER_STR: 10.0.18362.1.amd64fre.19h1_release.190318-1202

ANALYSIS_SESSION_ELAPSED_TIME: 1d10

ANALYSIS_SOURCE: KM

FAILURE_ID_HASH_STRING: km:0x124_genuineintel_processor_mae_unknown

FAILURE_ID_HASH: {d1b54118-491d-1f97-b145-f80457d2955b}

Followup: MachineOwner
---------

 

[PC Tailor]

Welcome to the forums my friend! Could you please post a link to the dump file?

I probably won't debug the dump file yet, as a WHEA error is pretty much solely hardware based.
First port of call is to remove any overclock you have, so reset the BIOS completely back to defaults and see if the issue persists.

Has the issue only occurred after the overclock?
Have you monitored CPU and GPU temperatures under load (look at CPU individual cores, not whole package)?

 

[Inverse93]

Welcome to the forums my friend! Could you please post a link to the dump file?

I probably won't debug the dump file yet, as a WHEA error is pretty much solely hardware based.
First port of call is to remove any overclock you have, so reset the BIOS completely back to defaults and see if the issue persists.

Has the issue only occurred after the overclock?
Have you monitored CPU and GPU temperatures under load (look at CPU individual cores, not whole package)?

Thank you! Here are the minidumps from the last two BSODs: https://we.tl/t-OqMtdDTUdk
I have undone any overclocking in my system and the bios options are the same as when it arrived in the mail.
The CPU temps are not so great. ~40c idle and maxes out at ~87c under load. I know that's not ideal, but is that really enough to cause crashes?

The issue started when I first overclocked it, which is the same day I had completed the build. It has persisted after I undid all overclocking.

 

[logainofhades]

What is the make/model of your power supply? Also have you tried to do a CMOS reset, since your overclock attempt?

 

[Inverse93]

What is the make/model of your power supply? Also have you tried to do a CMOS reset, since your overclock attempt?

PSU is a Corsair RM1000x modular power supply I bought refurbished from amazon.
I haven't tried a CMOS reset, will that undo the bios update I applied?
Since last night I also updated the bios to it's latest version and the chipset drivers to their latest version. I'll keep an eye out for another crash.

 

[PC Tailor]

I haven't tried a CMOS reset, will that undo the bios update I applied?

A BIOS reset will undo absolutely any overclock you put on. You said you already removed the overclock, assuming everything else is at stock settings, then it shouldn't make a difference, but the CMOS reset will make sure everything is set back to stock.

It will not undo the BIOS update, it only resets the BIOS settings, not the full BIOS firmware itself.

 

[Inverse93]

A BIOS reset will undo absolutely any overclock you put on. You said you already removed the overclock, assuming everything else is at stock settings, then it shouldn't make a difference, but the CMOS reset will make sure everything is set back to stock.

It will not undo the BIOS update, it only resets the BIOS settings, not the full BIOS firmware itself.

Ok, I will do that as soon as I get back home.
Again, since it only crashes about once a day on average it's going to be hard to confirm a fix until about two days after I make a change to the system. I'm also getting an 8 pin - 4 pin converter cable so I can make sure that voltage is not an issue for the CPU.

Another note: The system doesn't necessarily crash under heavy load. It has crashed more often while browsing the internet than it has during gaming.

 

[PC Tailor]

I'm also getting an 8 pin - 4 pin converter cable so I can make sure that voltage is not an issue for the CPU.

Do you mean an adapter? In which case I wouldn't if the PSU doesn't have the cable, it wasn't designed to be used that way.

You don't need the extra 4 pin typically, it's more for extreme overclocking as an extra mechanism.

 

[Inverse93]

Do you mean an adapter? In which case I wouldn't if the PSU doesn't have the cable, it wasn't designed to be used that way.

You don't need the extra 4 pin typically, it's more for extreme overclocking as an extra mechanism.

Yes, an adapter. So does that mean that the cable won't provide the correct voltage to the slot with the adapter? Ok, that means I'll probably want a different PSU down the line.
Also I've heard people use the term 'extreme' overclocking applied specifically to this situation. Does that mean something higher than bumping the CPU up 100mhz? I ask because the MSI OC feature that I first switched on only bumped the clock speed up to 4.8ghz from it's original 4.7ghz. Would that have been enough to 'fry' the CPU? If the culprit here turns out to be the CPU I don't want there to be any issues when I attempt to get it replaced...

 

[PC Tailor]

Yes, an adapter. So does that mean that the cable won't provide the correct voltage to the slot with the adapter?

Not necesssarily, it just means by design, your PSU wasn't made to function that way, and when it comes to a PSU, it should be the last thing to play around with (or go cheap on for that matter). Adapters (with PSUs) can cause more harm than good.

Also I've heard people use the term 'extreme' overclocking applied specifically to this situation. Does that mean something higher than bumping the CPU up 100mhz?

Yes, you'd be talking very extreme overclocking, taking the CPU to the maximum limits with high end cooling. 100MHz wouldn't touch it.

Another thing, the in built OC features can often get settings wrong that may generate instability. That's why many suggest manually changing each setting, so you know what has changed, and what needs to change.

If the CPU was fried, it probably wouldn't boot.

 

[Inverse93]

Update: I've set my BIOS to default and I've added the 4-pin CPU cable, which I didn't realize I already had (I didn't see that the 8-pin could split in half, duh)
I've also performed a clean sweep of my graphics drivers followed by a reinstall, which leads me to another concern
Could faulty drivers cause a BSOD with the error report I had? I did have issues installing the drivers properly and I had to use a combination of GeForce Experience and their driver browser in order to get both the control panel and GFE to run simultaneously. Today I noticed my mouse cursor was artifacting and disappearing when I mouse over to my second monitor so my suspicions are raised that it may have been an issue with graphics drivers all along.

Hopefully I've fixed it, I haven't crashed again in 24 hours and I'm keeping my fingers crossed.

Thanks for all your help thus far!

 

[PC Tailor]

Could faulty drivers cause a BSOD with the error report I had?

Technically possible, but highly unlikely with a WHEA error. Very rare for that not to be hardware at fault (even temperatures).

You're much more likely to encounter a different bugcheck if it was a third party module.

Either way, let us know how it goes my friend.

 

[Inverse93]

Technically possible, but highly unlikely with a WHEA error. Very rare for that not to be hardware at fault (even temperatures).

You're much more likely to encounter a different bugcheck if it was a third party module.

Either way, let us know how it goes my friend.

Okay. In the event that this isn't the fix, which it sounds like it probably won't be, how should I proceed forward? The caveat to this whole issue is that the crash seems to happen once a day at the most. I could try running one ram stick at a time to make sure it's not the cause, but wouldn't that mean waiting for a crash that may or may not occur each time? Would memtest be an appropriate tool to use just to factor out the Ram as the culprit?

I want to make sure I'm 100% certain that the CPU or Mobo is at fault before I spend the money on a replacement and have to return the old one.

Oh and one more thing. I added the second noctua fan to my heatsink to make sure it's not getting too hot. It currently runs around 38-40c idle on a hot day and maxes out at 87c when stress testing on cinebench.

 

[Inverse93]

https://forums.tomshardware.com/threads/i9-9900k-4-8-pin-power-should-i-connect-both.3412526/

This thread seems to imply that I needed to connect both the 8+4 pin power cables to the CPU socket. Still going to wait and see if I get another BSOD.

 

[PC Tailor]

https://forums.tomshardware.com/threads/i9-9900k-4-8-pin-power-should-i-connect-both.3412526/

This thread seems to imply that I needed to connect both the 8+4 pin power cables to the CPU socket. Still going to wait and see if I get another BSOD.

Again not usually needed, it's an extra safety mechanism, but there isn't a black and white at this point. Point is, if you have cable, yoy should absolutely plug it in, but you initially believed you didn't, so I stated its usually fine

Not always, but usually. It could well be if you were overclocking on a 9900k and it was drawing more than being supplied.

The advice was on the basis that you didn't have to cable initially!

 

[Inverse93]

Again not usually needed, it's an extra safety mechanism, but there isn't a black and white at this point. Point is, if you have cable, yoy should absolutely plug it in, but you initially believed you didn't, so I stated its usually fine

Not always, but usually. It could well be if you were overclocking on a 9900k and it was drawing more than being supplied.

The advice was on the basis that you didn't have to cable initially!

Yes, a couple days ago did I made the startling discovery that I can actually just use an 8-pin CPU cable and put half of it in the port, regardless of it splitting or not. 😄
I can't say I'm out the water yet, but I spoke with MSI a few minutes ago. According to them, their MPG Z390 boards should always have the 8+4 pin CPU power connected for stability purposes.
On the other hand, most people insist it's only for extreme overclocking. Perhaps this condition is unique to this series of MSI boards?
Anyways, I will update the thread if I experience another crash and if I don't in the next few days I will mark it as solved.
Thanks for your help!