Question Help with "WHEA Uncorrectable Error" and unable to generate crash dump ?

[laughingboy]

I'm getting a sporadic crash on my Windows 10 machine with a WHEA Uncorrectable Error. These crashes can occur when the machine is essentially idle and the core temperatures are low.

When this crash happens, the computer hangs on reset, or perhaps boot into bios if I wait long enough. Hitting the reset button does not get the computer out of the hang state. I can only recover by power cycling the machine.

My computer is configured to generate crash dumps but is never able to successfully do so when this error occurs. The system event viewer always has a logged event which states "Dump file creation failed due to error during dump creation." I have successfully forced a BSOD by hand with the "control scroll lock" mechanism and in that case it successfully creates a dump file.

I have checked for updated drivers and run chkdsk and dism fixup scans. The boot device is a PCIe SSD.

Thanks in advance for any help you can provide.

 

[Lutfij]

We're going to need a little more information. Please list the specs to your build like so:
CPU:
Motherboard:
Ram:
SSD/HDD:
GPU:
PSU:
Chassis:
OS: version(not edition) of Windows 10.

Can you get into BIOS and remain there indefinitely? Include the age of the PSU apart from it's make and model.

 

[laughingboy]

The PC is a new prebuilt MSI Aegis RS Gaming Desktop which I purchased just a little over two months ago.

CPU: Intel Core i9-10900KF
Motherboard: MSI Z590 PRO WIFI (CEC) (MS-7D09)
RAM: 32 GB DDR4
SSD: ADATA SX6000LNP 1TB
HDD: Seagate Barracuda ST2000DM008-2FR102
GPU: MSI / NVIDIA GeForce RTX 3080
PSU: Channel Well Technology GPU750V
Chassis: ?
OS: Windows 10.0.19042 Build 19042

I have seen no problems while I am in the BIOS setup screens. It is only the transition to loading Windows that fails after the BSOD. This will hang unless I power cycle -- hitting the reset button just leads to another hang (or sometimes I thing it will go to BIOS after a very long waiting period)

I have done the following tests / changes:

1. I cloned the OS from the SSD to the HDD and am now booting from there to see if this might make the BSOD go away or perhaps successfully generate a dump after a crash. It typically takes several days between failures so I have not yet seen a crash since this change.

2. I ran a 5.5 hour memory test (default settings, 4 passes) using memtest86 with no failures.

3. I have the free version of OCCT which lets me run CPU stress tests for up to an hour. I ran the "large / extreme" test for two separate one hour runs with no crash. All of the CPU cores hit the thermal throttling limit when running this test but none ever reached "critical temperature". This is much more stressful than I ever run the PC under normal circumstances, and I have seen the BSOD when the computer is essentially idle.

Thank you very much for your help!