Question BSOD "WHEA_UNCORRECTABLE_ERROR"; cannot get memory dump ?

[adijr]

TLDR: Cannot get dump files on BSOD, stuck at 0%, trying to get information on BSOD "WHEA_UNCORRECTABLE_ERROR"

- - - - - - - - - -

Hi all,

I have a 2016 custom-build PC with nothing fancy (no overclocking, etc). Every 2-3 years one of the windows updates leads to repeated BSOD that indicate some hardware/firmware issue, but then after the next update it's fine for a another 2-3 years.

In the latest BSOD-cycle earlier this year, I couldn't get out of it, and several things I attempted failed. So I decided to do a fresh windows 10 install - wiped the primary SSD and did a fresh install. Upgraded to 21H2. After a day or two, BSOD started again with WHEA_UNCORRECTABLE_ERROR.

I understand a likely culprit is a hardware piece or firmware/drivers related to it. I did a windows memory test overnight (no issues) and a chkdisk on the main ssd (which said it found some issues and corrected them), but the errors are still here. I'd like to get more information from the memory dump. My current issue: the BSOD is stuck at 0%, and I cannot get a dump file. I have tons of space (several hundred Gb) available on the SSD.

Any suggestions on where to go from here?

Thank you all in advance!

OS Version: Win 10 Enterprise 21H2 Build 19044.1706
Hardware:
CPU: Intel i7 6700k
Board: Gigabyte Z170X-Gaming 7
Ram: 64 Gb DDR4-3200 G.skill
GPU: Gigabyte NVIDIA GeForce GTX 1080
SSD: Samsung 950 PRO Series - 512GB PCIe NVMe - M.2 Internal SSD (MZ-V5P512BW)

--------------------------
edits:
updated mb BIOS
tested Intel Diagnostic Tool, no issues
tested Samsung Magician, no issues
ran Windows Memory Test overnight, no issues
ran chkdisk on main ssd, no issues.

 

[Colif]

updated BIOS on motherboard?

Can you get into windows?

if you can, I would run this: https://www.intel.com/content/www/us/en/download/15951/19792/intel-processor-diagnostic-tool.html?

if you can't, run this:
Prime 95 bootable - https://www.infopackets.com/news/10113/how-fix-bootable-prime95-stress-test-hardware

Prime 95 how to Guide: http://www.playtool.com/pages/prime95/prime95.html

I don't know if 950 uses Magician, might run some diagnostics on it if it does - https://semiconductor.samsung.com/consumer-storage/support/tools/
you might need the Samsung NVMe drivers from same site as well.

whea are hardware errors, if you get them on 2 installs its unlikely to be windows.
can be any hardware so need to test it all since you aren't getting any dumps
can be caused by overclocking so remove any you might have

 

[adijr]

Thank you. I appreciate all teh suggestions -- unfortunately I had tried these with no evidence.

I updated the BIOS on the motherboard about a month ago.
I can sometimes get into windows, and I ran the Intel diagnostic tool, all passed on the most thorough test I could set.
I ran Magician about a month ago as well with no issues
Did not overclock.

Is there any other way I can test which hardware is the problem?

 

[Colif]

have you changed the size of the page file? that can be one reason not to get a Memory dump.

Is there any other way I can test which hardware is the problem?

only way to figure out cause is test everything

So you ran memtest on all ram sticks individually?
what PSU do you have?
anything else attached to PC?

 

[adijr]

have you changed the size of the page file? that can be one reason not to get a Memory dump.

only way to figure out cause is test everything

So you ran memtest on all ram sticks individually?
what PSU do you have?
anything else attached to PC?

Thanks, power supply is EVGA SuperNOVA 750 G2.
There's a few extra HDDs for storage.
I guess one thing I can do is detach all HDDs, all peripherals, and all RAM sticks except one and see how that goes.

 

[adijr]

Very weird behavior I haven't seen or heard much about before:

After some February Win10 update, I started getting ~daily BSOD with WHEA_UNCORRECTABLE_ERROR. I couldn't fix this despite following many many pieces of advice/tutorials online, so I eventually reinstalled Win10 from scratch, updated to latest version and drivers, etc, but BSOD came back within a day or so.

I was ready to start doing hardware substitutions (somehow...) when things changed, a couple of weeks ago -- no more BSODs at all (!?!?), but computer has this peculiar freezing behavior after a few minutes from a boot: the mouse still works, but most of the OS is frozen. I can't click on the taskbar. Some processes freeze completely (I can't click in them). But others only ... half-freeze. For example, in chrome I can switch between tabs, and tabs with existing content are still interactive (e.g. if webpage has an interactive component of clicking on images to make them bigger, say, that works), and I can open new tabs but I cannot go to new websites (it just stalls).

I have no idea where to go from here as this seems a bit unorthodox freezing. Any suggestions?

A couple of other random comments:

• the computer seems to stay on the initial screen (with mb logo) longer than before. So maybe a hardware thing that's stalling?
• I can't find anything useful in the event viewer, but I'm also not an expert at looking at it

--------------------------
OS Version: Win 10 Enterprise 21H2 Build 19044.1706
Hardware:
CPU: Intel i7 6700k
Board: Gigabyte Z170X-Gaming 7
Ram: 64 Gb DDR4-3200 G.skill
GPU: Gigabyte NVIDIA GeForce GTX 1080
SSD: Samsung 950 PRO Series - 512GB PCIe NVMe - M.2 Internal SSD (MZ-V5P512BW)

updated mb BIOS
tested Intel Diagnostic Tool, no issues
tested Samsung Magician, no issues
ran Windows Memory Test overnight, no issues
ran chkdisk on main ssd, no issues.

 

[rgd1101]

why enterprise? where did you get the key from? where the installation media from?
malware scan the system?

 

[adijr]

why enterprise? where did you get the key from? where the installation media from?
malware scan the system?

enterprise because I have the OS from my college, this is what they offer.

Ran windows malware scan

 

[adijr]

as an update, later the computer stopped BSOD-ing but started not responding properly at odd times. It turned out to be an HDD problem with a secondary HDD.