System Instability/BSODs/Random Reboots

[CardinalDeVille]

Howdy!

I think I have this mostly figured out, but I want to bounce this off the community in case I am way off base. 

My "gaming rig" runs unRAID and is spec'd as follows:

• ASUS B550-F Gaming
• Ryzen 9 3950X
• 64GB (4X16GB) Corsair Vengeance Pro 3200MHz DDR4
• EVGA GTX 1070 
• 2x NVMe SSDs (Currently Removed From System)
• 2x 16TB  WD Golds (Magnetic NAS Storage)

 

Issue started with unRAID randomly rebooting; never showed anything particularly useful in logs, but it was apparent it was crashing because the array is encrypted, and needed unlocking to continue. Updated BIOS, seemed fine for a bit until earlier this week. 

 

Tried to Install Windows 11 on a SSD, kept getting a WHEA_UNCORRECTABLE_ERROR on certain actions. After a few attempts, Windows did manage to install with a particular Windows 11 ISO (Windows 10 outright refused and failed every time)

 

When trying to install ASUS Armory Crate (because I have a deep distrust for Windows Update doing driver installs) the system would BSOD every single time with the same WHEA_UNCORRECTABLE_ERROR

 

RAM has been tested with MemTest86+, max tests, all passed (it runs 10 times)

CPU seems fine, ran Prime95 for a long time (24+ Hours), no issues seem to be apparent there

GPU also seems fine, ran FurMark for a (24 Hours), seemed just fine

Ran AIDA64 System Stability Test on all options, no issues

 

I am beginning to think it is something with the motherboard, or more specifically the motherboard chipset. It seems to only tank when there are specific actions taken on the system (driver installs), and once an OS is loaded, it is stable when doing stress testing until you attempt to do anything related to drivers. 

 

I have another motherboard on order to be delivered tomorrow, but interested if anyone sees any holes in my logic in thinking that it is the motherboard. 

[OddOod]

Go into Event Viewer and find the warnings for WHEA issues. This will contain vendor and PCI ids of the device reporting the issues

 

It could be a dying drive or a dying mobo.

[CardinalDeVille]

I'll do that once I get windows back on the device; feels pretty unlikely about dying drives because it happens across several of my cache drives, and the NVMes are entirely removed at the moment. 

 

Glad you said dying mobo, because that is what I was leaning towards. Appreciate the synchronicity, @OddOod!

[CardinalDeVille]

Windows reports WHEA-Logger errors 46 and 47, but memtest showed no errors at all. 

 

I guess next steps is try the replacement motherboard and hope that its a ram socket issue (and not the ram itself) and if that is the case, then yay!

If not,