- Nov 2, 2011
- 8
- 0
- 66
Since a few days ago my main rig has started experiencing random BSODs. Interestingly enough, they come in stages: first the application I'm using freezes, then the whole desktop, including mouse pointer, and finally the blue screen cometh.
After that, the usual data collection process doesn't even begin - the percentage indicator stays at 0% and I have to manually shut down. Memory/kernel dumps are not created either.
This is, sadly, not the first time I'm dealing with random BSODs: around last June I had a very similar issue - the main difference being that the system would flash a blue screen before rebooting no matter what I set in Start-up and recovery, and still not generate dumps.
What fixed the issue back then was replacing the SSD, a Samsung 970 Evo which I bought in 2018 (same time as the rest of the build) with a 970 Evo Plus (due to the Evo no longer being available). Where it gets weird is that after RMA'ing the drive to Samsung (it was under warranty but the process would take a month and I needed my system back), they sent it back saying that they could not find any issue when they tested it.
As a last ditch attempt I also tried wiping everything and reinstalling Windows: when doing so I usually clean the drive via DiskPart to have the installer operate on an unpartitioned drive, but this time
would fail with an error message. I worked around the issue using the sanitize NVME function of my motherboard's BIOS, which allowed me to install Windows again, and get another BSOD within an hour.
So, once again to me the SSD looks to be the culprit, however it seems strange that
- The original SSD would come back with a clear bill of health from the manufacturer
- I'd get two bad drives in a row, from a pretty reputable manufacturer. Seems like exceedingly bad luck
But if the SSD is not the cause, then what? If it was the motherboard or RAM, swapping the drive last year would not have solved the issue. And it's not like I have swapped any other components recently.
After that, the usual data collection process doesn't even begin - the percentage indicator stays at 0% and I have to manually shut down. Memory/kernel dumps are not created either.
This is, sadly, not the first time I'm dealing with random BSODs: around last June I had a very similar issue - the main difference being that the system would flash a blue screen before rebooting no matter what I set in Start-up and recovery, and still not generate dumps.
What fixed the issue back then was replacing the SSD, a Samsung 970 Evo which I bought in 2018 (same time as the rest of the build) with a 970 Evo Plus (due to the Evo no longer being available). Where it gets weird is that after RMA'ing the drive to Samsung (it was under warranty but the process would take a month and I needed my system back), they sent it back saying that they could not find any issue when they tested it.
As a last ditch attempt I also tried wiping everything and reinstalling Windows: when doing so I usually clean the drive via DiskPart to have the installer operate on an unpartitioned drive, but this time
Code:
clean
So, once again to me the SSD looks to be the culprit, however it seems strange that
- The original SSD would come back with a clear bill of health from the manufacturer
- I'd get two bad drives in a row, from a pretty reputable manufacturer. Seems like exceedingly bad luck
But if the SSD is not the cause, then what? If it was the motherboard or RAM, swapping the drive last year would not have solved the issue. And it's not like I have swapped any other components recently.