PC keeps crashing with BSOD and then restarts after a year.
PC keeps crashing with BSOD and then restarts after a year.
Posted here before with the same issue and I addressed it by switching to a different PSU that seemed to fix the issue. Annnnd now it's back. A quick breakdown of the issue: PC handles every GPU, CPU, RAM, troubleshooting and stress test just fine. The PC works fine outside of gaming. The conclusion that I arrived at previously was that this was being caused by transient spikes in power from my GPU which my PSU was not equipped to handle. I bought a new PSU that was listed pretty high on the tier list that used to be in the PSU sub forum in here. I however, didn't upgrade the wattage. Putting in the new power supply seemed to fix my issues and my PC was stable as could be for nearly a year before the issue started showing up again. Specs: Ryzen 7 5800x G.SKill Trident Z RGB 16GB 3200mhz CAS14 RTX 3080 10GB EVGA XC3 Ultra ASUS ROG Strix 750w 80+ Gold (Current) Seasonic Focus+ 80+ Gold 750W (Original) ASUS ROG Crosshair VIII Hero Wifi Arctic Liquid Freezer II 280mm Windows 10 I recorded a video of what it looks like when it happens. The times when the crashes occur aren't replicable at any specific point in a game nor are they predictable timewise. I first tried to capture a video of it happening and had an hour long video of nothing happening before I had to stop. I came back and tried it again and it happened within 7 minutes. Tried it again and it happened within 2 minutes. For a long time I didn't even realize these crashes were actually BSOD because it would only appear on my left monitor. Generally I'll play on my left screen as that's my 144hz 1440p monitor. With RDR2 and some other games, I'll play on my right screen which is my 60hz 4k monitor. So what I would see was the game freeze, then both monitors just went black and then reboot. Bear in mind that somehow this doesn't produce any minidump files. So my question is do I just buy a new 1000w PSU to handle what I think might be transient spikes or is there something more to this that I'm missing that could be causing this?
The video appears slightly unclear, but it seems to indicate a WHEA_Uncorrectable_Error. Visit C:\Windows\Minidump and verify the presence of any minidump files. If available, return to your Windows directory and transfer the entire Minidump folder to the Downloads folder (desktop works if OneDrive isn't syncing). Compress the copied folder and include it in a post. Please adhere strictly to the steps provided—Windows discourages altering file locations. If this is WHEA, it likely points to a CPU or PCIe hardware problem. The most frequent cause now involves NVMe SSDs. During storage, these drives often can't save dump files because they become offline while attempting to save data. We can also review WHEA events in Event Viewer. Open Event Viewer, navigate to Windows Logs → System and choose "Filter Current Log" on the right. In the Event Sources dropdown, pick "WHEA-logger". If any WHEA events exist, highlight them, right-click and save.
When I reviewed it yesterday, the folder wasn’t a Minidump directory as expected. I tried the path you suggested and found none. I made a new one, but it didn’t appear. I verified the Minidump settings were active and they were. The location pointed to %SystemRoot%\Minidump was correct, yet dumps weren’t being created. I switched the path to C:\Windows\Minidump but still had no output. Using WHEA logger showed some events, though none were newer than March 24, 2024. Additionally, the logged times didn’t match the crash timestamps I observed. I think it would help to note that I receive a generic Kernel-Power Event ID 41 (63) in the event viewer for each crash, but I’m not sure if that provides useful details.
Adjusted settings to view the BSOD details and expanded options. No results in the last three codes, but some discussions suggest the top line points to an SSD problem. I own an NVME M.2 with Windows, plus a regular SSD and HDD. My concern is whether the drive with Windows could be the culprit or if it’s limited to that port. Could relocating the M.2 slot on my motherboard help identify if the issue lies there?
Yes, the issue seems linked to the storage device. Keep in mind it can't distinguish between a bad port/motherboard or a failing SSD, but a bad SSD is much more frequent. Also, we often encounter problems with M.2 drives not fitting correctly, so re-seating it might help. You might also want to look for firmware updates. If your board has several M.2 slots, choose the one farthest from the GPU since heat from the GPU can lead to crashes. Additionally, verify the SSD temperature—overheating (around 70°C) can be a concern.
The drive I’m using is in a slot that’s nearest to the GPU but sits above it. The alternative slot is farther away, positioned beneath the GPU and along the fan path. This setup could intensify overheating concerns.
It's tough to predict without real testing, but thanks to the shield, the top position might absorb more warmth. The outcome really hinges on whether the shield moves heat from the hot air into the NVMe or keeps it away from it.
I’m still figuring this out. I haven’t spent much time fixing it or adjusting settings lately. I own a Steam PC, not a gaming rig, so there wasn’t much incentive to tweak it. It’s been working just fine for gaming outside of that, which hasn’t kept me motivated. Recently, it’s started showing BSODs even when used normally. Sometimes it locks up on the post-screen, other times it boots Windows but then crashes shortly after logging in. In the BIOS, I saw the 3.3 voltage fluctuating between 3.008 and as low as 2.92. Could this be related to my power supply again?