Question Random-like and more often than before BSOD, maybe linked to the GPU or power supply unit?
Question Random-like and more often than before BSOD, maybe linked to the GPU or power supply unit?
PC Specs:
CPU
: AMD Ryzen 7 7700X 8 Core, 16Thread Processor
CPU Cooler
: Thermaltake 240MM RGB AIO Liquid Cooler
Motherboard
: MSI B650-VC Motherboard
RAM
: TeamGroup Delta RGB 32GB DDR5 6000 MHz (2x16GB)
SSD
: Kingston 1TB M.2 NVMe SSD, Western Digital SN850X 4TB M.2 NVMe SSD
GPU
: NVIDIA GeForce RTX 4080 Super
PSU
: ABS ATX 850W 80+ Gold ("CS850M "Cyclops" modular ATX 3.0 series" - Corsair?)
Case
: ABS Eurus H570 Obsidian Black (Dimensions: 19.17" x 8.50" x 18.25")
OS
: Windows 11 Home
Monitor
: Samsung Odyssey Neo G9 57"
BIOS version 1.C2
So as the title suggests, my PC somewhat randomly crashes. At first, it was only when playing Baldur's Gate 3 (could barely get past the main menu), and after some reading and hunting around, I came across a band-aid fix that was working for someone else, which was to throttle the GPU power limit using MSI Afterburner or similar, which I did - and at about 85%, the issue seemingly went away. The PC was purchased brand new in FEB 2024 (nearly 2 years old). The issue first presented itself maybe a few months after purchase, but was infrequent enough to shrug off as a minor bug or such. Maybe a month or two later, I began throttling my GPU power as described. The problem seemingly went away, and I mostly forgot about it save for the occasional Black Screen of Death, but I initially thought it was the monitor (I've read of people having issues with this behemoth).
I knew this meant I had a problem brewing, but haven't had the chance to try and properly tackle it. Suddenly it's happening more frequently and more unpredictably after a period of months.
Now, it can be at any time. Had it happen at the desktop within minutes of a fresh reboot after it had just happened in a game (Elite Dangerous). Had it happen with just YouTube open. Usually, it happens not long after starting up Elite, and then after a reboot, I'm seemingly fine a for at least a couple hours.
Figure it's time to get this addressed. After digging some in Windows Event Viewer, I think I finally found the issue, or at least some sort of logs for whatever seems to be going on.
A quick Google suggests these all point to the GPU or its driver files; it's the only common denominator. There's also maybe an overheating issue or possibly something PSU-related.
From what I can tell, every time I boot up / wake up my PC, I seem to get about 50 of the following log entries in Windows Event Viewer under 'Application' tab (I wish I knew about/thought to use this tool sooner!):
* Fault bucket , type 0 Event Name: LiveKernelEvent Response: Not available Cab Id: 0 Problem signature: P1: 141 P2: ffff97088f76b010 P3: fffff8037e5a9710 P4: 0 P5: ffff970887e36080 P6: 10_0_26200 P7: 0_0 P8: 768_1 P9: P10:
* Fault bucket , type 0 Event Name: LiveKernelEvent Response: Not available Cab Id: 0 Problem signature: P1: 1b8 P2: a P3: 0 P4: 0 P5: 0 P6: 10_0_26200 P7: 0_0 P8: 768_1 P9: P10:
* Fault bucket , type 0 Event Name: LiveKernelEvent Response: Not available Cab Id: 0 Problem signature: P1: 117 P2: ffffd58ab308f050 P3: fffff800b5aa9710 P4: 0 P5: ffffd58a9d3770c0 P6: 10_0_26200 P7: 0_0 P8: 768_1 P9: P10:
* Fault bucket , type 0 Event Name: LiveKernelEvent Response: Not available Cab Id: 0 Problem signature: P1: 117 P2: ffffdc0d11b9d010 P3: fffff80460cc9710 P4: 0 P5: ffffdc0d019f40c0 P6: 10_0_26200 P7: 0_0 P8: 768_1 P9: P10:
There may have been a few others of similar style, but it was late and such a long slog through so many repeating lines...
I have -NOT- attempted a "clean uninstall and reinstall" of drivers (though I have downloaded DDU in preparation), but short of that or a factory reset, I imagine the next step is to take her to a shop. Thought before I do all of that I should get some input or advice.
I DID run 'sfc /scannow' and DISM - it found some corrupted files, but seemingly unrelated. I also appear to have "Realtek Gaming 2.5GbE Family Controller has a Hardware IO error" which I assume explains my recent internet instability, but also I assume unrelated (on a side note, some advice on how to rectify that would also be welcome).
EDIT: I grabbed HWiNFO to start investigating not long after the first crashes, and have never been able to identify any clear problems, short of my CPU occasionally getting a bit into the red, max maybe 96C on the CPU die, but I read this CPU tends to run a bit hot and can handle that range(?) I also recall an article which I believe prompted me to get this tool, that a lot of 40-series Nvidia cards had a rather cheap thermal paste job. Related? My specific GPU is a Gigabyte Windforce: GV-N408SWF3-16GD rev 1.0.
Welcome to the forums, newcomer!
Motherboard: MSI B650-VC Motherboard
Despite its well-organized VRM area, this board doesn’t have a heatsink on top.
BIOS version 1.C2
It’s recommended to update the BIOS; you can find the latest version at https://www.msi.com/Motherboard/PRO-B650...FI/support. Ensure your motherboard’s chipset drivers are up to date before flashing the BIOS. After a successful flash, clear the CMOS for extra security.
PSU: ABS ATX 850W 80+ Gold
If you bought an assembled ABS PC, consider removing the side panels to read the labeled information on the case and the PSU (or send photos if needed) to identify the model and make.
Received feedback, please update the PSU information right away. Yes, Newegg.com, Advanced Battle Stations prebuilt PC. The only changes I made were adding the extra SSD, the WD Black. I haven’t even tried to overclock it—never needed to, and with the problems I encountered, I was discouraged from attempting it.
I’ll also start downloading a BIOS update; I just looked at them.
EDIT: The PSU label on the access panel side reads 'CS850M "Cyclops" modular ATX 3.0 series'. It shows at the power cable input: 180-240V~ 60/50Hz.
Edit:
I don’t see any other labels unless there’s something hidden on an unexposed side. I plan to flash the BIOS in an hour or so before going to bed.
Advanced Battle Stations Eurus Ruby, I think this is the product ID if that helps:
ER7700X4080S
Note: Over the past few hours, my PC has run for about 8 hours with a 2-hour sleep while watching a movie. No crashes have occurred. I launched Elite Dangerous and then closed the menu after around 10 minutes. I expect to get one as soon as I start playing, then reboot and everything will be fine for the night. I’ll update here if it happens, just in case.
Edit:
It’s like taking a car to a mechanic—noise has stopped. It’s too late to do much now. I’ll do a BIOS update tomorrow after work, and probably clean my drivers and reinstall them. For now, I updated my BIOS the first time this happened, which caused me enough concern. I’m not sure where it appeared in the timeline, maybe around the same time I started using HWiNFO. In short, although the issue seems resolved today, the logs remain...
I hope this meets your expectations. I’m uncertain if they included crucial details on a side that needs careful extraction, but that’s all I have left except to carefully unpack everything. From a quick search, it looks like a Corsair PSU—possibly the one shown here. Also, I’m certain it’s clear, so I added my GPU serial/model sticker for reference.
Additionally, I can’t remember exactly when or what triggered the CPU die hitting the 96°C mark. HWiNFO indicated a consistent spike, while average temperatures stayed normal—likely in the 50s or 60s at most. I don’t have a clear app responsible, though I don’t check it often enough to see the spike in logs or over time.
I'm not sure about that PSU and those temperatures are high, do they decrease when you remove all side panels from your case? Were the radiator dusted over the past approximately two years while you kept the system? Please include the .dmp files for us to review. You should host them on a platform similar to DropBox and share a link so we can examine them.
I was mistaken, my CPU temperatures are around 20s to 30s, averaging 45, with peaks at 78. Gaming increases them slightly, but they stay below the 70s-80s range. It was late and I was tired, sorry for that.
I've cleaned once; she needs another cleaning soon, though it doesn't look too bad. I'll add this to my ongoing list of chores.
I'll share more details later.
Times seem normal considering what you shared here (mid-90s is still quite high, but not the situation). Sharing dump files Lutfij suggested would be useful. If you don’t have another GPU to replace, can you swap in one?
You said it occurred when opening YouTube just open—so the GPU wasn’t under heavy use. However, I experienced a GPU failure over ten years ago. It began with game crashes and BSODs, followed by bleeding issues that eventually affected the OS. It was the GPU itself (I changed PSUs without success).
Does this happen in Safe Mode with networking or when launching Chrome, YouTube, etc.? If not, starting with drivers might be a better approach. The issue with YouTube is unusual—under heavy load it could point to a temp or driver problem, but since it’s in Safe Mode or just an Internet browser, it seems more hardware-related. Running dump files or using Event Viewer could provide clearer details.
I don't have another GPU to utilize, and I've been too occupied to return to this task. I'm out tomorrow, so at least I can flash the BIOS and reinstall drivers tonight or tomorrow and see what we get.
In the meantime, I notice the .dmp files attached to the event viewer logs, but I'm not sure how to locate them or share them. Even though I'm running as Admin with admin rights, the folder is locked. When I try to open it, a permission dialog appears, suggesting I already have the permissions the system needs.
UPDATE:
I finally managed to update the BIOS (AMI BIOS; 7D78v1K; 2025-07-04 – latest non-beta). It was flashed, DDU ran, and the newest Nvidia driver was installed, then I rebooted. Still seeing errors in the event viewer logs. I ran a Windows Memory Diagnostic, but it didn't help.
I'll note that she's remained fairly stable despite the logs – I've had a crash since my last post, but that's all so far. I still want to figure out what's wrong, as something seems clearly off.
Temperatures have been decent, though I experienced a VRAM spike around 97%, which happened when Elite Dangerous was open with Chrome and multiple tabs, including ads or similar, after several hours of use.
I feel like I'm running out of options except for a Windows reinstall and hoping for the best. Beyond that, it's the trip to the shop. To be honest, it's not just the cost that bothers me, but the worry about getting it to the right people.
And again, I seem to need some guidance on handling these .dmp files.