What causes my PC to crash and is there a dump file available?
What causes my PC to crash and is there a dump file available?
I accessed the computer, and when powered on, the RAM LED activated. I replaced the RAM (there were two compatible sticks at home), but now the system doesn't display anything, only a black screen, while the VGA LED remains on.
So I finally managed to get it to function. I had to unplug it and then connect power to make it work.
There was no overclock option available, but I set the fixed speed to 2400mhz.
I installed my old GPU, which worked fine, though it got quite hot after about ten minutes when only Windows was running.
Then I tried the original GPU, but it stopped working after a minute in Windows — it showed "No Source Input".
So you mentioned there were some GPU sag issues and it was getting quite warm, although sag isn't always a sign of death; I've heard it can lead to microfractures, shift the cooler slightly, warp the card, and damage the PCI-E slot. You also said it tends to run hot, especially when pushed to the limit for long periods. Stopping completely suddenly isn't a clear indication of health, wouldn't have expected that.
I don’t have much personal experience with GPUs failing this way, but my 1060 has been running for over nine years even with a decent tier C PSU. Maybe I didn’t stress it too much since I’m not a heavy user. The warranty usually doesn’t cover sag, as it’s often seen as [user error]. If you really want to try, you could attempt to remove and reattach the GPU heatsink with fresh paste. You might also need thermal pads of the right thickness—there are teardown guides on YouTube for this. Probably best after confirming with the manufacturer that the warranty isn’t valid.
Back in the day, some people tried baking their GPUs in an oven to reflow solder, but that’s risky. Components could fall off, and if not done properly, plastic residue might melt. You’d likely need a more advanced oven to avoid toxic fumes. Some folks who did this probably had industrial-grade equipment with proper ventilation.
It’s worth noting that baking GPUs is rarely practical nowadays. If it’s still working, it’s probably due to software issues—try reinstalling Windows, updating drivers, and disabling the MSI Center push installer in BIOS. For the motherboard, use board-specific drivers (like LAN and audio) and consider AMD B550 drivers if you’re using a compatible system. A fresh GPU driver installation might help.
If you’re still unsure, it could be a software problem. If that’s the case, perform a clean wipe—restart and reinstall Windows. The hardware symptoms are tricky to confirm, so it’s best to rule out software before jumping to hardware fixes.
At this point you should confirm whether the issue is related to hardware or software/driver problems. The best way to do this is to launch Windows in Safe Mode.
Running Windows in Safe Mode loads a minimal version of the operating system, containing only essential services and drivers. Usually, no third-party drivers are installed. This means you won’t be able to perform any meaningful tasks, play games, or expect your devices to function properly, since their drivers aren’t loaded. For instance, your display might appear at low resolution because only the basic Windows driver is used.
The advantage of Safe Mode is its stability due to the stripped-down nature of the system, which makes it reliable for detecting hardware faults like BSODs or crashes. If you encounter such issues in Safe Mode, it suggests a hardware problem. However, if the system behaves normally, the fault likely lies with a third-party driver or service not present in Safe Mode. In that scenario, another method can help identify the problematic service or driver.
Safe Mode can be challenging because the options available are limited. It offers internet access but also loads additional networking drivers. It’s important to continue using Safe Mode as long as possible—at least an hour or two—to see if a crash occurs. If you manage to trigger a BSOD in Safe Mode, it will point toward a hardware issue.
Check if it operates in Safe Mode using the suspect GPU.
Determine if it functions in Safe Mode with the previous GPU.
Find out whether it works in full Windows with the new GPU and the most recent driver for that card (apply DDU to remove other drivers).
Verify if it runs in full Windows with the old GPU and the latest driver for that card (apply DDU to remove other drivers).
I don't have an image with a suspected GPU issue; it's difficult to access the BIOS. The old GPU functions properly in both normal Windows and Safe Mode. I reinstalled drivers and removed the Nvidia audio driver. The old GPU works, but the problematic one fails on two systems.