Unusual random crashes, fluctuating speed, driver problems, computer feeling strange?
Unusual random crashes, fluctuating speed, driver problems, computer feeling strange?
I've experienced intermittent system failures recently, roughly every couple of weeks. During these crashes, the machine freezes completely and restarts without any warning messages. Some of my games, especially Valorant, are now showing unexpected microstutters that weren't present before. One crash even damaged the GPU driver, requiring a DDU setup. Additionally, my USB microphone isn't being detected anymore and no error codes appear. These problems seem connected—both crashes started around 2-3 months ago, with the microphone issue emerging recently. Your hardware appears to be stable otherwise, but the recent changes might be contributing. Any suggestions on possible causes?
Possible culprits include RAM, SSD, and PSU. Consider testing each component and replacing them if you have spare parts. It might be wise to use monitoring tools such as HWinfo64 to track temperatures and voltages, ensuring you're not facing issues like a failing cooler, dry thermal paste, or a disconnected fan. Good luck!
I completed memtest and achieved zero errors, which means I can eliminate RAM issues at this point. Thanks for the tip! Appreciate your assistance.
I’m unsure about the best method to verify SSD integrity. I looked at HWinfo data and found the 980 at 95% health with the 990 Pro at 99%, both reporting no errors. If this isn’t the right approach, let me know. I tested Cinebench multi-core and furmark simultaneously to stress the system for about 20 minutes without issues. The CPU reached 100°C during that time, which has been consistent since installation. I’m using a be quiet 360mm AIO, and the chip feels excessively hot—thermal paste is only a month old. GPU temperatures stayed normal, and the fan didn’t exceed 50%. In Cinebench, rendering speeds were noticeably slower in some cases, possibly around a quarter of the expected rate. HWInfo didn’t show any core throttling, which is unusual. This might indicate a CPU issue rather than an SSD problem, especially since I’m confident about the build from two years ago.
It's typical for certain squares in cinebench to run slowly—it mainly depends on the scene's complexity rather than your computer's specs. If you've set up the system consistently, you might replicate the issue again. Trying to isolate each part and swap components could help resolve a difficult problem. A clean Windows install would also be beneficial; backing up your current setup would prevent any damage.
The monitoring you reviewed provides a reasonable estimate of your SSDs' remaining lifespan. There have been some issues with certain Samsung models, so they shouldn't be entirely dismissed just yet. I concur that these processors are now rebranded toasters—Max temperature is 100°C for the chip—but after 20 minutes of stress testing, this is normal. I should note that this approach works well, but using a spare NVMe drive would also help confirm SSD health simultaneously. Personally, I rely on affordable WD blue drives and they have always performed reliably. Hopefully this gives you useful insight.
Sorry for not responding earlier—I just re-seated the RAM yesterday and haven’t had another crash. The frame feels much smoother now. I’ll wait a while to check if things stay stable or if problems return, and if it fails again, I’ll consider a fresh Windows install.