Unusual freezes and blue screens occur without clear reason.
Unusual freezes and blue screens occur without clear reason.
Hello there, my self-built PC from around 2015 is experiencing random freezes and blue screens that appear when it starts up or during gaming. These issues seem unrelated to each other, though they occur frequently. I haven’t identified a clear pattern based on component load. When the screen goes black, error codes usually show WHEA_UNCORRECTABLE_ERROR or CLOCK_WATCHDOG_TIMEOUT. My suspicion points toward the CPU and its heaters—though they’ve always been problematic before, they didn’t cause problems until now. I re-applied thermal paste, but it was completely dry when I removed the cooler, and swapping it didn’t help much. I ran the Intel diagnostics and saw no faults. I’d appreciate an expert’s take on what might be causing these problems and suggestions for further troubleshooting, including which part might need replacement. I’ve attached the Minidump files, PERFMON logs, and a HWiNFO64 sensor log from a crash session while playing Satisfactory. Thanks in advance! My specs are: Windows 10 Pro N 64 bit Build 19043.1706 (21H1), GIGABYTE G1.Sniper Z97 (BIOS F6 - 05/30/2014), Intel Core i7-4790K Corsair H100iGTX, 2x8GB G.Skill F3-2400C10-8GTX (DDR3-2400 / PC3-19200 DDR3 SDRAM at 1200.0MHz), NVIDIA GeForce GTX 980 Ti (ASUS Strix GTX 980 Ti DirectCU III OC), Samsung SSD 850 EVO 500GB (boot drive) ST3000VN000-1HJ166 (3TB), and a Super Writermaster DVD drive. DARK POWER PRO 650W (80 Plus Gold). Please let me know if you need more details!
These error codes usually indicate a stability problem, but they don’t clearly identify whether the issue lies with the CPU, RAM, or drive. The CPU seems to be the most likely culprit. Neither code provides specific details about the fault, so I’d begin by resetting the CMOS and reinserting the CPU. Make sure the contacts on both the CPU and motherboard are clean and straight. Thermal paste might affect performance if it gets into the socket, so keep both surfaces free of dust or damage.
Clock Watchdog Timeout is a blue screen issue on Windows when the CPU isn't handling interrupts properly. The WHEA uncorrectable error appears during crashes and usually points to hardware problems, possibly a faulty CPU or misplaced one. Try taking the CPU out of its socket, reinstalling it, and checking again. If the problem persists, consider sending it in for warranty service—since your unit seems outdated, a replacement might be the best option.
Thanks for your prompt responses. I removed the CMOS and examined my CPU; no unusual signs were detected (I’ve attached some images of the processes and the CPU). The copper heat sink on the water cooling unit appears damaged, with some cracks—this happened when I first reapplied thermal paste a few months ago, and it was hard to remove. I anticipate this might slightly reduce cooling efficiency, but not cause major issues. Would you think that’s normal? If you notice anything else unusual in the photos, please let me know. If you have any other suggestions, I’d be open to considering them. Thanks so far!
I might run Cinebench r23 to test CPU performance. If it becomes unstable, the program could crash. Repeated crashes during testing would suggest a problem with the CPU or motherboard.
I've already run some prime95 stress tests, but couldn't reliably crash the PC consistently. Sometimes it stopped in seconds, other times it didn't. I think running it more steadily for a bit longer might help. I'll try cinebench next and update you on the outcomes. Thanks.
I just completed a cinebench test without issues, though the CPU kept getting very hot during the multi-core session. I plan to run prime95 for a bit longer once I have some time. I also observed that two cores were consistently warmer than the others. Is this due to how cinebench performs the stress test or could it indicate a cooling issue inside the CPU? The only crash I experienced after reinstalling the CPU was a freeze right after startup, but it hasn’t happened under load yet again. I’ll try running some games that usually cause crashes to see if the problem persists, as I mentioned before it’s hard to reproduce. Thanks for your support so far.
Just a quick update: I just had a nother CLOCK_WATCHDOG_TIMEOUT error direcly after startup when i was trying to log in to windows, which happened directly again at the restart after that, and only after that the PC started up as normal. So the problem definitely did not get solved after opening it up, although it seems like it did help for a few days. So i am open for further suggestions. Thanks a lot.
Core performance varies with temperature changes. This largely depends on the CPU design and core layout on the chip. However, inadequate contact from a faulty cooler mount or similar problems might contribute too. Excessive heat can lead to instability. I’d check the CPU pins and pads on both the board and the chip, remove any old thermal paste, and clean the cooler as well. It’s unlikely that temperature alone is the main issue, but debris or old paste inside the socket could be problematic.
It seems you haven't received the notification yet, and I didn’t have much time to address the issue. However, I did experience some Bluescreens again as described earlier, which suggests the problem remains unresolved. Something I recently recalled is that the game "Hades" tends to crash my PC when using DirectX, but not with Vulcan. This might be useful information. Thanks for your suggestions. As mentioned before, I’ve already taken steps—replacing thermal paste, checking everything, and reviewing the pictures—to confirm everything is fine. I’m still unsure about any new methods you’re proposing. Could you clarify what additional actions I should try that I haven’t done yet? Also, do you know a software method to eliminate the possibility of a faulty motherboard? If so, I might consider replacing my CPU on eBay and testing it. Appreciate your help, even though this is quite old.