Computer suddenly stopped working, CPU running at half speed.
Computer suddenly stopped working, CPU running at half speed.
Over the past months I've faced some major stability problems with my setup. I don't know the root cause and I can't replace any components to test it. Here are my details:
Processor: R9 3950x
RAM: 3200MHz Corsair LPX DDR4 4x32GB (total 128GB)
Storage: Asus ROG VIII Hero (WiFi) X570 GPU
Memory Card: MSI 2080 Ti SSD
Cooler: Corsair MP600 2TB NVMe
Case: Asus Ryujn 360 Watercooler
Operating System: Windows 11
Software: All updates current; BIOS is latest, moved from 2020 to 2023 but issue continues.
Symptoms: No matter what I do—idle tasks, browsing, typing—I experience random freezes. The machine stays on but the screen remains, and sometimes audio plays. RGB lighting on the keyboard turns off completely. I can't shut it down with the power button; I have to unplug or cut power.
I keep monitoring via Task Manager on a secondary display so I can observe what's happening during freezes. Stats look normal when it crashes, and I've checked temperatures with Ryzen Master. CPU temps hover around 37°C, GPU stays low (0-2%), RAM at about 9%. Recently, the CPU drops from 4.3GHz to 2.3GHz in seconds and then stays stuck near 500MHz. Restarting doesn't seem to fix it—power cycling is necessary.
Other oddities: When I restart, the PC sometimes powers off briefly before coming back online, or it restarts instantly without a second power-off. If I boot from cold, it may briefly activate before shutting down again and then boot normally. Sometimes everything freezes while the CPU runs at 500MHz.
Despite passing memory tests, stress tests, and firmware checks, I'm still unsure. I suspect something with the motherboard or power management, but I can't confirm. I've tried reinstalling Windows using the reset method and even considered a full reinstall if it's software-related. The Windows Event Viewer shows only an unexpected shutdown.
I'm worried this might be affecting my system's health long-term.
Others might take a different approach, but my first step would be to reset the BIOS to its default configuration. Check Task Manager again to look for any unusual activity such as a Trojan or virus. If the issues persist, I’d attempt to reinstall Windows on a new drive. Should that not resolve the matter, I’d consider hardware faults. Not everyone has additional motherboards or processors; it could be the motherboard itself, or possibly the processor. Based on my experience with overclocking a 2700x, this seems likely. I hope it gets fixed.
Incorrect bios configurations or installation problems can cause this issue. The CPU shouldn't be the reason; if it was, the system would stop posting and get stuck at postcode 00, effectively dead. Usually, it's either the board or the RAM that's faulty here. It seems like the board is likely the problem. Personally, go into the tweaker menu, set core ratio to 45 (4.5g allcore), adjust vcore to 1.3v, turn off LLC or whatever the lowest setting with the highest VDROOP has.
Ryzen Master displays the results when the CPU runs at 500MHz. Notice the core clock stands at 30C with a speed of 535MHz...
Visit C:\Windows\Minidump to see if any minidump files exist. If present, return to the Windows directory and transfer the entire Minidump folder to your Downloads folder (use the desktop if needed). Compress the copied folder and include it in a post. Please adhere strictly to instructions since Windows doesn't allow file manipulation there.
The only available data comes from a single dump recorded after a recent BSOD incident. It shows everything halted while music continued playing, lights stayed lit but no input responded. The crash occurred roughly a minute later, displaying a Watchdog violation—expected behavior. This doesn’t seem to be the main issue. The dump points to nvlddmkm.sys, tied to NVIDIA. If you spot anything else notable, please let me know. On another point, I’m testing an update that might have resolved the problem. I considered changes between stable and crashing states, which led me to an iCUE update from version 5.6. Since then, the 500MHz issue appears to stop, and my RGB Strip works fine again. Previously, three strips would freeze while one operated normally. After rolling back iCUE to v4.33, the freezing ceased, and the issue has been resolved. I tried an overclock via BIOS, setting CPU speed to 3.8, which is lower than Ryzen Master’s settings. It might relate to CPU speed or voltage levels—temperatures were normal. I’ll monitor the situation and report back if it remains stable for a week. 4-10-23-dump.txt
The dump file is a cumulative DPC crash report. No driver surpassed its limit, yet together they did. The issue remains with a single driver, making it difficult for the OS to pinpoint the exact cause by examining only active DPCs.
The Nvidia driver was the sole driver with unresolved DPCs at that moment, limiting possible solutions. I’d execute DDU followed by updating the GPU driver to test for improvements. Please share any additional dump files if available.
I believe the Nvidia drivers aren't responsible for the issue. The problem stemmed from the system freezing unexpectedly. I'll let the device operate as it is now and monitor for any further occurrences. If it happens again, I'll reinstall the drivers using DDU. Up until now, performance has improved compared to the past six months. Assuming stability continues, it likely points to iCUE causing the disruption.