During a power loss, the CPU throttles itself and activates cooling protection, causing shutdown within moments.
During a power loss, the CPU throttles itself and activates cooling protection, causing shutdown within moments.
I experienced a power interruption today due to a kitchen stove activating the breakers. Following this, my PC instantly reduced its performance while navigating the UEFI BIOS, lowering the CPU clock and attempting to prevent overheating. It eventually settled at 550MHz and shut down once temperatures reached around 90°C. The system ran at approximately 1950x threadripper on an ASUS ROG Strix X399-E motherboard with a TR4 socket, paired with a Corsair Hydro cooler and a Corsair AZ1200 PSU. When I power-on, the CPU temperature rises steadily—about 1°C per second—and I can observe the thermal safeguards activating. The cooler fan spins noticeably, and the motherboard signals the CPU to throttle as temperatures climb. I feel the cooling unit operating, with audible movement in the tubes. At around 70°C, fans kick up more aggressively. Unfortunately, accessing Windows is delayed by boot time, and the throttling drastically slows down login processes before a shutdown occurs. I’m unsure what data would be useful; my PC skills are limited, so I captured UEFI images of any screens that might show relevant metrics. All recorded readings use the default settings, as I didn’t want to alter the system after its initial setup five years ago. I also recognize the RAM is underperforming—being an early adopter of Threadripper meant it wasn’t optimized, and the manufacturer suggested overclocking both RAM and CPU for better results, though I’m not sure if that was accurate.
Ensure you believe the CPU isn't overheating; locate the thermal limits setting in BIOS and turn it off. This should allow access to Windows for more troubleshooting.
I’m not sure about this setup. The CPU cooler feels fine, but the brackets are quite hot. I don’t have a clear idea of how warm they are. I’m concerned that turning off the thermal limits might accidentally harm the CPU, though I’m willing to hear any advice you have. Right now it seems hard to find a new motherboard or PSU nearby, so replacing them will likely take more time.
Notice the warmth of the radiator at the point where the lines connect to the tank. When the CPU is overheating and the temperature remains unchanging, the pump likely isn't working.
Great advice from @Kid.Lazer! I tried it and it stayed perfectly still—no temperature changes. It seems the movement in the tubes might have been just echoes. I’m considering switching to a new cooler as a first step, since this older socket is going to need a few days. I’ll go with something standard this time so I can check its performance later.
This cooler is clearly not functioning properly, with idle temperatures in the BIOS being excessively high. It seems a power surge may have triggered the failure or it was already malfunctioning. Simply inspecting the hoses and AIO components for warmth isn't sufficient to determine its condition. I recommend replacing the AIO without hesitation.
It's likely the pump motor continues to operate, which explains the vibrations. However, the pump seals might be damaged or another problem is stopping it from moving the fluid.
Thank you for your helpful ideas. It was actually quite challenging, but I understand the situation. The closest replacement option is a 7-day special order because the socket is a bit outdated. Since my vacation begins tomorrow and I won’t be able to test it until back home, I’ll keep you posted once I’m available. I’m sure this fix will work, so I’m confident in the solution. Have a wonderful summer everyone!