You might want to consider replacing your power supply unit.
You might want to consider replacing your power supply unit.
I forgot about this tool, but I'll keep an eye on it for next time I shut it down. Thanks.
evga psus often face problems with newer gpus, especially power spikes. That might be the cause. Consider trying a corsair rmx 850w. Your cpu temperatures look unusual—can’t help with that, but it seems something’s off with the loop or drive. Yes, that matches. However, it doesn’t fully clarify why idle temps differ from load temps.
Technically it's not water; it's a synthetic cooling fluid meant to be nonconductive. But if your loop runs GPU, CPU, RAD, and RAD together, that’s the core issue—directly transferring GPU heat to the CPU without enough capacity for the CPU to manage. I still believe the NZXT H510 case contributes to the problem—it's not designed for proper airflow, and the 5900X generates a lot of heat. Edited April 23, 2022 by An0maly_76 Updated, more info
You may want to determine the fluid's path. If the pump is moving in the opposite direction, it could be forcing res -- 240 -- gpu -- 120 -- CPU. This might make sense since it would chill the GPU and reduce performance on the 5900X. I believe you should consider a better ventilated case going forward. Edited April 23, 2022 by An0maly_76 Updated with more details
I might come across as lacking knowledge sometimes—this case isn't performing well, and the 5900X generates significantly more heat than the 2700X before it. Regarding the shutdowns, it could stem from weak capacitors in the PSU, or EVGA has had similar problems with theirs, which is possible for any component. Make sure the replacement parts match the dimensions, as sizes aren't interchangeable. I've used a Corsair RM850x in my previous builds.
There might be an obstruction in the connection between the CPU and GPU. The GPU handles its own cooling with fans, so if the coolant isn't moving heat efficiently from the CPU, that could be the issue. Are all your hoses the same size? Which one do you use?
The GPU no longer has fans, so I had to take them out before installing the waterblock. Any obstruction in the system could affect temperatures. However, if debris forms inside the CPU socket, it might lead to overheating. I’m not opposed to removing it for inspection, though it would be a bit of work—this part rarely exceeds 80°C and usually stays above 60°C. It seems more likely a power supply issue is causing the shutdowns rather than random temperature spikes. I have an extra CX430M ready; I’ll connect just the GPU to test later.
It could be your RAM that needs attention. I've noticed unstable RAM often leads to system crashes under heavy memory use. It might be the issue you're facing. Consider using memtest86 to check for problems.