Determine if your power supply unit is responsible for the crash during the stress test.
Determine if your power supply unit is responsible for the crash during the stress test.
Hello, welcome back to the forum! I just built my new PC with an Intel i5-13400F, Intel Arc A750, MSI Pro H610M-E board, and a 550W power supply. Windows 11 x64 is installed. To check stability, I ran AIDA64 stressing the CPU+FPU and Furmark testing the GPU. The system handled each test well. But when I tried to stress both at once, it crashed after about 30 seconds. I also noticed a “click” sound inside the PSU during startup—could be a switch mechanism. The monitors show the CPU drawing 80W and the GPU 200W, with a power meter indicating up to 420W total. What’s likely causing the sudden shutdown? Could it be a faulty PSU? How can I figure out the exact problem? Thanks for your advice!
It’s the Great Wall X5, though you might not be familiar with it because most of it is available in the Chinese market.
They often check RAM, cooling, or the motherboard before concluding. Look at temperatures and clock speeds first.
Click noise from the PSU triggers a protection circuit. It shows up when you can run just the GPU or CPU alone, but then shut down instantly when both are active—especially with the click. This strongly points to the PSU being the issue. If you have another PSU to test, that will confirm it.
I also agree with you. It’s unclear where to begin fixing things. I reached out to the seller and received a fresh PSU. This attempt shows the system can last about a minute under stress tests for both FPU and GPU before shutting down. Because I need to return the old unit quickly, I only swapped the PSU without checking the wiring. Now I’m feeling frustrated without clear next steps.
This setup seems different from what I expected. My current unit has a dedicated +12V 45A rail, and I’m considering swapping it for an Antec power supply. However, I stumbled upon something odd! When I set Furmark resolution to 1920x1080, the system ran smoothly during an eight-minute stress test on both FPU and GPU. But once I changed it back to the default 1280x720, it immediately shut down. I’m puzzled—why would resolution affect performance this way? The only logical guess is that lower resolutions cause more power swings for the GPU, possibly exceeding the PSU’s capacity. - If it doesn’t work with the new PSU, I’ll check the motherboard next.
Interesting. Does it fail while performing tasks you usually handle? If you need to apply certain synthetic loads to trigger a crash, I’d just skip that concern.