System powering down due to heavy usage.
System powering down due to heavy usage.
Details: AMD Ryzen 7 5800X with be-quiet cooler, 32GB RAM, 1TB SSD. Includes Corsair RAM and other components. Radeon RX 6600 XT, Corsair RM 750x80+Gold. Running Windows 10 64-bit with GIGABYTE B550M DS3H BIOS. No overclocking or modifications. Forum post for beginners; checking if my setup is correct and fixing the shutdown problem. Issue: PC powers off unexpectedly under heavy load—monitors cut out, USB devices disconnect abruptly. Seems sudden, not frequent. Possible causes: power supply or overheating. I recorded a crash log with CPU-Z and attached it. Online advice suggests power or thermal issues, but I haven’t found either. Any suggestions?
No answers available, just thoughts... your explanation seems to indicate the PSU is failing as it can't provide more power. It shouldn't be delivering less than before, but it is degrading over time. Have you checked for other PSUs? How long has this issue been occurring?
It has been happening for a couple of months now, beginning around January. The PSU isn't very old—about 1.5 years—and shouldn't be showing signs of wear at that stage, correct? It hasn't happened in other apps either; even during heavy VR sessions it hasn't occurred. When I attempt to recreate the shutdown through stress tests on Minecraft maps, it doesn't happen; it seems to occur randomly during gameplay.
I checked your logs... your GPU is getting close to 114°C before the log stops... that’s unusual. From what I found online about the RX 6600 XT, its thermal threshold is around 115°C... it likely reached its limit and shut down your system. VR might not be that intense compared to other high-resolution games since their resolutions are lower.
I've kept recording data and performed tests, discovering the hotspot hits 115 during new trials yet it remains active. I've tried DDU, reinstalled drivers, and adjusted the fan curve in the software. So far, stress testing hasn't triggered any crashes, though the problem's unpredictable nature means it's not fully resolved. I'm also unsure if using the 'hotspot' metric instead of the standard GPU temperature gives accurate readings, as driver-based stats match the regular GPU temp. Additionally, my VR headsets run at a higher resolution than my main monitor—playing games at 1080p, I suspect VR on ultra settings for simulations would be more taxing than standard VR flight modes.
I would disassemble the card and repaste it, and check if the pads need renewal too. A so called GPU service. It is less than 3 years but plenty use can cause dry paste too. unless you OC'ed it hotspot should stay below 80C with fresh paste and pads that fits.
99.9999999% confidence you have weak contact at least once on the die at one of the corners. Typical GPU temperature is normal around 71°C, but that 115°C hotspot isn’t. Even with an average of 71°C, it’s still a problem when the overall temperature is so low. After taking it apart, you’ll likely spot the issue yourself. Reapplying paste to a GPU is different from a CPU—make sure the adhesive covers every millimeter precisely. Don’t rely solely on the cooler’s pressure to fix it.
I will create a fresh thread for my concern Edited March 11, 2024 by BoneS71 I'm new and didn't intend to take over a discussion
It indicates your system stopped because the power was cut. You might attempt to turn it back on and switch the PSU power, or simply disconnect the power cable—both should trigger the same error message. I suspect your system sent a cut power signal to the PSU when the GPU hit its thermal limit. Since you've tried this multiple times and it reached 115°C without shutting down, it's unclear. Still, service the GPU just in case it shouldn't reach such temperatures.