Device shutting down unexpectedly only during gameplay sessions.
Device shutting down unexpectedly only during gameplay sessions.
I assembled this system in December 2020. It operated flawlessly at first, with no problems I remember. Recently though, whenever a game launches, it opens smoothly, runs without lag or tearing, and the clock speeds remain stable. However, when the screens—connected via display ports—go dark, I’m shown the boot screen again (the one you see when you power on or restart).
Other problems appeared as crashes occurred. On my ROG STRIX BF450-F motherboard, the SATA ports ceased functioning except for two. Eventually, I discovered the issue: the latest BIOS update enabled fast boot and automatic CPU overclocking. Disabling those two ports restored functionality, but the reboot problem persisted.
I updated drivers for CPU, GPU, BIOS, chipsets, USB hubs, Bluetooth, and reinstalled Windows. I even tried reverting to a 2020 BIOS version and its drivers, but that didn’t help. All drivers and BIOS are now up to date.
In the Windows Event Viewer, errors cleared except for EVENT 86 and EVENT 46. EVENT 86 indicated an SCEP certificate enrollment failure via a specific URL, while EVENT 46 warned the system to stop powering down unexpectedly.
After extensive testing with Cinebench, Heaven, and MemTest5, I concluded:
1) Cinebench showed stable performance with minor CPU temperature changes—CPU appears okay.
2) MemTest5 confirmed no RAM issues.
3) Heaven Benchmark revealed choppy video, high GPU utilization, and crashes after 18 minutes of playback.
I’ve taken steps to fix everything: cleared errors, updated all components, and even rolled back to an older BIOS version. Despite these efforts, the rebooting issue remains unresolved. I’m still trying to figure out what caused it.
It might be a PSU or GPU; attempt to track power usage using an overlay from AMD. If it increases and the system stops, likely the PSU is failing. Also, check the AMD Pro driver—resolved lag and unexpected shutdowns.
AMD Pro Edition will be installed, hardware chosen, and drivers reinstalled. Next, I'll run the overlay with Heaven to check for power usage increases. Appreciate the advice. Starting now.
To ensure a smooth update, it's wise to use Display Driver Uninstaller to remove outdated drivers prior to installing fresh ones.
GPU Update: Installed new driver via uninstaller to remove old display drivers. Updated AMD Pro for fresh GPU, chipset, and related drivers. Launched Heaven without overclocking. Video initially lagged, similar to instance 3 from the original post. Power usage remained low while GPU performance stayed stable. Over time, GPU consumption rose significantly—reaching up to 100% usage, clock speed hovering around 1284–1287MHz, VRAM clock at 1750MHz. Power draw jumped from 22W to 103–110W. Currently, the Heaven video stalls on a transition frame, no motion detected, but GPU continues running at those speeds. No crashes observed yet...but that’s not surprising.
After a while, the Heaven video played without issues, just as expected. I left it running while I got water and came back to see the "Tada" login screen—meaning it crashed again. There were no changes in power or clock speeds once the graphics card started working, but I’m not sure. Do you all think the problem is with the GPU, or could it be a BIOS issue that needs more power or less?
I don’t have an extra GPU available right now…but the board does have an HDMI port. Would you try using that instead? Check if it’s still causing issues. I’ve already updated the BIOS, and I even rolled back several updates to the 2020 setup from when I first assembled the PC. Nothing has been stable so far.