Random PC crash without BSOD. Asus GPU lights up three LEDs.
Random PC crash without BSOD. Asus GPU lights up three LEDs.
Hello. I'm not very experienced with computers, but I can help with troubleshooting and updating what's needed. To begin, things have become more problematic. I first noticed the problem after switching from an older Ryzen to a newer generation. At first, I let everything run normally for a day, ran benchmarks, and everything seemed okay. Then I turned on EXPO, and it worked smoothly for a month—no errors or crashes. I always keep my software updated and drivers properly installed. However, over time, I started experiencing random restarts without any BSOD. It just powers off and restarts as if the power was cut. Initially, I tried restarting again and trying to recreate the issue, but sometimes it doesn't happen. Usually, when I play games, they crash. At first, it happened with demanding titles like NFS Heat or Battlefield V, but modern games like Modern Warfare 2 work well. I also play lighter games such as Timberborn and Dota 2 without issues. Recently, even Modern Warfare crashes completely. Even games like Dota and Timberborn stop working after a few weeks. Not sure if this is connected, but my 3080 Ti lights up three LED blinks on the PCIe cable. I've searched online and found it's due to insufficient power. My Thor 1200W PSU seems sufficient—my ASUS 2080 Ti OC and 3950X run fine even with overclocking, and my old PC handled 550W before. In my opinion, the 3080 Ti and 7950X should work with a 1200W unit. I've asked others, researched, but no solution has worked. I've tried reverting to stock mode, changing RAM settings, disabling global C-state in BIOS, updating drivers, undercoating the GPU, adjusting power curves, and even using different cables. None helped. When games cause crashes, I shut down the PC, unplug it, wait 10 seconds, then restart. It stops the blinking lights and lets me boot again. But whenever a game starts, the problem returns. I updated drivers manually from NVIDIA or the manufacturer site—no luck. I changed the BIOS to x670-i to the latest version from Asus, but it didn't fix anything. I tried undervolting with PBO AMD Eco and thought performance loss was worth it for fewer crashes. I also considered GPU issues because of power LED flashes when it fails. I attempted undercoating the GPU with Afterburner, lowering the curve to below 1900, and raised it back to 1900 after selecting 925. It didn't resolve the issue. Still, the GPU lights up three LEDs. When I see those lights, I shut down the PC, unplug it, wait ten seconds, then try again. It restarts, boots, but games always cause the same crash. I updated the M.2 BIOS to 1616 and thought that might help, but it didn't work. The RAM issue was resolved by using a different stick, but the problem remains. I tried multiple RAM configurations, but nothing changed. I suspect the CPU might be the culprit—recent power levels exceeding 1.4V could trigger instability. I attempted undervolting with AMD Eco and thought it would prevent crashes, but performance dropped instead. I also considered the PSU, thinking maybe insufficient power was the cause. Older parts like the 2080 Ti and 3950X still worked fine. I have an extra 1200W PSU, but it didn't solve the problem. RAM seemed fine since the GTX Trident was set correctly. The issue persists even with different games and settings. Last night things got worse. Now I can't do anything—PC crashes during Windows boot, loading screen, or even in a browser. I tried reinstalling Windows after updating the BIOS, but it's still broken. I'm stuck. I've tried everything: different cables, power sources, BIOS versions, RAM changes, and even PSU upgrades. The drive died after a crash, so I'm worried about hardware failure. I have important files, medical records, schoolwork, and personal stuff stored. Everything I needed was lost when the drive failed. I'm in Cali near Fresno Hanford. If anyone can help, please—anyone willing to pay or offer pizza would be appreciated. I'm desperate. I'm willing to drive 4 hours if needed. I can't afford new parts because of bills and a family loss. My PC specs are: - AMD Ryzen 7950x - ASUS Strix X670-i ITX - ASUS Rog 3080 Ti OC 12GB - G.SKILL Trident Z5 Neo AMD EXPO 32GB F5-6000J3636F16GX2-TZ5NR - ASUS Rog Thor 1200W Platinum - Samsung 980 Pro 1TB - Samsung G9 Odyssey - 1500VA UPS.
Have you checked any sensors before restarting? (overheating/temperature might be overlooked.) Your undervolting effort likely helped a bit, but it probably didn’t fully resolve the issue—especially with this situation in mind. It’s worth reviewing the situation carefully. Most likely, the power supply isn’t delivering consistent voltage. This could explain the GPU indicators and might cause shutdowns if the board’s safeguards aren’t functioning properly. The PSU’s rating doesn’t matter much if it can’t maintain steady output; otherwise, things may become unstable. I’d recommend testing the PSU or comparing behavior with another unit.
Hi, just wanted to share my experience. I used another PSU with the same 1200W rating and it performed just as well. Temperatures were also within acceptable ranges. I'm using a Ryujin 360 for the CPU and have seven extra fans set up in a push-pull configuration to maintain good airflow. My GPU stays under 70% even at full load, which is much better than the 95% I saw before. It seems the issue might be a brief spike somewhere, but I don’t have a way to capture it. As a result, I can't run Windows for more than a minute. 20230517_004209.mp4
I recall the igpu was active since it appeared in dxdiag, showing the display adapter and most of the apps I use. It seems you can disable them? How would you do that? I’m curious about this. Earlier I remember the event viewer displaying a common NVIDIA error with letters like nvdklms, and some said AMD drivers might be causing issues. But I’ve never had it before—just one time. Was it happening on the same day?
Attempting to recall the steps I took feels confusing. I believed I turned off something in Device Manager, but the check doesn’t confirm that. It looks like the issue might be in the Asus BIOS settings. [Search suggests it could relate to: Advanced/NB Configuration?] The BIOS seems more suitable for future adjustments or corrections. I haven’t experienced problems with both Nvidia and AMD drivers, yet I’m unsure if I spent much time figuring out what caused the problem... assigning blame to Windows and disabling the iGPU ended up being the right choice.