Game crashes unexpectedly on the computer, RAM indicator stays lit, and the power supply needs to be switched.
Game crashes unexpectedly on the computer, RAM indicator stays lit, and the power supply needs to be switched.
In recent weeks, my computer would unexpectedly stop working, freezing completely without any BSOD or restart. The screen would freeze, the keyboard and mouse would turn off, and all connections would drop—lights stayed on while fans continued running. Pressing the reset button on the case or MOBO didn’t help, nor did holding the power button for a few seconds. The only way to restore it was to switch off the PSU switch and perform a hard restart.
Here’s what I own: an Optimus AMD waterblock, an EVGA waterblock for the KINGPIN GPU, Corsair tubing, Bykski fittings, and a Lian Li O11 dynamic reflection plate. My PC has been running smoothly for about nine months with Aida64 on a 7-inch monitor. Under load and gaming, CPU stays around 65°C, GPU at 50°C, and two water sensors around 31°C. M.2 drives are at roughly 45°C.
In BIOS I’ve made a few changes: enabled Resizable BAR, activated PBO, set CPU cooler to ignore, turned DOCP on for RAM to run at 3600, and I’m using the latest BIOS version, GPU drivers, and all updates are complete.
The first problem started a few weeks ago—initially everything worked fine, but after some time lag appeared during gaming. The OSD showed the CPU clock at 547MHz, while the power profile indicated high performance. I couldn’t get it back to normal (around 4.5GHz). After a restart, it improved, but then it would freeze again about a week later. Since then, it’s been stable except for random freezes.
The only indicator in MOBO is the DRAM light. I ran memory diagnostics and memtest86 overnight (six passes), with no errors found. Some online users suggested it might be a failing PSU or MBO, but I can’t replace them yet. Before buying another expensive part, I’m considering disassembling my rig to test it further.
Could this issue be related to RAM or the MOBO DRAM slots?
If the temperature is dropping quickly during the memory test and you're seeing a DRAM error on your motherboard, the issue probably lies with the RAM. Turn off DOCP, Resizable BAR, and PBO, then check again. Consider using 2x8GB modules instead of 4x8GB to determine if that resolves the problem.
I’m testing the 2x8 sticks again. Before, I reset the BIOS to default, kept DOCP, ReSizable BAR, and PBO off, but the issue persisted. Just hoping it’s related to RAM since I really don’t want to replace anything just to fix this.
I used 2x8 sticks but it kept happening. Ran each stick through memtest86 with three passes and saw no problems. Reinstalled Windows and updated all drivers to the latest versions. During the first run of the OCCT memory test, it worked again. I ordered a new RAM module and plan to try that next week when they arrive. Any other suggestions on what might be causing this? I’m considering replacing the motherboard and PSU just in case. I took apart my whole system today, cleaned everything thoroughly—only been together for about six months, so it didn’t seem bad. No visible dust or blockage, which is a plus.