Vram instability could trigger a reboot, but it's often unclear why. This was the most puzzling problem I encountered.
Vram instability could trigger a reboot, but it's often unclear why. This was the most puzzling problem I encountered.
Something similar to "clock_watchdog_timeout" makes sense. I might go ahead and purchase one now—your point is valid, and I also considered buying one before for backup. It feels like the time has come. Just keep in mind that I faced the same issue with a 50% power cap; the PC doesn’t shut down immediately, but after a few minutes of freezing (sometimes it just crashes without freezing), which could point to dirty power or PSU wear.
Run "MemtestG80" to verify VRAM. A ready Windows version is available: https://github.com/ihaque/memtestG80/blo...estG80.exe. Remember it supports up to 4 gigabytes of RAM, but a batch script has been created to start many instances for testing. You can find it here: http://pastebin.com/HCmYMwhM. The authors suggest limiting to 62 MB to avoid excessive CPU load. WARNING: I haven't tested it personally and can't give an opinion. If others know more about it, please share.
I just tested it, the standard setting. At -150mhz there were no issues in all tests except one where I saw around 200K. At -300mhz I got about 6k errors, and below 0 after 10-15 minutes. But running a benchmark at -500 still caused freezing.
It seems there might be some confusion about your hardware setup. You mentioned having two 16GB RAM sticks, but it looks like you're asking if the GPU is failing. Double-check your components to confirm the configuration.
Other random fixes are kicking in—ChkDisk, SFC/DISM, updating BIOS, repositioning CPU, inspecting rail voltages with HWMonitor, and reviewing temperatures. It’s like the system is trying to fix itself again.
Yes, two 2x16 boards have separate partitions with Windows (10 and 11), and the issue appears consistently on both. Temperatures and rail voltages look normal, according to HwInfo64, but the biggest discrepancy is a 3.285V reading on the 3.3V rail. The PCIe side shows voltages between 11.989V and 12.063V, though I can't verify them precisely at freezing because it freezes. The BIOS is up to date, and I've had the same problem before.
Sorry for the confusion, I purchased it via a company which gave me a 1-year warranty instead of the usual 2. It seems the issue began about a year and a month after purchase. If I had waited, I would have sent it to a repair shop instead. Thanks to everyone! I plan to test it on a friend's PC this weekend, so I’ll know for sure.