A few months ago I bought an Acer Predator Orion 5000 P05-650. The system comes with a Windows 11 Home OEM which is worthless to me so I installed my retail Windows 10 Professional on it. Whenever I'm playing a particularly performance-intensive game, there's a random chance that the entire PC freezes and you need to hold down the power button to turn it off and on again. There is no bluescreen. No trace of an error is left in Event Viewer. No minidumps are recorded. Reliability monitor just lists the power being cut off since that's the only way to reset the PC after the freeze. The freezing isn't even related to how much is going on in a game graphics-wise. It seems to be entirely random. Also I manually set the fans to maximum via PredatorSense whenever I'm playing a demanding game to keep it cool.
I got some third-party advice telling me that I could try disabling Windows 10's ability to automatically acquire device drivers. I did, after doing a complete fresh reinstall of Windows 10, and I painstakingly installed every hardware driver manually by downloading them from Acer's website using my PC's serial number. That actually helped a bit, because it seems that the PC's operating temperatures went down somewhat, and the freezing happens less often now. But less often is still not never, and it's still unknown what is actually causing it. The only warning that remains in Device Manager is: Resource Hub Proxy device - This device is currently waiting on another device or set of devices to start. (Code 51) Dependencies: PCI\VEN_8086&DEV_7A4C&SUBSYS_164E1025&REV_11\3&11583659&0&A8. Device Manager is claiming that it already has up to date drivers and nothing can be done about that warning. Also there's this event in Event Viewer regarding the WiFi adapter that comes up daily. "The description for Event ID 7010 from source Netwtw10 cannot be found." I don't use WiFi. I only use wired Ethernet.
I've done all of the normal diagnostic procedures, checking for corrupted system files, etc. I've tested the GPU with Furmark and there were no problems. I've tested the RAM with Windows Memory Diagnostic and there were no problems. Same with Memtest86. The only test program that I've managed to use to replicate the freezing issue outside gaming is Prime95 CPU stress test, specifically the "Large FFTs maximum power usage" test, and even that test can usually run for quite a while before it freezes. So, could it be a power supply issue?
Today for the first time the PC actually froze and rebooted in a way that looked like a GPU fault, because the colors on the second screen went crazy and the computer actually managed to record a minidump before rebooting instead of freezing, even though it still didn't manage to get to bluescreen. The memory dump shows the error "ntoskrnl.exe"
I have not flashed the BIOS but the only update I can see on their website for this PC is "Update AMI Kernel to CRB028(BKC WW46-2022)". There's still over a year of warranty left for this PC so I don't really want to risk bricking the machine in case the problem can be solved in another way.
This is the first time anything like this has ever happened with a PC of mine, and also the first time I've ever bought a pre-built machine.