Weird random freeze

May 17, 2018
3
0
10
I’m having a weird freeze on my desktop pc every day. It can happen from anywhere between 5 minutes after booting, or 6 hours into gaming. The picture freezes and the sound disappears, and I have to hard reset.
In the beginning, I thought it was my RAM, because the freeze happened instantly after boot if I switched sockets on the DIMMs. But no errors come up after several cycles of memtest86, and I ended trying a different set of DIMMs and it still crashes.

Nothing overheats, and nothing comes up in event viewer except the hard reset.
I’m suspecting my mobo, but I’m not really certain.

Specs:
CPU: AMD FX 8350 4.0 Ghz (never OC’d), and it still crashes with AMD turbo core and cool n quiet off

GFX: MSI GTX 1070 Gaming X 8 Gb
Mobo: Gigabyte 970-UD3P
RAM: 16 gb (2x8) Kingston HyperX 2133 mhz DDR3.
PSU: Corsair RM650X.
OS: Win10 Pro 64 bit

Another weird thing is, that my RAM come up as single channel in speccy, but the other set I tried come up as dual channel (as they should).

I’m usually pretty tech savvy, but I can’t seem to figure this out.
I couldn’t find any other posts like this, sorry if they exist.

Thanks
 
Solution
First you boot the PC into safe mode to see the PC will freeze or not, if it does not, you know the problem is from hardware only, and it not from the software, like driver or OS. If the PC will freeze even in safe mode, and update all the MB driver, like the chipset https://support.amd.com/en-us/download/chipset?os=Windows%2010%20-%2064 even the BIOS, if the MB does not have newer one. Using DDU to uninstall the GPU driver, then reinstall the GPU driver again. And check the OS with system file check (SFC) scan: https://answers.microsoft.com/en-us/windows/forum/windows_10-update/system-file-check-sfc-scan-and-repair-system-files/bc609315-da1f-4775-812c-695b60477a93

If the PC runs fine in safe mode, next test the PC with minimum...
First you boot the PC into safe mode to see the PC will freeze or not, if it does not, you know the problem is from hardware only, and it not from the software, like driver or OS. If the PC will freeze even in safe mode, and update all the MB driver, like the chipset https://support.amd.com/en-us/download/chipset?os=Windows%2010%20-%2064 even the BIOS, if the MB does not have newer one. Using DDU to uninstall the GPU driver, then reinstall the GPU driver again. And check the OS with system file check (SFC) scan: https://answers.microsoft.com/en-us/windows/forum/windows_10-update/system-file-check-sfc-scan-and-repair-system-files/bc609315-da1f-4775-812c-695b60477a93

If the PC runs fine in safe mode, next test the PC with minimum hardware, because it is not the overheat problem, so use like only one stick RAM, only the boot device, other low power consumption GPU, also disable the cores in the BIOS too ( Note the MB is good enough for the fx8350, but who knows, because the freezing may or may not relate to the MB VRMs, which to power the CPU).

 
Solution
May 17, 2018
3
0
10
I already tried a full reinstall of Win10, and everything is up to date, including BIOS, so I don’t think it’s software.

I will try the low power setup. The VRM should definitely have enough power for the CPU, as besides the GPU and PSU the build is around 3 years old and I have had zero problems until now. But I have seen orher people have problems with this CPU with age, so I will definitely try to disable some cores.
Thanks for your answer.
 
WHen motherbaords start to 'head south', they can be very hard to narrow down what is precisely causing the occasional freeze.....

If your motherboard is well ventilated, RAM is known good, and temps are under control, an intermittnet motherboard failure is sure more likely than not, but, it won't hurt to borrow someone else's GPU or PSU, just to rule those out. (COmplicating the matter, some moterboards might work fine with one model GPU, but be unable to handle the rapid voltage/current requirements of a newer GPU that actually might use a few less watts, oddly enough....; and you'd tend to see that with more older AMD mainboards than on Intel boards, IMO)
 
May 17, 2018
3
0
10
I have another PSU I can try to switch in. I’ve had the newer GPU for a year if that has any significance, with no problems. I’ve run PRIME on the CPU and Furmark on the GPU for 5-6 hours with no overheating or freezing. So I guess it’s mobo or PSU?