- Dec 3, 2020
- 4
- 0
- 10
Hello there,
For some months now, my PC has had a problem where it would first start to bluescreen, with WHEA_UNCORRECTABLE_ERROR, so heres what I have done:
Reseated RAM
Tried new RAM
Replaced old HDD / Removed HDD temporarily
Replaced old 750W PSU with a 1,300W PSU
Tried default PSU cables (was using extensions)
Reinstalled windows 3 times onto my M.2 SSD
Reset windows 10 times
After resetting / reinstalling, not installing anything to confirm it was a hardware problem
Ran SSD benchmarks
Increasing the power limit for my GPU
After doing some of the above, my PC just started crashing. No bluescreen. However I am convinced there is still a hardware problem so I:
Reinstalled the CPU, put new thermal paste on. Arctic Liquid Freezer 2 reads 30C Ambient on my 3700X, 60C while gaming.
Tried turning on and off OC elements in BIOS, cleared CMOS multiple times, ran PC at default values..
I looked for bloated capacitors on the motherboard, since there could be a problem with power delivery to the components.
Took apart 5700XT and replaced thermal paste with thermal grizzly kryonaut
But heres where it gets really interesting:
Suspecting my 5700XT GPU was the problem, I replaced it using a low TDP GeForce GT 710. After installing that, my PC runs absolutely fine. No crashes whatsoever...so I conclude
that the GPU was the problem the entire time. Tested in games and casual work for around 4 months constantly, no crashes. In the meanwhile I manage to cop a 6800XT (a little upgrade)
whilst planning to RMA the seemingly faulty 5700XT.
I install the 6800XT, DDU the old drivers and install the latest drivers from AMD. Then my PC starts crashing AGAIN. I nearly pull all my hair out. Heres my theory which I need some advice on:
The problem was not the GPU at all, but rather the motherboard dealing with power to each of the components, mainly the GPU, which has the largest power requirement in the system. The reason why it didn't
crash beforehand was maybe becaseu the GT 710 doesn't require as much power as the two others, and therefore didn't yield any issues. Once a GPU with a larger power draw was placed in, the motherboard
couldn't supply the power it needed, causing a crash. Another theory could be the utilisation of the PCI-E 4.0 lanes via the GPU and the M.2 SSD, I am yet to try a new SSD or disable PCI-E 4.0 but I don't
think that's the case here.
My system specs are as follows:
CPU: Ryzen 7 3700X
RAM: 32GB 3600MHz Corsair Vengeance RGB RAM
GPU: Aorus RX 6800XT 16GB OC
MOBO: MSI MPG X570 Gaming Edge Wifi
CPU Cooler: Arctic Liquid Freezer 2 280MM, Ambient 30C gaming 60-65C
SSD: PCIE 4.0 Samsung 980 EVO 1TB, contains OS and applications on the OS
HDD: 3.5 inch 4TB Western Digital Black 7.2k RPM, contains games
PSU: EVGA Supernova 1,300W G2, seperate 8pin+8pin channels routing to the GPU
If anybody could suggest something, my last resorts are to swap the CPU or motherboard but I would like to know everyone else's theories beforehand. Thank you.
For some months now, my PC has had a problem where it would first start to bluescreen, with WHEA_UNCORRECTABLE_ERROR, so heres what I have done:
Reseated RAM
Tried new RAM
Replaced old HDD / Removed HDD temporarily
Replaced old 750W PSU with a 1,300W PSU
Tried default PSU cables (was using extensions)
Reinstalled windows 3 times onto my M.2 SSD
Reset windows 10 times
After resetting / reinstalling, not installing anything to confirm it was a hardware problem
Ran SSD benchmarks
Increasing the power limit for my GPU
After doing some of the above, my PC just started crashing. No bluescreen. However I am convinced there is still a hardware problem so I:
Reinstalled the CPU, put new thermal paste on. Arctic Liquid Freezer 2 reads 30C Ambient on my 3700X, 60C while gaming.
Tried turning on and off OC elements in BIOS, cleared CMOS multiple times, ran PC at default values..
I looked for bloated capacitors on the motherboard, since there could be a problem with power delivery to the components.
Took apart 5700XT and replaced thermal paste with thermal grizzly kryonaut
But heres where it gets really interesting:
Suspecting my 5700XT GPU was the problem, I replaced it using a low TDP GeForce GT 710. After installing that, my PC runs absolutely fine. No crashes whatsoever...so I conclude
that the GPU was the problem the entire time. Tested in games and casual work for around 4 months constantly, no crashes. In the meanwhile I manage to cop a 6800XT (a little upgrade)
whilst planning to RMA the seemingly faulty 5700XT.
I install the 6800XT, DDU the old drivers and install the latest drivers from AMD. Then my PC starts crashing AGAIN. I nearly pull all my hair out. Heres my theory which I need some advice on:
The problem was not the GPU at all, but rather the motherboard dealing with power to each of the components, mainly the GPU, which has the largest power requirement in the system. The reason why it didn't
crash beforehand was maybe becaseu the GT 710 doesn't require as much power as the two others, and therefore didn't yield any issues. Once a GPU with a larger power draw was placed in, the motherboard
couldn't supply the power it needed, causing a crash. Another theory could be the utilisation of the PCI-E 4.0 lanes via the GPU and the M.2 SSD, I am yet to try a new SSD or disable PCI-E 4.0 but I don't
think that's the case here.
My system specs are as follows:
CPU: Ryzen 7 3700X
RAM: 32GB 3600MHz Corsair Vengeance RGB RAM
GPU: Aorus RX 6800XT 16GB OC
MOBO: MSI MPG X570 Gaming Edge Wifi
CPU Cooler: Arctic Liquid Freezer 2 280MM, Ambient 30C gaming 60-65C
SSD: PCIE 4.0 Samsung 980 EVO 1TB, contains OS and applications on the OS
HDD: 3.5 inch 4TB Western Digital Black 7.2k RPM, contains games
PSU: EVGA Supernova 1,300W G2, seperate 8pin+8pin channels routing to the GPU
If anybody could suggest something, my last resorts are to swap the CPU or motherboard but I would like to know everyone else's theories beforehand. Thank you.
Facebook
Twitter
Instagram