[SOLVED] Need some help with my GPU

Mar 11, 2021
6
0
10
Specs:
Mobo: GIGABYTE GA-970A-D3
CPU: AMD FX-6300
RAM: 8GB DDR3 1600MHz
GPU: ASUS RADEON R9 290
PSU: Seasonic S12II-620 Bronze 620W


First things first, my PC started acting up a while ago, constant blue screens after turning on any video stream for the first time, after the BSOD it worked properly. After the seasonal dusting, this issues usually went away for a couple of months, then started to occur again.

Recently, the GPU flat out decided to give up, won't boot past the Windows 7 animation. During the bootup, I can briefly see the Motherboard image and it is slightly pixelated (more than it already is); no weird lines across the screen or anything.

I took the GPU out, vacuumed it a bit, as usual, then put it back in - nothing, same as above.

I borrowed another GPU from a friend, an R9 270, and it works properly, so I don't think it's a Motherboard issue.

I took my 290 out again, cleaned it properly with pressurized air from a compressor, cleaned off the excess burned out thermal paste with
isopropyl alcohol and applied a new blob then put it back in. It booted normally the first time, in 640x480 resolution (or whichever is when the driver isn't recognized) and it prompted me to restart after it installed something for the GPU. After that restart however, it went dead again. Same story with the pixelated motherboard screen, not booting past the Win7 animation, and a new weird oscillating sound from the case.

At this point, I have no idea what course of action to take anymore, any help would be appreciated.
 
Last edited:
Solution
I know and that is why I suspect the 12v rail is going on the PSU and cannot deliver consistent and stable voltage to the GPU and therefore you get critical system (kernal power) errors that cause the system to shutdown before damage is done. Can you go into the the Event viewer to see what errors are being reported

What PSU do you have, could be dying as that sounds like a coil whine to me. But the system is on the whole getting on so can you try the GPU in your friends system just to rule that the GPU is not gone. I would really suspect the PSU though as when more power is needed it give up the ghost and the gpu is being starved of clean 12v current leading to the bluescreens and instability.
 
Mar 11, 2021
6
0
10
What PSU do you have, could be dying as that sounds like a coil whine to me. But the system is on the whole getting on so can you try the GPU in your friends system just to rule that the GPU is not gone. I would really suspect the PSU though as when more power is needed it give up the ghost and the gpu is being starved of clean 12v current leading to the bluescreens and instability.

The PSU is a Seasonic S12II-620 Bronze 620W, bought it with the other parts back in Nov 2014.

Thing is with those bluescreens, after it BSOD once, it wouldn't do it again no matter what I watched/played, for the remainder of the day. After turning it on again the next day it's like it knows a day has passed and it hits me again with the BSOD. I honestly have no ideea, after the first BSOD, gaming doesn't affect it, a restart doesn't affect it, just waiting for the next day turn on to screw it up. Maybe a cooling issue? Fluctuating power? I'm at a loss.
 
Mar 11, 2021
6
0
10
Have you checked temps with HWinfo64, CPU and GPU?
I haven't recorded any temps as of late, but I don't really keep a heavy load on the PC, apart from the browser, if that counts.

As for testing on another rig, I don't really have anyone readily available to switch and plug a GPU with, but I agree that I should maybe try on another system to maybe rule out the PSU. The R9 270 that my friend gave me is powered only from a 6 pin connector, while the 290 is powered by a 6+4.
 
I know and that is why I suspect the 12v rail is going on the PSU and cannot deliver consistent and stable voltage to the GPU and therefore you get critical system (kernal power) errors that cause the system to shutdown before damage is done. Can you go into the the Event viewer to see what errors are being reported

 
Solution
Mar 11, 2021
6
0
10
I know and that is why I suspect the 12v rail is going on the PSU and cannot deliver consistent and stable voltage to the GPU and therefore you get critical system (kernal power) errors that cause the system to shutdown before damage is done. Can you go into the the Event viewer to see what errors are being reported


Sorry for replying so late, I swapped the 270 in place again to check the event viewer, and it doesn't seem to be anything of use. Just a couple of fatal alerts 40 and 70 one after the other (which are connection related? idk).

I noticed however, as I left the case open this time laying on the side, is that the 270 only has one fan spinning. I ran furmark for a minute or so to increase the stress, nothing spectacular happened. Didn't power off, didn't bluescreen.
 
Mar 11, 2021
6
0
10
So after a few weeks I finally got around to swapping the PSU with a new one, nothing really happened, still nothing after Win7 logo.

I added pic related that I mentioned before, these black dots that appeared weren't there before.

At this point I pretty much have no idea what else to do, other than to declare it dead.


t2FkYxS.jpg