Blue screen with White stripes on Radeon HD 7950 (Possible VRAM issue?)

Dethan1999

Reputable
Oct 11, 2015
198
0
4,760
Hi there, I received a troublesome card from a friend to attempt to "repair".

It is an HD 7950.

Basically, the card seems works fine starting up Windows 10. It works fine when running on Idle Speeds. But as I put a more significant load on the card, my screen will turn blue and have white vertical lines going down. The white lines are hair-thin.

From what I've researched, I believe it may be either a VRM or Memory issue. The Temps are actually pretty decent. In the 30 range while idle, but I don't get to see the temps under load before the screen messes up. Part of me thinks that the issue happens when the Memory speed changes from Idle to load. But I'm not sure if that makes sense because I've been able to run stress tests, and it will actually work under load for roughly 30 seconds then the issue happens. So I'm not sure if the issue occurs EXACTLY when the Memory speeds increase.

Possibly a heating issue? I re-seated the cooler and applied more thermal paste. If I'm identifying the Memory chips correctly, it appears that they aren't in contact with any heat sinks (if I'm looking at the right thing).


Anyways, any help with this would be great. Also, I'm wondering if underclocking the Memory speeds will help. Maybe if I find the speed that screws it up, I can lower the speeds to prevent it.
 
Solution
It sounds like an overheating problem, if you're unable to monitor the temps before crashing, then here is one way to find out. Take the side panel off the case exposing the innards of the machine. Then you can try 2 things. 1. try placing your hand near or on the gpu to see if it's hot. 2. if it is overheating that's the issue, it could be a crappy cooling system from the manufacturer. If so disassemble the original and replace it with a new one. HOWEVER If it's a VRAM issue, you may need to get a new card entirely.

Justin Tullis

Commendable
Mar 11, 2016
39
0
1,560
It sounds like an overheating problem, if you're unable to monitor the temps before crashing, then here is one way to find out. Take the side panel off the case exposing the innards of the machine. Then you can try 2 things. 1. try placing your hand near or on the gpu to see if it's hot. 2. if it is overheating that's the issue, it could be a crappy cooling system from the manufacturer. If so disassemble the original and replace it with a new one. HOWEVER If it's a VRAM issue, you may need to get a new card entirely.
 
Solution

Dethan1999

Reputable
Oct 11, 2015
198
0
4,760


I agree, it sounds like a overheating issue, but I'm not too sure it is. The temps are fine idle. And it only takes about 15-30 seconds under load for it to happen. I re-seated the cooler and re applied thermal paste, still no change. My friend gave it to me because the manufacturer said its defective with no warranty. I think it's the VRAM. But I think it's only triggered when the speeds reach a certain point. I'm gunna try to Underclock the Memory Speeds to try to figure out what speeds triggers it.