PSU failure or GPU about to die?

cesaritox09

Distinguished
Mar 23, 2008
4
0
18,510
Hi everyone!

This is my setup:
Intel i5 4440@3.30GHz
Asus H81M-E Motherboard
Thermaltake TR2 TR-600 600W ATX12V v2.3
MSI Radeon RX 480 4GB Gaming X Twin Frozr Edition
8GB (4x2 DDR3 1600MHz)
2 HDD 7200 RPM (1TB + 3TB)

I have a question, when AMD released 18.6.1 driver, I proceed to update and for some reason artifacts started to showed up in screen. Decided to roll back to 18.3.4 but still it showed up some artifacts. It stayed like that for 3 days and after that, it didn't showed up anymore artifacts during 3 weeks. Yesterday I was playing The Division and alt+F4 the game and artifacts started to showed up again.

This morning I start the pc now the artifacts got a bit worse, instead of the lines I got now like a huge square that disappears/appears. I decided to shut it off and turn it back on and it has been doing well, no artifacts. One thing I noticed is that the artifacts show up when PC is idle, while I'm gaming they don't show up, which is weird. I even ran a stress test (Furmark) twice and everything was ok.

What I need to know is, if either the gpu or psu or any other component is going bad? PC has 5 years, while gpu has 3 years. A friend of mine says it could be the psu, since that psu has bad reviews and it could be causing some failures in the gpu. I'm attaching a picture of the artifacts that usually shows up. Sadly I don't have another gpu/psu to test. One thing to mention is that once the motherboard started acting weirdly, one day it only detected 4 GB out of 8GB; I moved both RAM to another slot and pc turn on but didn't even passed into bios, moved back and it worked ok. Three days later I clean up the pc and changed the RAMs to the slots where it didn't let boot the PC, but it booted fine.


Thanks for any help or advice.

iY5i4vJ.jpg
 
Solution
As someone said - down clock the GPU to see if they are still there, if not - then it still doesn't mean its GPU or PSU because less power is going to GPU to run it.
The only sure way is to put GPU in a friends computer and see how it runs for a couple of days, let him borrow it.

The PSU doesn't "cook" your GPU slowly, I really don't know where people get these ideas from. The only thing a bad GPU can do is give out noisy voltage and not constant - it doesn't cook, it isn't a microwave or a cooker.
It just normally makes something go pop instantly, it isn't roast pork.
Could be both actually.

The TR2 PSU is absolute garbage and likely has been slowly cooking the GPU over the past few years. I am surprised that the PC is actually still running at all.

I would replace the PSU with a quality unit 1st.


Something like:


Seasonic Focus Plus
Corsair RMX
EVGA G2 or G3


Then check and see how the GPU is after, but you need a good PSU to protect your components either way.


 

Rexper

Respectable
BANNED
Apr 12, 2017
2,132
2
2,510


"Gold certified" refers to efficiency only. Efficiency isn't the issue here.
 
As someone said - down clock the GPU to see if they are still there, if not - then it still doesn't mean its GPU or PSU because less power is going to GPU to run it.
The only sure way is to put GPU in a friends computer and see how it runs for a couple of days, let him borrow it.

The PSU doesn't "cook" your GPU slowly, I really don't know where people get these ideas from. The only thing a bad GPU can do is give out noisy voltage and not constant - it doesn't cook, it isn't a microwave or a cooker.
It just normally makes something go pop instantly, it isn't roast pork.
 
Solution