Question Need help ruling out PCI-e slot or GPU itself


Oct 7, 2016
As the title suggests -and I know this has been asked before- I am having a problem figuring out if it's a bad PCI-e slot or a bad GPU. I have a MSI Armor RX570 4gb, AMD; and I have been using it normally for the past two years (bought it early 2020). I bought it with one year of usage, but owner said they applied fresh thermal paste, had it cleaned properly, etc., and I never had any problems with it. Everything ran well, all the games I wanted to play, I could; but two months ago, after some weeks that I replaced my old i5-4670T with a i7-4790K, it suddenly started to artifact. Very much so, that it happened very shortly after I turned on my PC, and after the artifacts, the screen just turned into a solid color (often gray, but sometimes brown) and I could hear the glitched audio behind, and turning off my PC by holding the power button was the only option. The card was in a 16x slot, so I moved it to the other 16x slot, but this one powered in 4x (since my card had 4 VRAM, I figured it wouldn't be much of a problem whilst I didn't get a new MB) and it worked flawlessly. I figured this would be because I removed my GPU several times to apply new thermal paste to the new CPU, since it was my first time and I was attempting to get the best thermal paste applied to keep it as cool as it could get. Anyway, I could run everything again (with some games that ran better before, now with some performance issues due to obvious reasons); until today.
The same thing happened again, but in a sudden manner: I opened Valheim, was loading up the game and the screen just went to the solid color, with the PC not wanting to boot up after. It constantly gave out the three RAM beeps, so I switched the RAM sticks several times, even tested with some old ones I have around; until I unplugged and plugged the GPU back in, and it worked again, to my surprise; except that this time, the GPU gave artifacts suddenly when playing Worms Armageddon, which is a low spec game. The PC froze, popped the solid brown color, and I had to restart it, but it wasn't booting up again. I switched it to the old 16x slot and ran a stress test using Heaven Benchmark 4.0, with all the settings popped up to max: and it had no issues. It ran without any artifacts or weird screen tear, but after I stopped the program and had the GPU clock go down and also the temperature, it started doing all the weird things on the screen that it did months ago. Now, the PC won't even start with the GPU in any PCI-e slot. If I keep unplugging and plugging the GPU back in, or twitching it in the slot a bit, it eventually turns on, but it either artifacts to a system crash, or it doesn't loads the AMD drivers.
Anyway, after all the text, I'd like to ask for help to some possible tests I could do to rule out the proper component. I currently have no other PCs I can test the GPU at or any other GPU I can test on my MB. I'm trying to talk to a friend to see if I can test the GPU at his rig but, well, to keep any personal matters out of this question: it's complicated.

Last but not least, here are my specs:

Windows 10 Pro
Intel Desktop Board DH87MC
i7-4790K 4Ghz
AMD RX570 4GB VRAM (MSI Armor)
16gb RAM (2x 8gb 1333mhz DDR3 (if I'm not mistaken))
2x 500GB HDDs
1x 3TB HDD

If there's any tests I can run to rule out the PCI-e slot or the GPU, I'd love to know. Here's hoping it's the MB, since GPUs are really expensive as of the date of this post; specially in Brazil, which is where I live. Thank you all for your time and sorry for the long text! I attempted to be as thorough as possible. Anyway,
I also say its GPU, My brother had a x99 board with a dead PCI-E slot and it was crashing the system and hanging on boot, but if the screen artifacts, its usually GPU related, and not the slot.

Also another question, What Power supply do you have, sometimes a weak PSU can cause weird issues like this, I don't see many OEM PSUs able to running that GPU reliably.