Question 8-pin PCIe slot Broken on GPU

[Banr]

So, I was having some major instability problems with my computer before, and every time I would touch the PCIe Cable that goes from the PSU to the GPU, my computer's image would just freeze, and the OS would keep running, or sometimes, it would hard crash altogether and nothing would work.

At first, these hard crashes would happen once a day, and then, after a week my PC wouldn't boot anymore and I'd have a VGA warning on my MSI motherboard light.

I thought it might've been the PSU, so I bought another one, but the problem continued.

So now the problem must lie in the GPU, right?

Today, I submerged the 8-pin PCIe connector SLOT ON THE GPU with Isopropryl alcohol, and all of a sudden my PC boots again, and it seems to be stable. It ran for about 3 hours.

After this, the same stuff started happening. I reseated the 8 Pin PCIe cable, the computer is turning on, but it'll come back, like it always does.

It seems like the 8-pin PCIe connector is causing all this unstability. What do I do?

It's not the cable because I bought another PSU and tested with a brand new PCIe cable. It must be the GPU 8 pin slot.

 

[Lutfij]

Make and model of your old and new PSU? Make and model of your GPU? Submerging the GPU every time you encounter a lock up isn't exactly a long term solution.

I'm leaning on the fact that perhaps your PSU might be unable to deliver necessary power to your entire system when taxed after a certain amount of time. Either that or your GPU is having thermal issues.

On second thoughts, please list the specs to your build like so:
CPU:
CPU cooler:
Motherboard:
Ram:
SSD/HDD:
GPU:
PSU:
Chassis:
OS:
Monitor:
include the age of the PSU apart from it's make and model. BIOS version for your motherboard at this moment of time.

 

[Eximo]

1# Take some good close ups and post them here
2# List all of your hardware, can't really say what is worth doing or trying without knowing what you have.

I can't imagine it got dirty enough that cleaning would have helped, probably just a coincidence and you manipulating it made the difference.

If the PCB is cracked or there is a cold solder joint and getting the GPU warm causes it to disconnect, that I could see.

 

[Banr]

Make and model of your old and new PSU? Make and model of your GPU? Submerging the GPU every time you encounter a lock up isn't exactly a long term solution.

I'm leaning on the fact that perhaps your PSU might be unable to deliver necessary power to your entire system when taxed after a certain amount of time. Either that or your GPU is having thermal issues.

On second thoughts, please list the specs to your build like so:
CPU:
CPU cooler:
Motherboard:
Ram:
SSD/HDD:
GPU:
PSU:
Chassis:
OS:
Monitor:
include the age of the PSU apart from it's make and model. BIOS version for your motherboard at this moment of time.

Thanks a lot for the answer.

I didn't submerge the GPU, only filled the PCIe slot with Alcohol, but I agree it isn't a good solution.

My old PSU was a Riotoro 500W, my new one is a MSI 650W. My GPU is an AsRock RX 6600.

My GPU isn't having any thermal issues, as a matter of fact, these freezes happens even in BIOS.
Also, by my monitoring software the GPU Hotspot is at less than 60ºC.

I think it's hard that it's a Power issue because the new PSU just arrived today. I bought this one specifically because I thought it was a bad PCIe cable on the old PSU.

CPU: R5 5600
CPU cooler: AMD Stock
Motherboard: MSI B550
Ram: Kingston 16GB
SSD/HDD: Kingston NVME
GPU: RX 6600
PSU: MSI 650W
Chassis: Aerocool Cylon with everything removed for more airflow
OS: Linux Mint and Windows 10
Monitor: Philips 27 Inches don't really know the model

include the age of the PSU apart from it's make and model. BIOS version for your motherboard at this moment of time.

Age of PSU: 0 days olds

BIOS Version of motherboard: 2.C0

 

[Banr]

1# Take some good close ups and post them here
2# List all of your hardware, can't really say what is worth doing or trying without knowing what you have.

I can't imagine it got dirty enough that cleaning would have helped, probably just a coincidence and you manipulating it made the difference.

If the PCB is cracked or there is a cold solder joint and getting the GPU warm causes it to disconnect, that I could see.

Thanks a lot for the attempt to help me

Ok, I'll take some good pictures when it stops working again. I posted the hardware on the comment above.

It's not dirty at all, I checked before and after cleaning it.

Hmm it happens even in BIOS, and when it gets really bad the PC doesn't boot.

 

[Eximo]

I don't mean the chips are getting warm, the entire board heats up too. If there is something physically wrong with the power connector on the board, or your motherboard, that could be the problem.

 

[Banr]

I don't mean the chips are getting warm, the entire board heats up too. If there is something physically wrong with the power connector on the board, or your motherboard, that could be the problem.

But like, even in BIOS? Like after some time it starts crashing 3 seconds after getting into the desktop, then it crashes even in BIOS.

 

[Eximo]

Yes, say there is a small crack, or something else loose. Computer is off and it cools down makes contact. PC heats up and if the board is under any stress and it starts to warm up over ambient, it can change shape and lose connection, causing your crash.

One thing to do is a complete teardown. Uninstall all components, re-insert every cable, just to make sure it isn't something loose. Can breadboard the system outside of the case just to be sure before reassembly. If you identify any one component that makes the system unstable, then it is already apart and easier to replace.

Aside from something obvious, a flaky component can be just that. Might be bad video memory, or an internal short, anything.

 

[Banr]

Yes, say there is a small crack, or something else loose. Computer is off and it cools down makes contact. PC heats up and if the board is under any stress and it starts to warm up over ambient, it can change shape and lose connection, causing your crash.

One thing to do is a complete teardown. Uninstall all components, re-insert every cable, just to make sure it isn't something loose. Can breadboard the system outside of the case just to be sure before reassembly. If you identify any one component that makes the system unstable, then it is already apart and easier to replace.

Aside from something obvious, a flaky component can be just that. Might be bad video memory, or an internal short, anything.

Bro, you might be onto something. I have an Old PC here, with a FX 6300 and iGPU

U think its a good idea to test the GPU on this old pc to see whatsup?

 

[Eximo]

Absolutely.

 

[Banr]

Absolutely.

Just tested the GPU on another computer. The GPU works flawlessly there. I'm starting to think the PCIe on the MOBO is bad, what you say?

Should I RMA the MOBO?

 

[Eximo]

Try removing and re-inserting the CPU at the least. If that does nothing, then yes, probably something wrong with the board.

 

[Banr]

Try removing and re-inserting the CPU at the least. If that does nothing, then yes, probably something wrong with the board.

Did that, no avail. Something wrong with the MOBO, yeah? You say, to RMA right now or not?

 

[Eximo]

Assuming the motherboard is still covered, sure try an RMA. AM4 motherboards aren't exactly the most expensive things in the world if you have to get a new one.

 

[Banr]

Assuming the motherboard is still covered, sure try an RMA. AM4 motherboards aren't exactly the most expensive things in the world if you have to get a new one.

It is still covered. It has a 3 year warranty for all defects. I mean, my MOBO was expensive, over $100 so I'd love to keep it

Thanks a lot for the help bro!!!