[SOLVED] GPU or Motherboard Damaged?

[sundayz]

A few months ago my old GPU died while I was playing GTA V and I had to replace it with a new gpu. I've been having issues with my computer where it would randomly crash and the screen would go a solid colour and the pc becomes unresponsive and needs to be restarted through the power button. Very rarely, some "static" would flash on the screen too. I thought this was a PSU problem so I replaced the psu with a good, brand new one but the issue is still there.

The crashes would happen every time I started the PC in the morning, after it had been off for a few hours. It would get past windows loading, then 2 mins later it would crash no matter what. Eventually after 2-3 crashes it ran fine for hours with no problems. Yesterday it was really bad and the PC kept crashing, after like 6 or 7 crashes in a row I removed the GPU and I am using the integrated GPU with no problems.

The GPU ran fine at 100% load once it got going... so my question is, is this a faulty GPU that I have to return or could this be damage done to the motherboard from the previous GPU that fried itself? The motherboard is 8 years old at this stage.

My thoughts are that I should get a replacement for this GPU and build a new PC to cover both cases...

 

[cdrkf]

A few months ago my old GPU died while I was playing GTA V and I had to replace it with a new gpu. I've been having issues with my computer where it would randomly crash and the screen would go a solid colour and the pc becomes unresponsive and needs to be restarted through the power button. Very rarely, some "static" would flash on the screen too. I thought this was a PSU problem so I replaced the psu with a good, brand new one but the issue is still there.

The crashes would happen every time I started the PC in the morning, after it had been off for a few hours. It would get past windows loading, then 2 mins later it would crash no matter what. Eventually after 2-3 crashes it ran fine for hours with no problems. Yesterday it was really bad and the PC kept crashing, after like 6 or 7 crashes in a row I removed the GPU and I am using the integrated GPU with no problems.

The GPU ran fine at 100% load once it got going... so my question is, is this a faulty GPU that I have to return or could this be damage done to the motherboard from the previous GPU that fried itself? The motherboard is 8 years old at this stage.

My thoughts are that I should get a replacement for this GPU and build a new PC to cover both cases...

That is a tricky one as there are lots of possible causes. It's possible the motherboard could be damaged, also it might be worth checking the motherboard temps and maybe change the thermal compound on the motherboard chipset heatsink as at 8 years old that could have dried out which might cause problems.

It would be helpful to know some specs of the machine - also what gpu did you have and what have you moved to? If you have changed between brands (i.e. gone from an nVidia card to an AMD card or vice versa) you can run into conflicts with the old drivers that can cause problems.

Does your motherboard have more than one PCIe slot, even if the other slot isn't full x16? If it has a full size slot (even at x8) it might be worth trying to run the gpu in that slot, just to test if the main graphics slot has been damaged with the board failure.

 

[sundayz]

Specs:

• i3 2120 (Current CPU)
• Intel xeon 1220 v2 (Old CPU, still working fine, tried both)
• 2x4GB Kingston RAM + 1x8GB Corsair RAM DDR3 running at 1300mhz
• Old GPU: AMD Sapphire HD 7850 1GB (dead)
• New GPU: MSI Armor AMD RX 570 8GB
• PSU: Corsair RM 550x
• Motherboard: ASUS P8H77V-LE

I tried replacing the CPUs and I changed the thermal paste, both cpus are fine, problem persisted. Temperatures are OK.

I think I had the card running in the old PCIe slot. Yesterday I tried moving it to a different one, didn't help the problem at all. I think I have 2 more I could try just in case. Could the old GPU failing cause problems to multiple PCIe slots? The i3's integrated graphics work fine.

These were all the parts I had before my old GPU died (Aside from the PSU) and they were working fine together. Only things that have changed is the GPU and the fact that the old one had a bit of an unpleasant death. I dont think there were sparks but the PC just shut off and the old GPU would only work in safe mode and wouldn't install any drivers

 

[cdrkf]

Specs:

• i3 2120 (Current CPU)
• Intel xeon 1220 v2 (Old CPU, still working fine, tried both)
• 2x4GB Kingston RAM + 1x8GB Corsair RAM DDR3 running at 1300mhz
• Old GPU: AMD Sapphire HD 7850 1GB (dead)
• New GPU: MSI Armor AMD RX 570 8GB
• PSU: Corsair RM 550x
• Motherboard: ASUS P8H77V-LE

I tried replacing the CPUs and I changed the thermal paste, both cpus are fine, problem persisted. Temperatures are OK.

I think I had the card running in the old PCIe slot. Yesterday I tried moving it to a different one, didn't help the problem at all. I think I have 2 more I could try just in case. Could the old GPU failing cause problems to multiple PCIe slots? The i3's integrated graphics work fine.

These were all the parts I had before my old GPU died (Aside from the PSU) and they were working fine together. Only things that have changed is the GPU and the fact that the old one had a bit of an unpleasant death. I dont think there were sparks but the PC just shut off and the old GPU would only work in safe mode and wouldn't install any drivers

Hmm - so I gather you are using the i3 due to the graphics issue, as the Xeon is the better cpu of the two....?

With regard to the thermal paste - have you also checked the thermal paste on the motherboard chipset cooler (looking at images online, this is the heatsink that sits below the ram slots and above the PCIe slots)? It might not be the issue but worth a check.

Given the problems you are describing my thought is the motherboard may have been damaged, the fact you have tested two different cpu's out rules them out imo.

If possible it would be a good idea to test the new gpu in a different machine, my guess is it will work fine - in which case a new motherboard (or a full platform upgrade as you suggest) might be the best solution.

 

[sundayz]

Hmm - so I gather you are using the i3 due to the graphics issue, as the Xeon is the better cpu of the two....?

With regard to the thermal paste - have you also checked the thermal paste on the motherboard chipset cooler (looking at images online, this is the heatsink that sits below the ram slots and above the PCIe slots)? It might not be the issue but worth a check.

Given the problems you are describing my thought is the motherboard may have been damaged, the fact you have tested two different cpu's out rules them out imo.

If possible it would be a good idea to test the new gpu in a different machine, my guess is it will work fine - in which case a new motherboard (or a full platform upgrade as you suggest) might be the best solution.

Honestly I didnt even know I could take off the motherboard chipset heatsink. The temperatures for the motherboard are at 25-35 degrees celsius so I think it's maybe better to leave it be. It's been there for 8 years maybe taking it off will do more bad than good

Thanks for suggesting I try the gpu in another PC first. I'll check if it fits in the small PC I have downstairs but I don't know if the power supply in that thing can handle it. I might give it to my friend to test it out in his pc. Would be nice to avoid having to send this gpu back.

And you're right the xeon is the better CPU but it has no integrated graphics. The motherboard only supports DDR3 so it's probably time to get a new one anyway...