[SOLVED] GPU - Broken, Disconnected, or Other Problem

[sb1132]

Hello community!

I bought a Sapphire Pulse RX 580 8GB last year, and it held up extremely well. Last night I was playing some video games with friends, and all was going fine. I went to bed, and this morning I woke up to an unresponsive PC. Since my CPU has integrated graphics, I used that to boot up my computer and use it normally, but I just couldn't figure out why my GPU suddenly stopped working overnight.

What I've tried:

• The obvious; unplugging/plugging in the GPU (both into the motherboard PCIe and the power connector), restarting the PC
• Updating my AMD Radeon drivers to the latest release
• Updating my BIOS to the latest stable release (I couldn't get the absolute latest as my CPU is Raven Ridge, and on the website it advised against downloading the latest BIOS update if your build includes a Raven Ridge CPU)
• Checking the voltage of the PSU-to-GPU power connector (with a multimeter; every set of pins has a current of 12v, save the ground pins)

Some unique problems:
- earlier I checked on device manager (while connected to the iGPU) and when I enabled "show hidden devices" it listed my RX 580 under the display adapters. But now it doesn't, and I can't seem to get my build to register it. I've tried rescanning and manually adding a "legacy device" to no avail.

So the ultimate question: is my GPU broken? Is there any way I can test to see whether my GPU is functional, given that because of quarantine I can't test my GPU on a friend's PC? Is there, in general, a solution to my problem? I just don't understand how it can cease to function overnight...

 

[DSzymborski]

It's hard to test without another PC, really. I'd probably go ahead and start an RMA just to be safe.

What PSU is testing this? If you have a low-quality PSU, it's more likely the GPU is dead than if you had a higher-quality one. Though that's not a substitute for testing.

 

[sb1132]

It's hard to test without another PC, really. I'd probably go ahead and start an RMA just to be safe.

What PSU is testing this? If you have a low-quality PSU, it's more likely the GPU is dead than if you had a higher-quality one. Though that's not a substitute for testing.

I bought the product March 31 2019, so it's been over a year though. My PSU is a Corsair CX450M. Would that have contributed to the problem?

 

[DSzymborski]

I bought the product March 31 2019, so it's been over a year though. My PSU is a Corsair CX450M. Would that have contributed to the problem?

Most GPUs have three-year warranties, at least two at a minimum.

The PSU isn't suspicious. I'd have been more comfortable with a 550W, but the PSU is a good budget PSU unlike some of the junk we see in similar questiosn.

 

[sb1132]

Most GPUs have three-year warranties, at least two at a minimum.

The PSU isn't suspicious. I'd have been more comfortable with a 550W, but the PSU is a good budget PSU unlike some of the junk we see in similar questiosn.

That's great news. One last question before starting an RMA -- is there any possibility of a broken PCIe slot? I have two on my motherboard but I can only use one for my GPU because of size constraints (small case). So is there a way to determine whether the PCIe slot is still functional? It'd be a waste to go through all the trouble of getting a new GPU to realize that the PCIe slot was the culprit all along...