Question Once stable system burning out multiple video cards

DavidRexGlenn

Reputable
Jun 10, 2019
11
1
4,515
Even though I am more or less a noob when it comes to building PCs, I built a tower two years ago.

Motherboard: Mag B460M Bazooka (MS-7C83)
CPU: Intel i-5-10400 2.90 GHz
PSU: EVGA 500 BA 100-BA-0500-K1 500 W ATX12V / EPS12V SLI CrossFire 80 PLUS BRONZE Certified Non-Modular Active PFC
Original Video Card: GIGABYTE GeForce GTX 1650 SUPER WINDFORCE OC 4GB
Replacement Video Card: MSI GeForce GTX 1660 SUPER 6GB

No problems until I decided to jump from Ubuntu 20.04 to 22.04 this past August. Decided to go the fresh install route, created a USB boot drive with installation image. Found out that because I set the BIOS to Fast Boot I could not access the BIOS to set the Boot Drive order with a USB keyboard. Because I am older and don't see well enough to find the jumper, I pulled the video card to access the battery for a BIOS reset. Reinstalled video card, installed new OS, everything worked great for about 3 weeks until the video card crapped out.

Ordered a new card similar to the one I previously had. Installed card and everything was great again. Now three weeks later, this new card has bit the dust. I mainly use the system to stream movies and such as well as an hour or two of FPS games via Steam. It's not like I am rendering videos 24/7. According to the BIOS the system temperature hovers around 86 F. Nothing is overclocked. I do my best to make sure the dust doesn't build up and choke the system.

Before I return this card for a replacement, and possibly burn another one out, I thought I would try to determine why this is happening. Strange that both cards only lasted 3-weeks. The fans on both cards will spin up when power is first applied but then go silent after the switch is made to the CPU's GPU

Could I have damaged the contacts of the port where the card plugs in when I reseated the original card after I reinstalled the battery?
Could this be a video driver issue related to Ubuntu 22.04? I am using the same NVIDIA drivers as I did with 20.04. I've searched for such a correlation but have not come across anything.

Any assistance in troubleshooting this issue appreciated
 
Last edited:

DavidRexGlenn

Reputable
Jun 10, 2019
11
1
4,515
hmm i think there were some video from it from linus mentioning this issue with zero RPM GPUs, but basicaly whats happening is this:
low GPU usage = zero fan speed
combined with no monitoring of GPU internals like VRMs alongside with high case ambient temperature (low gpu, high CPU usage scenario) = GPU dyin due to overheating

edit: it was gamer nexus, video here
View: https://youtu.be/8eSzDVavC-U?t=221
Thanks, this gives me an idea as to how to prevent this in the future