Has anyone noticed a high failure rate in 30 series?

[0cean]

I have several hundred 30 series cards and almost all my Strix and 3080 FTW3 have this issue, they will be mining fine on my Asus Z390-P and as I am phasing that board out for the Biostar Expert/Pro 2.0 boards I find my GPU tend to “die”

They work completely fine, board swap, and then they just show 0 GPU

When I remove the GPU to my Windows test bench the GPU will boot but no longer post. I have to assume some driver issue is happening but how am I to fix these issues if the GPU shows up empty in the PCI Slot on my Windows machine?

[keaton_hiveon]

if the gpu isnt showing up in windows its not gonna be a driver issue if its happening on linux too. i have 100+ 30 series, mostly evga and FE though. most have been running for 1-1.5 years and havent had a single one fail aside from a few asus cards fans (3060 phoenix, 3060ti mini, and strix 3080)

[Mininggago]

Which OC have you been using?

[0cean]

Right which is why im trying to narrow my possible issues out here

[0cean]

Not really relevant here tbh

[keaton_hiveon]

Sounds like something is getting shorted out in the process of you removing and reinstalling them in the new system.

[0cean]

In your eyes what point of failure would cause the short? I unplug my server PSU before doing anything to a rig

[keaton_hiveon]

It’s tough to say without knowing the setup/process/conditions etc. How many have you had fail in this way? Have you sent any in for rma/ to see exactly what part failed?

[0cean]

EVGA and asus send back blank error reports for the RMAs

Id say 20 my last 100 3080tis have died

[keaton_hiveon]

That’s crazy high, gotta be something with your setup or process.