Michapolys
Yeah,

I can see some issues with the way some EVGA implementations of the 30 series were engineered. It is not really just EVGA at fault here. More like a combination of things, with EVGA being the last wheel, thus being the one taking the blame here.

Unfortunately there does not seem to be a way to contract the engineering department, to let them know of what can be done to mitigate or even nullify the issues.

So the situation is what it is.

Would you mind going into more detail? What issues have you seen and what steps would you take to solve the issues?
 
I'd love to learn more about these issues before it's too late.

There is no doubt there are power delivery issues with RTX FTW3 series cards.....some fail...some dont.....but many are.

jacoffey85

Would you mind going into more detail? What issues have you seen and what steps would you take to solve the issues?
 
I'd love to learn more about these issues before it's too late.

The side-effects have been repeatedly reported. As for the causes, they are just a couple of things and they cannot be nullified by the end user.

Best you can do as a workaround is to disable any sort of power-saving setting on the GPU and the CPU. This will take care of any power related issues, be it reboots or burned GPUs, with an estimated success rate of 80%. It should work for most setups.

As for the details, they should be shared with EVGA engineers as only they can do large scale changes to their designs and share the associated design details. If you happen to find any of the people involved, send them over to discuss possible actions.

jacoffey85

Michapolys
Yeah,

I can see some issues with the way some EVGA implementations of the 30 series were engineered. It is not really just EVGA at fault here. More like a combination of things, with EVGA being the last wheel, thus being the one taking the blame here.

Unfortunately there does not seem to be a way to contract the engineering department, to let them know of what can be done to mitigate or even nullify the issues.

So the situation is what it is.

Would you mind going into more detail? What issues have you seen and what steps would you take to solve the issues?
 
I'd love to learn more about these issues before it's too late.

Its called "everything working great and Card OC's like a banshee ...for the first 3 months...then ..out of blue..you here a pop..and then the Card is dead, followed by a fixed red led over one of the 3 PCI-E power connectors" = obvious power delivery fuse that "popped"....= obvious power delivery problem at the pcb electrical engineering level....you want to "learn more" suggest a little googling n reading as there are many many threads out there......
 
If you can get your mitts on an Asus Strix I would strongly suggest gong that route for this generation...as I dont see the same repetitive failures with those Cards, although I dont hangout on ROG Strix GPU Forum....given the issues with the ROG Z690 Hero Boards ....looks like perhaps some of these QA faults could very well be related to the pandemic and lack of employees...(I saw even JayZ pointed that out).....makes sense.

rjbarker
If you can get your mitts on an Asus Strix I would strongly suggest gong that route for this generation...as I dont see the same repetitive failures with those Cards, although I dont hangout on ROG Strix GPU Forum....

I've had mine since December 2020.

schulmaster

rjbarker
If you can get your mitts on an Asus Strix I would strongly suggest gong that route for this generation...as I dont see the same repetitive failures with those Cards, although I dont hangout on ROG Strix GPU Forum....

I've had mine since December 2020.

Yeah unfortunately would have been my first choice this time around (always been an Evga GPU Fan Boy....Motherboards is Asus all the way).....but availability of the Asus Strix Cards was extremely impossible,,,,while the Evga FTW3 was impossible ;)

rjbarker
 
 
Its called "everything working great and Card OC's like a banshee ...for the first 3 months...then ..out of blue..you here a pop..and then the Card is dead, followed by a fixed red led over one of the 3 PCI-E power connectors" = obvious power delivery fuse that "popped"....= obvious power delivery problem at the pcb electrical engineering level....you want to "learn more" suggest a little googling n reading as there are many many threads out there......
 
If you can get your mitts on an Asus Strix I would strongly suggest gong that route for this generation...as I dont see the same repetitive failures with those Cards, although I dont hangout on ROG Strix GPU Forum....given the issues with the ROG Z690 Hero Boards ....looks like perhaps some of these QA faults could very well be related to the pandemic and lack of employees...(I saw even JayZ pointed that out).....makes sense.

I appreciate the response. As a fellow electrical engineer, I was wanting to see if the guy with 4 posts calling out engineering flaws was going to share the details behind his findings, or keep making generic posts. As expected, it was the latter.

i am disappointed evga chose to cap their XC3 at stock voltage, worse off then even FE. 
Please unlock 3080Ti XC3 to 390W. 

kim666
i am disappointed evga chose to cap their XC3 at stock voltage, worse off then even FE. 
Please unlock 3080Ti XC3 to 390W. 

In a thread discussing failures related to possible power-delivery issues, asking for a way to push the power-delivery components even harder seems ironic.

"EVGA
All of the cards were earlier production run cards manufactured in 2020. Under an X-ray analysis, they appear to have "poor workmanship" on soldering around the card's MOSFET circuits that powered the impacted cards."


https://www.techpowerup.c...bricked-rtx-3090-cards