Hi,
 
I recently installed the graphics card in question into my custom loop and had horrible performance. leak testing went fine, there were no leaks. However, when I tried to run anything the computer would crash. I'd be lucky if I could even get past the windows login to be honest. It was almost like it was stuck endlessly booting. When I did finally get past the windows log in, I either had sever lag when doing anything, a disappearing mouse, another reboot, or a blue screen of death. The gpu rarely performed well on benchmarks such as furmark and when viewed through task manager it only ever used less than 1GB of memory. I swapped components in and out and the only constant problem was the gpu. I called EVGA and they opened an RMA for the card. I sent it in and it got there Tuesday last week. I called today to get a status of the RMA and they told me there were no issues with the card and they'll be sending it back to me as is.
 
I'm surprised this was their response. For as hyped up and expensive as the card is it was incredibly underwhelming. I personally think it's unacceptable they're sending the same exact same card back but arguing with some tech support guy on the phone isn't going to solve anything. So I'm writing here to see if anyone has any ideas as to what my issues could be. I will say that I have 2 cables powering it and using both connections on one cable. I'm not opposed to trying 3 individual cables, but I'm putting that out there since that seems to be a common thing people suggest. Temperatures were fine as well while running the benchmarks. It's just whenever I boot my pc, the motherboard debug light for "vga" is always lit. according to the manual, that means gpu failure or not detected. I'm currently using a rtx 2070 with that very same motherboard.I was really excited to get this card and build my first custom loop (which in itself was a great experience) but this card is ruining it for me.
 
My build is below. Any help/ideas are appreciated.
 
motherboard: msi mpg x570 gaming pro carbon wifi
cpu: amd ryzen 5500x
gpu: EVGA GeForce RTX 3090 FTW3 ULTRA HYDRO COPPER GAMING
psu: cosair RMx 1000W

is there a documented case where they have admitted there was something wrong after saying themselves the device was fine? They were pretty adamant on the phone that their testing was the be all end all in this situation. they did say they'd do what they can over the phone to help when I receive the card again, but I don't know what they could possibly do over the phone to help. especially considering it's a custom loop that requires draining for testing other components.

are you on a riser cable by chance? just curiosity out of something you mentioned.

mmomega
are you on a riser cable by chance? just curiosity out of something you mentioned.

This might be a great question, are you using cable extensions for the PCI-E power connectors? On the set you daisy chained, did you chain port 1-2 or 2-3?

I would also look at possible reloading your windows operating system if possible.  DDU is a great option, but if the GPU came back from evga with no issues it would cause me to atleast look at the setup first.  I would talk to evga to see if they have notes on the rma to see what they did to determine the card was fine just to be comfortable with their response too.  Adding the third power separate power cable would be highly recommended as well.  
 
My thing would be after you try the third power cable and the DDU to remove the old drivers, to try and reload windows from BIOs and then go from there.  Make sure to back up your files first and be prepared to spend some time going through this process.  Sometimes a fresh load of windows is your best bet if everything else mechanically checks out.  

WaterCooledSloppyJoe
I will say that I have 2 cables powering it and using both connections on one cable.

This thread is a month old and was bumped by a spammer. Let it die.