Second GPU failing ASUS Maximus Hero XII

[dparisi]

I'm trying to setup a second GPU in my system.

Specs:

CPU: Intel Core i9-10850K
Motherboard: Asus Maximus XII Formula
RAM: Corsair 8GB 3600 x4
PSU: EVGA P3 1200W
PCIEX1_1: Internal USB3 port header x 2
PCIEX16_1: EVGA 3080Ti FTW3 Hydrocopper
PCIEX1_2: Empty
PCIEX16_2: Asus Strix 2080ti EKWB block
PCIEX1_3: Empty
PCIEX16_3: Empty
M.2_1: Samsung SSD 970 Evo Plus
M.2_2: Empty
M.2_3: Samsung SSD 980

Both graphics cards have two separate cables coming from the power supply going to the cards.

I previously had the SSD in M.2_2, but I moved it to M.2_3 as it was not detecting the 2080ti at all.
Now when I have a clear cmos and start the computer, the two GPUs detect and function.

As soon as the 2080ti is under load.. mining or longer 2nd display connection, it eventually flickers on and off 3 times than ceases to show up anywhere. (Device manager, BIOS, etc...)

If I restart the system, it does not show back up.
If I kill the power to the power supply when the computer is off, and then give it some time, it will show up again.

Could this be an issue with how I installed the water block to the 2080ti or is this more likely a faulty mobo?

Given that I have a hard tube custom loop (my first one at that!) I have some reservation from constantly tearing it down to move cards or storage around as it involves draining the loop.

ANY insight would be greatly appreciated. I ran the HWIN64 and GPUZ during a crash if that would help?

 

[EarthDog]

Only way to tell if it was a bad block install is to put the air cooler on and see...

 

[freakdiablo]

Yeah, check card temps - the part about needing to let it sit after load makes me think the card is overheating or close to it. Either of those programs should work.

 

[dparisi]

Sooo the stock air cooler wont fit in the case if I am running both cards with the lower rad. The 2080ti was mining and running fine when previously installed (by itself) with the air cooler...

The temps were fine when it crashed. Nothing high (31.4 C temp and 45 C hot spot)

Maybe I remove the power from the 3080ti and run the 2080ti and see if it is still crashing under load. IF this does not crash... I pull it out of the computer and test it in my friends computer with an air cooler re-installed, running at the same time as his GPU?

I was looking for any significant changes at the same time as the 2080ti crashing. Funny the W on the 3080ti jumped up from ~20W -> ~100W as soon as the 2080ti crashed.

 

[EarthDog]

You've got a 1.2KW PSU... unless it's failing, that isn't it.

 

[dparisi]

Re-did the thermal paste and tried to make sure everything was seated correctly.
Upon re-installation, the card worked for 11 mins and 30 seconds under load and then disappeared from being detected. I am at a loss for what is going on....

I guess drain the whole system AGAIN... and maybe use a vertical mount riser I have kicking around and re-mount the air cooler?

 

[Johan45]

Have you tried a different BIOS?

 

[dparisi]

Have you tried a different BIOS?

I am on the most recent BIOS. Try an older version?

EDIT: Version 2403 (most recent does not detect)

Reverted back to version 2301... no change.

Could this be a hardware issue with the motherboard??? It is still under warranty and don't know how to rule anything else out.

 

[Johan45]

I am on the most recent BIOS. Try an older version?

EDIT: Version 2403 (most recent does not detect)

Reverted back to version 2301... no change.

Could this be a hardware issue with the motherboard??? It is still under warranty and don't know how to rule anything else out.

I'm not certain that it is HW related. The GPU works, it's not overheating from what we know but just disappears from windows/BIOS/UEFI. The screen flicker reminds me of a driver crash but it's only affecting the one GPU.
If you have a spare/ separate drive you could install Windows and drivers etc on to, just the minimum, it would help to rule out a SW issue.

 

[Johan45]

One other thing I thought of is trying the PC in safe mode which won't load any drivers. Check to see if both GPUs appear in the device manager. You can get to the windows troubleshooting page by holding the shift key and restarting the system then advanced startup options

 

[dparisi]

Oh man... Air cooler back on, using a riser to run the GPU hanging off the front of the computer. Been running stable for a day now.... So its safe to say that it has something to do with the install of the EK water block?
Maybe put fresh thermal pads on as I had tried re-doing the thermal paste...?