This is mostly an FYI, as I have been in touch with EVGA and now waiting for an "RMA approval".
Been building custom Water Cooled Systems for many years now and only had an issue with one other Evga Card (GTX 680 Classy), many years ago.
 
Have had my 3080Ti FTW3 Ultra under water since August (used a new Version 2 EK Vector Block). Has been rock solid with great OC'ing results and gaming (mostly CP77).
 
- 14 Nov: Day before yesterday had the first "hard crash" in that time, as in frozen screen and buzzing sound (I know most are familiar with this.
- Didnt think much of it as I also OC my 9900K so did go into my BIOS and made a few tweaks and all good, continued gaming for a few hours and shut down at the end of the evening.
- 15 Nov: Started up my PC and saw my first "No Post", as in Maximus XI-E Board with Post Code 97 w White LED light on (indicating GPU error). Cycle powered PSU and was able to get past the Post Code 97 and able to game for 3 hrs. During the gaming I periodically exited the game and tried to repeat the 97 code by going through a number of Cold Boots and Restarts (could not repeat the 97 Code).
- 16 Nov: Went to start up PC, once again Post Code 97 w White LED light on my MXI-E Board. So Cycle powered as previous day, however, this time when I powered up:
 - PC started up and immediately heard a small "pop" from within the GPU and AX1600i Mobo went to shut down mode (red LED on PSU).
- Cycle powered to reset PSU, although PC (Mobo) would go thru an entire post process right to "AO Code" (which is all normal for MXI-E), no video signal to monitor.
 
What I tried (knowing the "pop" I heard means likely dead):
- tried all video inputs on GPU, no video signal.
- tried different vBIOS switch on GPU, still no video signal.
 
I ran an HDMI cable from monitor to Maximus XI-Extreme iGPU output and all good, PC works fine. (glad the GPU didnt take out any other HW, Corsair AX1600i safety feature obviously worked as it should).
 
Anyhow, a drag for sure. As Im waiting for RMA Approval (which I would fully expect), I'll remove the Water Block and have a good look at the PCB (take pics if I see obvious damage) reinstall air cooler (test in another rig I have on hand), get it ready to ship.
 
Should be no issue with the EK Block, as it is a version 2 Vector model (modified stand offs vs version 1) and Ive installed many EK Water Blocks going back to GTX 680's.
 
On a side note, as I had Evga on the phone, I wanted to do a cross ship, unfortunately as the Card was purchased by one of my crew (I work on a CG ship), its considered 2nd hand so no cross shipping available (I had no idea this was the case).
The day I got the Card last Aug I had my crew member purchase the Card for me (while he was in town off the ship), he came back to the ship, gave me the Card (of course I paid him right away).
I did have to go through Evga via email to properly Register the Card as the original Invoice had my crew members name on it, but was all good.
 
Anyhow, was hoping they would have approved the RMA over the phone and not have to wait another 2-3 days for approval by email? Is this normal? I explained everything over the phone in detail.
 
Will keep anyone interested up to date.
 
Cheers
 
EDIT: I recieved an "RMA Approval" from Evga a few hours after my phone call to Customer Support.
Same Day as above: Loop drained, Water Block removed (nothing looked obviously wrong) and no sign of a short or anything looking fried.
Stock Cooler n thermal pads re-assembled, used Kryonaut paste and will test in another system later this evening:
Test System:
Asus Rampage IV Extreme
Intel 3930K (Stock)
16G Corsair Dominators (2133 Mhz)
Corsair AX 1200i
The above system is what I tested the Card on for several days on stock cooler back in Aug (was all good).
 

nick_shl
Since you have to mount original cooling system for RMA, have you check card itself for any damage signs?

Yes of course...already drained my loop, Card is out and Block taken off.
I dont see anything obvious at all. All looks normal, but Im not sure if a capacitor "popped" whether there would be visual indication.
I just heard heard the pop, but no smoke or smell etc.
 
After I re-install the stock cooler I will try it out in another rig (older Rampage IV Extreme / 3930K build).
 
RMA has already been approved, but from Vancouver Canada to Brea Cal and back, would expect 3 weeks out of commiss.
 

I think you'd have to check the fuses/components with a meter.  As I recall the card that buildzoid fixed didn't have any visual damage either.  Just a bad vrm stage and blown fuse.  I guess it makes sense that it was the middle plug that failed as it pulls the most power.

ObscureEmpyre
Ouch! Hoping my water-cooled 3080 Ti doesn’t crap out on me. I have other GPUs I can use if it does, but that means modifying my loop temporarily. Anyway, if EVGA informs you what the problem is, please share. A dead card with no visible signs of damage is quite intriguing.

Yeah no kidding.....was like ****...no wayy!!
As said guys, as a "trail leading up to it", this is after almost 4 months of no issues (granted light use as I work on ship on a 4 week rotation):
- Day 1 I had 2 hard crashes playing CP77 (frozen screen n loud buzz), but that can also be related to unstable CPU....so made tweaks in BIOS and gamed for a few hrs w no issues.
- Day 2 Started PC greeted to "no display" and Code 97, was able to get past this by cycle powering, then all good for a 3 hrs gaming (even ran a bunch of benchmarks).
- Day 3 started PC and no display and Code 97. Cycled powered and didnt get past Code 97.
After 2nd power cycle is when about 2 Sec into the boot process I heard the "pop" and my PSU went into "safety shut down mode".
 
And yeah, no obvious signs of failure on the Board after removing Block / Backplate.
 
Would be great if I got some feedback from Evga once they get the Card, but I kinda think maybe they dont. They likely throw it in a pile for the service guys and send out a replacement.
As far as the loop goes, Im set up pretty good for draining n removing Card, have done so many Block swaps even that is pretty quick :)
 
I do have an Evga 1080Ti SC Black and an EK Block I could throw in the system, but also have a 12900K and Asus Z690 Apex Board showing up next week, so not really sure if Im going to strip completely down or wait or? Problem is I also cant find any DDR5 !!!

ty_ger07
Sounds like a VRM MOSFET stage failed and blew that PCI-E plug fuse. It's been intermittently happening to quite a few folks. If the fuse works as intended, there shouldn't be obvious damage.

That would be the "pop" I heard ...the fuse? w no visible damage.
 
I guess the question is "why" ? Likely a $M question.
I dont see those Fixed Red LED's while a Water Block is installed (but do have the Board Post Codes and White LED next to the Post Codes that identifies faulty GPU.
Of course once the stock cooler re-installed I could see the fixed red LED above #2PCI-E power plug.
 
I'm curious, is there a fixed Red LED for each of the 3 PCI-E power plugs on the GPU PCB?
 
As a side note, the thermal pads on the pcb and the EK backplate looked normal, all the normal indentations I would expect to see.
 
I'm always about "why"...
For 24/7 gaming I had the Card clocked at +120 / +675 w no added voltage (wasnt required)
The Card was capable of up to +180 / +1175 for benchmarks only w a little added voltage.
 
For what its worth running GPU-Z while gaming w 24/7 OC's the Card would draw around 400w......if I ran Kumbuster would get up to 445w..
Max on PCI-E_1 Slot 56w
PCI-E power #3 was always around 75w - 80w, substantially lower than #1 and #2, but that seems common with mnay cards including my previous non Ti 3080 FTW3..

Oh here we go on the LEDs. Just happened to stumble on to this post just now.

https://www.reddit.com/r/...p&utm_source=share

Bummer dude! Really sorry to hear it!

kevinc313
Bummer dude! Really sorry to hear it!

No problem ....it is what it is right?
Funny both my Asus Z690 Apex, 12900K will be here tomorrow, was thinking all I need now is the DDR5....
 
Anyhow 1080Ti back in ....leak testing now.