EVGATech_AdamB
Hello jeffyt! I couldn't find your RMA information so it must have been processed under a different member account than your forum account. I will send you an e-mail so I can work with you on this. All of our RMA replacements are tested before shipping out and also considering that there have been new-in-box replacements as well, it honestly doesn't sound like the failures are due to the graphics cards. After cycling through that many cards it is hard to imagine that there couldn't be another component at fault causing the failures. The power supply would be the most likely culprit and I would agree that you should wait until you get the new PSU before installing the card again. 

jasoncodispoti
Reddit can be a harsh place... I hope that my posts (eliteassissin07) did not come across as being such. From the information that you shared on Reddit and now here, I stand by my original assessment that the power supply is to blame. 
 
You mentioned that initially the system was shutting down under heavy load and you had to use two separate PCIe power cables to resolve this. What was occurring in this situation, as you noted, was that you were overloading the +12v rail the reason this was resolved with the usage of two separate cables is that this distributed the load across both rails instead of just a single rail. The Corsair RM750i is a 750w power supply... the 750w is applicable to the +12v rail. Per Corsair documentation on this power supply when running in single rail mode you end up with a single +12v rail that can deliver a max load of 62.5A's and 750w's. I am assuming (due to it not being documented) that when running in multi rail mode it runs with two +12v rails. With two +12v rails that means that each rail is only able to deliver a max load of 31.25A's and 325w's. When running low power devices this is fine and it acts like a breaker in your house by not allowing the devices connected to any one single rail to consume more than 31.25A's or 325w's of power. However if you connect something like a EVGA RTX 2080ti FTW3 Ultra for example to this power supply running in multi rail mode you have to be sure to use two separate PCIe power cables one from each rail as a single rail will not be able to deliver sufficient power. All of that too say that I am not really sure that this power supply is going to be large enough to supply sufficient clean power to your system especially when running in multi rail mode. For the purpose of this conversation I am going to assume that you are using a EVGA RTX FTW3 Ultra with a power limit of 373w. Even with two separate cables powering the GPU at full load you will be putting about 187w on each of the two rails leaving only about 138w of head room. This IMO is really not enough head room on each rail as you have to take into account that other devices in your system will also be feeding off of the +12v rail. You could switch the power supply over to single rail mode, but even than you are cutting things close on power... Your power supply is 750w I would leave 15% unused (think of this as extra head room for spikes etc) so just factoring in 15% head room (112w) and the GPU at 373w that only leaves 265w for everything else in your PC which depending on your build may be enough, but I think its going to be close. 
 
The other reason I suggested the power supply as being the culprit is because I also tired running an EVGA RTX 2080ti FTW3 Ultra HC off of a Corsair HXi750 and technically as long as I did not load the CPU up it SHOULD have been sufficient for the GPU, but for some reason it just didnt seem like it was up to the task... I didnt have any crashes, but I had observed odd behavior with the GPU such as the LEDs dimming (on the GPU) for a couple of seconds when a load was applied and the voltage on the +12v rail dropping slightly for a couple of seconds and than normalizing. I suspect two things were at play here I dont believe the cables on the Corsair HXi750 are of sufficient quality to deal with the load of an RTX 2080ti FTW3 Ultra and I suspect that these power supplies are not actually able to output at the advertised values and when you are already tight on wattage this can spell trouble. I ended up selling my Corsair power supply to a friend of mine who is using it without issue to power his RTX 2070 and purchased a single rail EVGA Super Nova. Unfortunately I believe that the quality of the Corsair power supplies especially those at the higher output limits suffer from some quality control issues which appears to be reflected in recent reviews, go check out the reviews for the Corsair 1200w AXI series... 
 
I wish you the best of luck and completely understand how frustrating things like this can be, but I have confidence in EVGA support to be able to help you. Oh and do your self a favor sell that Corsair power supply and get your self an EVGA Super Nova =). 

jasoncodispoti
Did EVGA support suggest power supply issue as well? 

Hi, just wanted to follow up with a little more information. I really want to make sure my other hardware is not at fault, especially the PSU. I already monitored the BIOS 12v readings like Robert suggested but I also ran a benchmark for about an hour. I have attached the results as an attachment/picture with multiple graphs that OCCT generated and I combined.
The test was running using DirectX 9 instead of DirectX 11. DirectX11 would just stop after 2 seconds. This is after using DDU to remove everything while in Safe Mode, disabling internet, installing the newest nvidia drivers. Division 2 uses both DX11 and DX12 but both seem to crash after a few minutes of loading. It was around 95% gpu usage, 250 fps but limited to 30% power limit and -500 core/memory. The card was running at about 500-550mhz gpu clock and 6500mhz memory clock during the test.
The first card just had simple fan issues but the second and third are similar.
A very important discovery I made was when I dropped both core and memory clock -504, set Power limit to 30% and launched Borderlands 2. Everything was good for about 20-30 seconds. I got in game, ran around and then I started to get very strange flickering. VERY similar to what had been happening with my first replacement card, the same one I sent in a week or so ago, just a lot more intense. (more flickering and while the GPU is severely downclocked) I almost feel like that card was at the very beginning stages of getting to this stage, like some sort of hardware is failing on these cards but the EVGA tech told me that they found nothing wrong with the card I sent back. 
Another detail is when restarting the computer. I am now at the point where when I reboot my computer, it will start artifacting the moment it hits the Login screen. It'll take a few seconds of a glitchy screen and then it will be back to normal. I am assuming it is when the tool is started and the -500 downclock/30% power limit is applied. If I reset the downclock to default or 0, so no boost or downclock, it will instantly start artifacting and black screen. Sometimes I will need to replug the HDMI cable to get image, but Remote apps on my phone can access the computer too. After a short while, the system just completely lags out and needs to be restarted. Sometimes, if the last setting was 0 instead of the downclock at -500, it will take 4-5 reboots until I can actually open up the tool and set the downclock back to -500 before it starts artifacting, freezing and crashing the system.
The card now NEEDS to be at -500mhz on both core and memory clock to actual display the desktop... and this all started when it suddenly started artifacting during the top 1/8 of the screen while playing The Division 2 and crashed. After that, it wouldn't even boot back up until I applied the downclock which it would get in game, but crash after a few minutes while on all LOW settings. So it seems like it's degrading quickly too.
So I am really sorry for the huge amount of text and understand that everything really seems like it's just the GPU failing/artifacting but again, I just want to make sure it's not my system causing any issues. I really don't want this happening again.
Are there any other possibilities that I need to look at? Any other applications I can try using to stress test to ensure nothing in my system is causing these issues?
Thank you.
 -----------------------------------------------
Thank you for the long description of the issue and all the troubleshooting involved. Although the power supply does show a steady voltage, it's possible it may be unstable underload but the only way you can test this is by using a multimeter tool to get an accurate real-time reading other than that, the graphics card is is more likely to be at fault especially if you are able to test it in another system and the issue persists.

Once you get the replacement I would highly recommend to test the card in a different PC before installing it in your main rig.

I'm sorry to hear about the continued trouble with your replacement card. It looks like your most recent card did fail our testing with artifacting. The RMA before that passed our testing, and the first RMA failed testing (fan noise). I'm assuming the issue is card related from your previous troubleshooting, so I've setup another Advanced RMA so we can get you a fully functioning replacement ASAP.

bcavnaugh
@ "One thing I never really understood, why should I try it in another PC?"
Because it could be in general a bad computer, Hardware, Software, Drivers, OS.
So this is why members are asked to test the card in another computer.

GTXJackBauer

jeffyt
 
Also it's not your standard flickering. It was a very strange broken up image. Like if I was on my desktop or chrome, mostly when I was just typing on forums/reddit, my desktop wallpaper image would appear for .00001 second and disappear. If I was in game, whatever was on screen would just flash again for .00001 seconds and disappear. All very randomly. It wasn't just a black screen flicker, but some very weird broken/scattered image flicker. It was present on the second card as well, for around 6 months, got a bit more frequent, but EVGA never found anything wrong with it... maybe I should have just dealt with the flickering :D

Sounds like you were experiencing screen tearing since I assume your TV isn't Gsync capable as majority aren't.  Always sync your FPS to your screen's Hz aka refresh rate. 
 
A few things come to mind. 
 
1. Bad PSU
2. Not so clean power from the outlet/bad weather?
3. On a bad surge protector or a simulated Sinewave UPS
4. High Temps overtime will degrade components/bad case flow/high ambient temps
5. Loose connections or bad grounds
 
It's VERY highly unlikely it's the GPU after the 2nd time tbh.  Very rare would you get 2 bad GPUs back to back, let alone 3-4+.  Something else is going on and usually testing out parts with back up ones helps investigating this but understand not everyone has extra parts laying around. 

A new PSU is worth a shot tbh.  My 1st choice that I prefer is Seasonic Prime PSUs, EVGA/Corsair are all considered top notch and ironically I've had issues in the past with a AX1200i but all have been fixed since and still trucking along just fine.

Give us a run down of your PC system, screen included and what it's connected to from case to wall to better assist.

Vlada011
I can't say... RTX2080Ti have more problems then any previous generation. 
I will not avoid strong words to spare emotion of people who invest money... I understand when someone buy hardware don't like to hear that he made mistake and that line is worse... because people with limited budget didn't bought Turing. That was product from first day for guys who don't care. Not only for price, they not care how much performance they got. Now reality is cruel, only K|NGp|N performance and difference compare to GTX1080Ti are worth... but his price is not even close to price of RTX2080Ti... But if you check score that's performance you expected from reference model. When guys OC K|NGP|N or FTW3.
customers with limited budget who save money for new GPU looked in silence what happen with high end graphic cards prices and bought second hand Pascal more then ever.
 
I doubt this problem have anything with EVGA.
You have luck because you bought their card to get new one in box.
Because with others you would probably debate weeks with their support.
 
I don't know why some brands have support and force you to write every single information about PC when technical support give you advice and it's obvious that they don't know to recognize difference between motherboards.
I send question to ASUS how to disable CPU Fan Error if PC Boot and fan is not to CPU Fan header. I installed ASUS Fan Extension card and didn't want anything on CPU Fan Header.
Wasting of time. Question of forum, wasting of time as well. At the end I found among many option in BIOS because it's not like on older ROG Motherboards and resolve problem. After 2 days arrive answer from ASUS to set setting I don't have in BIOS. Obviously that technical support didn't look motherboard type at all.
My first opinion was... God, imagine RMA. OK I have different procedure in East Europe, but people worldwide, in USA example...
On twitter ROG site there is some question to give opinion about ASUS ROG and advice.
My advice was Buy ASUS ROG on place where you can back and replace very fast, if work paradise, if not hell. No middle.

Attached Image(s)