I've had my EVGA 3090 FTW3 Ultra installed for about 3 days now, and ive noticed about 8 or 9 times, while at desktop, in battle net launcher, or browsing, that the screen will go black for a few seconds, the fans will spin up on the card, and then it will return. It gives an error event id 4101 - "Display driver nvlddmkm stopped responding and has successfully recovered." in the event viewer. Also, when it happens, it glitches out the battle.net launcher app. It will make it show nothing until i force close and reopen the app. It's a fresh install of the latest updated version of windows 10. I wasn't having any issues previously with my EVGA Rtx 2080ti.
 
Should i RMA the card?

I am having the exact same issue on my 2060 Super SC Ultra, battle.net also has to be restarted after the driver crashes and recovers. I get the issue when videos or gifs autoplay on Reddit or Twitter (tested on Microsoft Edge and Google Chrome), or when content has to load on the Windows 10 Mail app, however it doesn't always happen, it's very random.
 
I did actually send it in for RMA but I quickly returned it to myself after I tested on my EVGA 970 and had the exact same issue, followed by some blue screen of death caused by nvlddmkm.sys as well, because I didn't want to replace a new card that I bought recently (around August), and the bsod issue occurring on my EVGA GTX 970 was enough to tell me that the 2060 Super wasn't the issue.
 
My configuration is i7-4790k, 16gb ram, Windows 10 20H2 and I'm using driver version 456.71. But this issue previousl yoccured on much older drivers, and older versions of windows 10 such as 1903 and 1909 and 2004.
 
Unfortunately I don't really have a solution, but I have been troubleshooting this for the last  few months as the issue is very inconsistent and occurs randomly so it's very difficult to reproduce for me personally.
 
Things I've previously tried are reformatting my pc and reinstalling Windows 10, and using Display Driver Uninstaller.
 
It never happened on my 970 before either, until I upgraded to 2060 Super, just like how you didn't have the issue before. But unfortunately I did start to get the issue on my 970 after swapping it back to test.
 
It has to be a software or driver issue though...
 
Edit:
On nvidia's game ready driver community thread regarding the latest driver, it appears to be an acknowledged issue, you can see this in the stickied post
GeForce 457.30 Game Ready Driver Feedback Thread (Released 11/9/20)
 Some users may experience random TDR while using a web browser [3171517]
 
It seems like this has to be related to our problem. I know it's a report for the newest driver, but there's no reason it can't be related to older drivers since I was having precisely the same issue.

What are your Nvidia power settings at?
 
There is a bug in the current driver that causes intermittent, random black screens if you have the card set to "Optimal Power" or even "Adaptive".
 
Set the card to Prefer Maximum Performance and see what happens.

Hitting this issue as well on an EVGA RTX 3090 FTW3 Ultra. Usually when I exit a game or something I'm hitting this.
This also seems to be occurring on Version: 457.51 (release 12/02/2020)
My thermals are around 68C - 73C during gaming. Idle temps are 28C - 33C.
Went to NVIDA control panel and going to set the Power management mode to "Prefer maximum performance".
Will update the thread if I see this issue after changing the settings.
 
 
Steps incase anyone is too lazy to google:
1. Launch NVIDIA Control Panel
2. In the 3D Settings pane select "Manage 3D settings".
3. In the Global Settings scroll table find "Power Management mode" and switch to "Prefer Maximum Performance"
 
Computer specs:
CPU: i9-10920X
Fan: NZXT Z63
RAM: G.SKILL Ripjaw V Series 128GB
PSU: Corsair RM850x
MB: ASUS ROG Strix X299-E Gaming II LGA 2066
HD1: Samsung 970 EVO Plus 2TB
HD2: Samsung 970 PRO 1TB
HD3: Samsung 860 EVO 2 TB
HD4: Crucial 1TB
Case: NZXT H510 Elite

So the crash persisted after switching to "Prefer Maximum Performance" but I have done the following and it seems to be playing nice for 5 hours straight:
1. Open File Explorer and go to C:\Windows\System32\ and search for nvlddmkm you should see "nvlddmkm.sys"
2. Right click it and go to properties and click the Security tab.
3. Click "User" and click Edit and give yourself (User) Full Control
 
I also did this and not sure which one fixed it for me:
1. Launch X1
2. In the GPU Clock section enter in -100
3. In the GPU Memory Clock section enter in -100
4. Apply
 
Hope this helps! Cheers!

I'm currently experiencing this exact issue, as well. Went from a 2080Ti FTW3, that was operating fine, to a 3090 FTW3 that would do this periodically.
It would occasionally yield a "25" debug code on my MOBO, as well, but that was never a guaranteed precursor to the issue occurring. I could be A0 or E3 and it would still happen.
Same Event Viewer System Event "Display driver nvlddmkm stopped responding and has successfully recovered."
 
Mine manifests a little differently. I won't get game crashes, per say, but I'll have everything go black for 5-8 seconds and then everything comes back, usually not the same, either.
I'm usually not doing anything serious when it happens; maybe watching a YT video with a couple monitoring apps in the background.
This has persisted over a couple Driver versions, too. Right now, I'm on 460.79 (12/9/202)
 
I've yet to try the High Performance setting on my Power Plan, so maybe that will take care of it.
 
Even if underclocking handles the issue, you have to admit that's completely unacceptable, both from a hardware and driver point of view for any card, especially these $1800 3090's. 

Ok so what ended up solving the issue for good for me without under clocking is actually setting it to Maximum Performance mode and then rebooting the system:
1. Launch NVIDIA Control Panel
2. In the 3D Settings pane select "Manage 3D settings".
3. In the Global Settings scroll table find "Power Management mode" and switch to "Prefer Maximum Performance"
4. Reboot
 
I have not hit the issue in 2 and a half days so please disregard the under clocking suggestion as this did not help and actually raised my thermals.
Hope this helps!

Its the same issue 3080 owners also have. There is a whole stickied topic about this above.
Currently, only setting prefer max performance solves it. There a few other rumours too, but nothing else confirmed.
And, sadly, nvidia also seems not to confirm having this bug, because they dont see it on their telemetry, lol... Of course they dont, because hundreds are now using it with prefer max performance, so that they are not annoyed by random crashes.

In the past I've experienced two black screen flashes and 1 driver crash. Just had it occur again and I triewd tdik543's fix. Will report back if the issue persists.

I always updated my drivers to 461.09.

Did it happen again College_Debt? Or does the fix of tdik543 work?
tdik543, do u still have any blackscreens?