Okay gang. I'm one of those poor souls who bought the RTX 2070 Super card (see below). As has been reported on many (oh so many) forums this card is... well... crap. The card crashes inconsistently, but all too frequently. It crashes the card only. The OS stays up and running (nothing like listening to a video...). Tried an incredible variety of fixes including the most obvious (driver re-install, physical re-install of card, different IDE slot, etc.).
 
My system is pretty damn beefy enough so that I don't overclock anything.
 
PC:
Device name Dad_Office_PC
Processor AMD Ryzen 7 3700X 8-Core Processor 3.60 GHz
Installed RAM 64.0 GB
Device ID FCC4BD67-A374-44E1-935A-BA7A9D43ABDF
Product ID 00326-11250-10002-AA121


OS:
Edition Windows 10 Home
Version 21H2
Installed on ‎11/‎7/‎2020
OS build 19044.1566
Experience Windows Feature Experience Pack 120.2212.4170.0
 
GPU:
EVGA GeForce RTX 2070 SUPER FTW3 ULTRA GAMING, 08G-P4-3277-KR, 8GB GDDR6, iCX2 Technology, RGB LED, Metal Backplate

Crashed while your doing what ?  Gaming ?
Answer: Anything, unfortunately. In fact I gave it the old 'benchmark' try of 'just run the OS in safety mode for 24 hours' thing and it never made 24 hours before crashing. Have had the PC just sitting on the OS home screen during a phone call and came back to no video plus GPU fans going full blast.
 
Have you looked @ this yet ?  
 no but as part of my ongoing MCSE certification process I've gone through everything listed (and then some)

Run GPUz & look under the sensor tab for Temps & other good information on what is happening ?
have run multiple tools for machine overall 'health' including mobo temp/voltages, CPU temp (liquid cooled system), RAM temp (cool little heatsinks they make these days), GPU temp, PSU temp and voltages
 
Ever run DDU for a clean driver install?
 Yep. Repeatedly. 
Which Nvdia driver are you using ?
 Just updated to the one they released in February 2022. Don't recall the version at the moment and the PC is shut down.
What monitor, what type connection ... if DP is it certified DP cable & monitor firmware up to date ?
 Monitor isn't the problem, but 32" AOC top-shelf OLED. Not using DisplayPort, but did try to see if that helped using the DP port on the card... nope.

Did you try all the steps found here? 
Ad infinitum. Been building PC and servers for literally 4 decades now. In fact, just finished building a nice gaming rig for a friend's son this AM.
 
Basically, this smells like a hardware problem with the GPU. Sure 'feels' like I'm hitting some instruction on the chip that piece is flawed and it just pukes. Those 'flawed' areas can grow over time if the original flawed area causes micro-fluctuation in the voltage. Unfortunately, the tools available won't get that 'granular' for monitoring the GPU. Well... my tools would, but disassembling the card to using a series of voltmeters on the chip connections would pretty clearly violate the warranty so no can do.

Understood on all the above. 
 
1.) What use case during crash? Anything; including just trying to 'sit' on Windows SAFE mode for 24 hours (never makes it even a few hours much less 24) as a benchmark 'minimum' use case. This to me is the 'well that settles it' test. If the thing pukes using nothing but a basic OS driver simply to drive the desktop repeatedly, but not 'on demand' meaning I can't 'trigger' a crash even running ridiculously maxed out settings in pretty hard core video games, video rendering 4k with preview in CyberDirector, BlueRay playback software running a 4k movie. None of those will automatically 'trigger' the failure. Whatever is done the card WILL fail within a 2 hour window. I've had it fail within minutes of rebooting the OS and simply letting it run multiple times then 'third time's the charm' it will stay stable for an hour while I do browsing, play a game, etc.
 
2.) Log checking - I've checked Windows OS logs, my SPLUNK server logs monitoring this machine, the logs in the various applications (Steam's are... damned disappointing bluntly in the lack of detail), as well as running a diagnostic tool I built myself via PYTHON that runs a multitude of checks across monitoring systems and produces a compilation report (interesting how two pieces of software can report a different temperature supposedly talking to the same temp probe).
 
3.) Yep OS instability is far more frequent an occurrence than hardware faults (usually... however... faults tend to run in 'batches' and this card has a history... to the point they short-stopped retail sales WAY before they could possibly have recouped RD and assembly line costs). In this case, however, I'm able to confirm the underlying OS (and even application) are running just fine. I've 'listened' to YouTube videos, BlueRay movie soundtrack, a variety of PC game soundtracks, etc. The video shuts down with no warning and the GPU fans go into 'full blast' mode despite the temp on the PC being well within tolerances. This thing has great air-flow, a good environment, and a liquid cooled CPU. 
 
Actually as a long time (oh so long... 4 decades now) troubleshooter your best thing you can possibly do is NOT try to assume anything one way or the other. Simply look at your testing results and understand what, or what combination, of things can cause the symptoms and work from there. My take is that this card is faulty. By the way putting in my old card into the same slot with same everything... works fine. This card was supposed to be the 'upgrade'.

Did this issue just recently start to happen or is the card always done this and are you the first owner of the card?

First owner. Bought it brand new from NewEgg to replace older AMD Radeon RX 480 (8GB) card. It happened once or twice the first few months and I wrote it off to patch issues, driver conflict between apps, etc. since it did seem to happen initially when I was playing either a video game at high resolution or using CyberDirector to do video editing. Then about 5 weeks ago it started happening frequently and then two weeks ago it started happening constantly.

guiwhiz
AMD Radeon RX 480 (8GB)

Previous card being Radeon RX 480 means you might still be having remnants of AMD drivers in your Win 10 install. If you haven't already done -complete- DDU that included AMD drivers too you should do it (in safe mode when disconnected from net). Even better, if you haven't already tried completely fresh Win 10 install and you have time for it, you should try it with only Nvidia card.
 
Disable fast boot and disable CSM if you haven't already.
 
Also, your old Radeon has lower power requirements than 2070. Lot of the times when people put in new card with higher power requirements and that starts crashing that can point toward PSU or connections between PSU and card. Make sure you are using all connectors and that you are not using piggyback connectors. You might also want to try running Win 10's ultimate power profile and setting 2070 to 'prefer maximum performance' during troubleshooting to check for possibility fluctuations in voltages are making your system unstable.
 
Also, your mb is PCIe 4 while 2070 is PCIe 3. In theory that shouldn't matter, PCIe is backward compatible, but just in case something odd might be happening try forcing in BIOS your video card PCIe slot to Gen 3. And I assume you already made sure your mb is on latest BIOS.