lately I have been having serious issues trying to fold (7.6.21).  My rig will work for a couple of hours and then suddenly it stops folding altogether or turns off.  I love being woken up by the sound of silence.
Ryzen 2700 (Asus TUF Gaming X570-Plus Wifi)with my 1080Ti SC2 Hybrid (461.09) Power Supply is a 1000PQ
Windows 10 Pro 20H2 Build 19042.746 
no overclocking, EVERYTHING is on stock speeds.  Ram is at 2933.
 
20:37:44:WU00:FS01:0x22:Completed 360000 out of 1000000 steps (36%)
20:37:45:WU00:FS01:0x22:Checkpoint completed at step 360000
20:47:57:WU00:FS01:0x22:Watchdog triggered, requesting soft shutdown down
20:57:57:WU00:FS01:0x22:Watchdog shutdown failed, hard shutdown triggered
20:57:57:WARNING:WU00:FS01:FahCore returned an unknown error code which probably indicates that it crashed
20:57:57:WARNING:WU00:FS01:FahCore returned: WU_STALLED (127 = 0x7f)
 
It will attempt to keep folding and will get to 99.99% and then hang
 
20:57:59:WU00:FS01:0x22: cuda-device 0 specified
20:58:09:ERROR:Receive error: 10053: An established connection was aborted by the software in your host machine.
20:58:14:WU00:FS01:0x22:Attempting to create CUDA context:
20:58:14:WU00:FS01:0x22: Configuring platform CUDA <-- hangs here 
 
Then when it has hung and I have to force close the application, my system refuses to fold until it has been rebooted.
It goes through the motions (folding bar moves as if I am folding), but my card is not being accessed (GPUZ reports everything as being idle same with UPS software), nor is my cpu.
 
Is it my card that is failing because it appears to hang on the CUDA platform?  I have also tried to go back to different driver versions, as far back as 442.74 and still the same results.  Card will fold for a while and then just hang and go through the motions.  
 

Really not sure if this is of any value but I had the same symptoms during the TZC on my win 10 gaming rig. Finally it just would not boot. Took a while for me to dig up a spare cpu to test the mobo and it boots to bios but goes into a boot loop for a bit first. Temps all looked great when folding (gpu only) but when I pulled my corsair cooler off I saw a lot of brown bubbles in the tim. So, if using the stock tim on a clc I'd say pull it and double check contact. 
Like I said, not sure if this is of any value but check before they are both toasted like mine.  

Are you running the beta flag by any chance?

Single GPU rig, or do you have both GPUs in there? If both it might be a configuration issue. The card sounds fine, but you can always test it with https://www.ocbase.com/ which has a built-in artifact scanner. 

>>>0xc0000005
 
So that's the current event log error, not the one from October correct? As you said it can be RAM, but also registry/file system damage, and program corruption can also trigger it. You might open a CMD prompt and type sfc /scannow just to rule it out.
 
I would suggest you uninstall the F@HClient, and then make sure to manually delete the %\AppData\Roaming\FAHClient folder if it still exists. Then I'd DDU your NVIDIA driver and install the latest one. Then install the latest 7.6.21 folding client.  

DDU your drivers, reboot, and re-install. Run a system file check and then do a Windows Update check as well.

Well I'm confused, but that's fairly normal.  So if your system only has kernel power / unexpected power loss in the event logs and nothing else then it's likely a hardware issue. But if it has the time to write other eventlog errors like Exception code: 0xc0000005 that you mentioned but didn't clarify, then it's possibly a software cause. 
 
Doesn't sound promising for the GPU though, you might swap your 1080 and 1080 Ti with each other and see if the error follows the GPU or stays.

Memtest came up clean.  I don't know of a way to test video ram outside of linux with GPU_Burn and you need to compile it yourself..which for me might as well be brain surgery.  Is there a windows version of this?  I have not been able to locate one.  I want to be able to test the actual GPU memory, like you can with memtest and regular ram.  I am getting sick and tired of having to reboot my system daily, sometimes 2-3 times.  If I don't fold, everything works great...

can you try another card in it and swap that out - if it works fine with another GPU ... u know it's a Card issue
 
you can always underclock the ram  and see if that works
 

https://blog.pcrisk.com/windows/12778-how-to-fix-application-error-0xc0000005
 
https://www.ionos.ca/digitalguide/server/configuration/resolving-0xc0000005-errors/ 
 
https://techcult.com/wiki...tion-error-0xc0000005/
 
id definitely would test with another card first - even if you have to borrow one
 
if it fails with a known good card you know its not your card and it something else on your system
if it doesn't then you know it's the card
 

Well got a replacement at the beginning of March, and now the replacement card is checkerboarding.  This means it's going to cost me another $80+ to send a card to EVGA and HOPE I get a working one.  This is awesome.

troy8d
Sorry to hear this, sounds like some really bad luck to me.  Hope you have better luck with the next one. 

+1!!