I have not attempted to OC my 1070 FTW yet, but Doom is the most GPU intensive game that I currently own and while playing it I often see the card's core clock peaking out at about 2ghz. Normally I'd be quite happy about that but I have also been getting frequent crashing to a black screen(only mouse pointer shows) that requires me to do a hard reboot. I don't want to be too quick to blame the card for not being able to sustain that clock, but I played through the game on my GTX 680 before upgrading from it and I never experienced any crashes like that. Are those clock speeds normal for a stock FTW? I'd like to test the card's stability before assuming it to be faulty but I haven't played with GPU clock settings since my 680 was new so I'm not sure what the best stability test would be? Furmark? Is that basically like the intel burn test for GPU's?
 
That might be a noob-ish question but like I said I'm a bit out of practice with this.

A brief component over is:
Core I7 3770K @ 4.0 ghz, EVGA GTX 1070 FTW, Crappy Asrock Motherboard, 16gb gskill ripjaws, 1000W Seasonic PSU
 
Nothing changed from when I was using the GTX 680 without problems. I just bought the 1070

CoercionShaman
So did you use DDU when you installed?  Have you tried in debug mode?  Just go through those steps in that post first.  A lot of people are able to resolve issues with these simple steps.
 
Debug mode might also show an unstable factory overclock.  It happens.  Other similar posts in the forums.
 
Also, what OS are you on?

I will follow that guide when I get a chance later, at work for now so I wanted to plan how to test when I get home. I used driver sweeper and the clean install option with the most recent drivers on windows 10. I didn't know about debug mode until your reply but that looks like a good place to start. I will actually be relieved if it crashes in debug mode, crappy drivers or something I'd be fine with but I hope to not have to return the card for an unstable factory OC. Is there also a certain period of time that running Furmark or something should indicate stability? Like an equivalence to prime95 for so many hours or passing so many loops of burn test for the CPU?

CoercionShaman
I'm pretty sure Furmark is blacklisted by NVIDIA and will throttle down because of it.  Use Heaven or Futuremark.  You will generally see issues in the first pass.
 
Also, if you are on Anniversary Edition, it is a known issue that it resets a lot of settings in the NVIDIA control panel.  Be sure those are correct as well.  If you are reinstalling with DDU to be certain as per the forum guide, you would have to reset that anyway.

Thanks for your help on this. A few updates: I don't have anniversary edition but I have previously gone through control panel to tweak some of the settings anyway, such as set the power control mode to prefer maximum performance. I have downloaded the latest version of heaven and ran several passes without it crashing. Then I was able to give it a +100 bias for a boost clock of 2076mhz. I was able to run a few passes in Heaven without is crashing at this speed. It did eventually crash once for me but that time(unlike when I tried to push the core clock over 2.1ghz) it did not report it as a display driver crash, but a more traditional "program stopped responding" and I did not see the discontinuity of any of the trends in MSI afterburner which my understanding is that is what occurs when a crash occurs due to an overclock.
 
I'm going to try a few more passes at this speed to verify if I can get through it 3-4 continuous times. By my question is, is that a good or typical overclock speed? Since heaven is directx 11, and Doom is dx12, could the later dx version contribute to an added tendency to crash from instability at the stock FTW speeds? To be able to add even a little bit of extra core clock gives me added confusion about whether or not the card is being problematic at factory FTW speeds.