Hi Fennicillin, 
 
I'm sorry to hear that you're still having issues with the unit, however it does sound like there may be more going on here than just a faulty card. 
 
I'm sure you've heard by now that we verified the failure on the first unit, however the last card you sent in passed testing as well as the one we sent out to you so it would seem unlikely that we would be able to detect the issue once but not on any subsequent cards experiencing the same problems. Of course saying that is easy, and this doesn't mean the the issue is not present while you're trying to game, however can you please note which cards have you been using to test in the interim?
 
If the card can pass a benchmark multiple times without issue, even though this is a synthetic environment, we can use this information to help determine the health of the card since the same commands and operations that would be occurring during game rendering can be performed without engaging other processing and transport systems to help reduce the complexity of the scenario. Maybe running a higher resolution benchmark would trigger this?
 
I'm not here to claim that I know everything, but what this usually indicates is that the issue lies in the supply chain, and since everything seems to be working fine with other cards this would likely point toward:
-PSU (if the power requirements are significantly different; this would include power via the PCIe bus)
-Software (if there are driver conflicts, etc. with the 900 series geforce driver you are using)
-Firmware incompatibilities (pretty rare, motherboard may be having some weird issues communicating with this specific card)
 
In any case, EVGA would like to assist you in resolving this issue however it sounds like more troubleshooting is needed to get to the bottom of this. If you would like to troubleshoot here, maybe the community can provide valuable input and suggestions. Otherwise, I will email you directly and we can try to move forward from there.

How are the temps on the 980 Classified when you get the crash?

Sorry to hear about your frustrstions. Hopefully EVGA can get you sorted and get everything running for you.

Did you check the windows event viewer on those systems for BSOD code information? This can help guide us in the right direction.
 
If the issue is specific to a game, or game engine, then I'm not sure how we can generalize this to hardware failure; temps don't seem to be affecting the issue, and PSU capacity (at least) is not a question in these systems. 
 
The necessity to downclock the card to increase stability does point toward overclock or timing stability issues (even potentially power fluctuation rather than failure), however this is still relatively vague given that benchmarks seem to be unaffected. You wouldn't happen to have an overclock on your CPU would you?

If the card fails in system #1 and system #2 (friends), i don't think it would be due to monitor setup on system #1. It's rare, but i have seen people over the years go through 3-4 RMA's before they received one that works like it should.
 
If EVGA thinks it might be power related, offer to cross-ship RMA the 1000 G2 power supply too.
 
 
 
 

MSim
If the card fails in system #1 and system #2 (friends), i don't think it would be due to monitor setup on system #1. It's rare, but i have seen people over the years go through 3-4 RMA's before they received one that works like it should.
 
If EVGA thinks it might be power related, offer to cross-ship RMA the 1000 G2 power supply too.
 
 
 
 

Yeah, they can just go ahead and charge me $160 over 2 weeks of turnaround time, take my power supply and send me back a janky RMA to add more problems to my list.

Fennicillin

MSim
If the card fails in system #1 and system #2 (friends), i don't think it would be due to monitor setup on system #1. It's rare, but i have seen people over the years go through 3-4 RMA's before they received one that works like it should.
 
If EVGA thinks it might be power related, offer to cross-ship RMA the 1000 G2 power supply too.
 
 
 
 

Yeah, they can just go ahead and charge me $160 over 2 weeks of turnaround time, take my power supply and send me back a janky RMA to add more problems to my list.

Talk to a RMA supervisor, see if you can work something out.  Maybe they will offer a free EAR (EVGA Advanced RMA) for the GPU and PSU.   
 
 
 
 
 

I even begged for them to downgrade me to a lower end 980ti because I falt out don't trust their binning anymore. Which I shouldn't have to do! And they said "sorry Charlie, it's policy." They're trying to empty the bucket on their 3 year old refurbs.

Fennicillin
Las time I tried to work with them to advance the RMA for the two junk cards they told me they already bent over backwards to give me "extended testing."

Push for it again, you shouldn't have to pay to return a faulty replacement card. If they want to help resolve it, ask about advanced RMA on the power supply too, if they think it's a power issue. If they think it needs more troubleshooting, stay on the phone with the tech until they admit you need another RMA. If the supervisor is unwilling to help you, ask to speak to his boss. 
 
Your not the first person to ever have to do 3-4 RMA's on a EVGA graphics card, don't let anyone tell you otherwise. It's rare to happen, but it has happened.
 
 

I agree, it's no fun having to deal with RMA's.  Are you dealing with US support or EU Support?
 
 

Tried calling Customer service with two hangups on their end and now a busy signal.
 
EDIt: Tried calling 10 times over the past hour. Rings once, silence, rings again, then hang up on their end. This doesn't happen when I call anyone else.