I bought this Supernova power supply back in 2015 and didn't install it for several years (after ordering, my UPS tripping on dirty power turned out to be actual line workers a block over and not my old PSU).  Eventually I installed the EVGA PSU and it seemed to run my 2nd gen/sandy bridge system fine, except every so often I'd find the computer wouldn't start up after it went to sleep/powered down.  It basically ignored the button press to power up, although coming back an hour or two later it would run (maybe not the first attempt, but eventually).  I attributed it to my older sandy bridge-era hardware and used that as part of the reason to replace my computer.
 
Enter my newer system, where I have an evga Z370 ftw board and migrated the evga gs 850 PSU over to this 9th gen system.  I assembled it over the weekend and I've observed the exact same behavior with the power supply refusing to respond to the power button on the case and also not respond to the power button on the motherboard.  I've unplugged all the modular cables and then used the 24 pin connector + PSU tester and observed the PSU power up, but then still had the PSU refuse to power up when I reconnected all the modular cables again.  The problem has been intermittent, but I kind of think it's happening more frequently lately (and it's no fun when your new build refuses to start).
 
I've done a quick search for others with the same issue with GS PSUs and came up with similar stories.
 the urls are stripped but they're posts - I-think-my-SuperNova-850-GS-is-failing-m2601784, Supernova-850-GS-PSU-problems-2nd-unit-m2488599, Possible-issue-with-850-GS-and-booting-m2396853
 
As part of my diagnosing the problem, I completely disconnected the case's power switch at the motherboard header and figured I'd just rely on the motherboard's power button.  The computer didn't post and I resumed writing this post on my old computer, but several minutes later the computer with the evga PSU has just fired up on its own, without even pressing the power button on the motherboard.
 
I think I need to open an RMA for this PSU, but is there some more advanced logic going on if you use a motherboard's power button vs just a short at the PSU's 24 pin cable (as the PSU tester does)?  I'm kind of at a loss as to what I can diagnose here, but the same intermittent behavior from the same PSU in two different computers sure does appear to indicate the PSU.

Thanks Matt.  I did get the e-mail and have started an RMA.
I've had a number of EVGA parts over the years and it's the first time one has displayed issues before being retired, so EVGA still has a pretty good record in my book.  Now I just need to locate all the spare cables/PSU bag/etc.....

I just wanted to drop a line that Christian was helpful in getting my RMA issued and the cross-shipped G3 arrived today, with the full compliment of cables, power supply tester, etc.  That's good because the GS cables aren't compatible (even when the pin positions are the same for the peripheral/drive cables, the connectors are keyed slightly different).  The replacement PSU sure looked like it might have been new and is working well in my new build.
 
I wanted to give a day or two to observe that I don't see the issues with sleep/resume and will get the old/RMA GS PSU in the mail soon.  Thank you EVGA Matt, Christian, Jacob/Peter, etc.

I'm happy with the cross-ship replacement (Shipped From EVGA: Shipped On: 7/10/2019) and used EVGA's shipping box to return the not-working-right 850 GS today.  Should be at EVGA in Brea/92821 by Friday.