Thanks all for the feedback.
Try replacing the powered USB riser, or dedusting the slots, testing components like RAM, CPU, disable OC, disable all sleep and low power states, lower the miner's intensity, etc. There are so many things that can be wrong causing your problem.
I tried 3 different brands and also tried the system with just one card(plugged into the motherboard directly) so I ruled that out. Also it's super clean as it's been in a filtered room, but will try de-dusting them.
RAM and CPU could very well be related, I'm thinking maybe the heatsink isn't doing it's job, but from what CPU-Z reports, it's nice a chilly around 50-60c under load. Is there a log to file method? Maybe something happens right before it powers itself down that I'm missing since I can't catch the heat readings.
Tried 3 different power supplies as well. It's odd because it doesn't seem to be an OS crash, just simply powers itself down. In Bios I have all OC disabled, and have it to resume on power failure (such as being unplugged from the wall, when you plug it back in it should power up).
It does this, but if I let it "crash" then try this method, it won't power back up automatically, I must hit the power button.
This is very odd, since I can kill the power from the wall or PSU, then resume power, and it comes right back up.
But if this "crash" happens, the only way to get it to turn back on is the power button.
Also this happens even when not mining, so that's another odd thing. I'm using Windows 10, various re-install methods, next is to try Windows 7 I guess, but I hear Windows 7 can only read a certain number of GPU cards for mining, I can't find a solid answer on this for mining.
It was running great for 3 days straight when I first built it, then it just started powering itself off constantly, after running for 1-2 hours.