AMD Ryzen 7 1700X (stock speed but 1.375v, LLC3 manually)
ASUS Prime X370-PRO (newest 0810 BIOS)
32GB Corsair CMK16GX4M2B3200C16 (4x8 single rank, 2x16GB kits) @ 2933MHz
EVGA GTX 1070 FTW
Samsung 960 EVO NVMe M.2 (boot drive, there are a couple HDD’s too)
Corsair HX750 PSU (bought new 2013)
So lately I’ve noticed that if I leave my computer running but walk away for a time, instead of going to sleep it shuts down. It will also occasionally shut down instead of going to sleep when I put it in sleep mode manually. I don’t remember this happening when I had an RX 480 in the system, although that was a single 8-pin connector and the 1070 has two 8-pins, leading me to believe this may be a PSU issue. I am considering ordering a tester to make sure, but was hoping someone else may have some insight.
I have had a few other strange issues with this build as well, such as during OC testing stability the system crashes to a black screen and the video cuts but the computer doesn’t power down, although power to the keyboard drops and I have to use a long-press on the power button to reset (this also happened on the 480). This seemed to have been solved by upping the CPU voltage to 1.375 and the LLC to level 3, although I don’t normally run the CPU overclocked anyway.
RAM hasn’t been stable with all four sticks at 3066MHz although it was for two sticks before I bought a second kit. Haven’t tried for higher on the new BIOS, just 0807. Installed 0810 today.
Thoughts or common experiences? The sleep issue doesn’t happen all the time, but often enough it’s concerning.
Was the 480 in the same system or in a different system with the same PSU? Might be a PSU issue, but hard to say from afar.
2 Kits bought separately can always lead to issues, it’s always recommended to buy kits of however many sticks you want to use. This is because even in the same product line they tend to use different chips. And even when they have the same chips, the sticks might have different parameters for them to be stable, because chip quality varies as with any other type of chip. Regardless of that, having a lower clock on 4 sticks is perfectly normal compared to 2 sticks, because it puts more strain on the Memory controller. If you’re lucky you can run them at the same speeds, but there’s never a guarantee.
480 was in the same system, and my previous system (FX 8350, Gigabyte board, 16GB 1866 DDR3) with the same PSU. Didn’t have any issues then, upgraded from an HD 6970 that was showing its age.
Oh I was aware of the risk and prepared to return them if need be. I am pretty happy with 2933MHz though, I just didn’t want to have to dip down to 2666 or even 2400MHz. I could do some testing with the memory clocked lower though and see if it happens again, but as it seems to be fairly random I’m not sure if that would tell me anything, and would require a lot of wait time. FWIW HWiNFO64 spits out identical ID information for all four modules although that may mean nothing. I’m not as familiar with RAM as I’d like to be.
That’s what I was expecting but it’s been a surprise. No errors when I run MemTest64 at 2933.
Hm… did the issue start appearing before or after you added the additional RAM sticks? You could try ripping out the 2 additional (or original for that matter) sticks and try if the issue persists. Might be that Windows can’t put the board to sleep as it wants (due to different chip parameters on the RAM sticks maybe).
If the issue is just with sleep, then just don’t use that function… disable it and turn off when your away from desk (with an 960EVo M.2 it should boot quick enough). Not a solution to your problem, but should make it at least usable until a better understanding of whats going on can be had.
It’s an option at least 
