Ok understood CW. Could it be possible that even though the memory used is in the QVL list for the motherboard that it would still cause BSODs? I don't really feel like futzing with Memtest so would next steps be updating to latest BIOS, installing all RAM and then retesting?
I'm not aware of a better way to be sure memory is compatibile with a motherboard than the motherboard manufacturer's QVL so an incompatibility would be low on the list of possible causes in my mind. The QVL specifically states your memory has been certified for 2/4/8 DIMM configurations which I would hope means they've thoroughly tested those configurations for stability.
As far as what to do next, my understanding is you've at least divided the memory into 2 sets of 4 and tried each set in slots A2, B2, C2, and D2 for an extended period of time without experiencing any BSODs. That would suggest the memory is okay, if true.
It's possible for a slot to be bad. It's usually quite obvious if a slot is bad but perhaps a faulty slot can be more subtle than I've personally experienced. The memory chart in the motherboard manual doesn't show it as a possibility but another test I'd try is 4 DIMMs using A1, B1, C1, and D1 to see if the problem returns in that configuration. Since the memory is likely okay a bad slot amongst that group of slots could be revealed that way. One test in that configuration would be sufficient assuming the computer can actually boot in that configuration. I don't think the memory chart was meant to show all of the possible configurations which will work but I don't know for sure.
However, I'd still want to update the BIOS before replacing hardware. Whether it's the first or last thing you do depends on your comfort level. If you can get the system to boot with an A1, B1, C1, D1 configuration it might be worth using the system as you normally would (like you did previously) if you have the patience to do that test again. If you'd rather not, I'd update the BIOS and try all 8 DIMMs set at DDR4-2667. That should get your motherboard in as stable a state as ASRock have been able to achieve using 8 DIMMs and supported settings.
Ok thanks for the detailed response CW. I've updated the mobo to the latest BIOS (3.3), will reinstall all 8 DIMMs and see what happens.
OK I reinstalled all RAM, initiated a GPU render and the system is still running strong as of last night. Any final thoughts? Tests?
I think the best test would be to use it without mercy and see how it does with everything you need it to do. I've seen BIOS updates fix stability issues with many systems and it wouldn't surprise me if a BIOS tweak was needed to make an 8 DIMM quad channel configuration more reliable but I'm afraid I don't know for sure.
Hopefully you've got it working properly now but please do let us know if the problems return. Good luck!
Got a BSOD. UNEXPECTED_KERNEL_MODE_TRAP caused by amdppm.sys. This latest occurrence is typical: late at night when the system is idle.
Here's the minidump.
Some things to remember are that this is a new motherboard and this all started with an unsuccessful OC of my CPU as well as botched cleaning job which forced me to clean the underside of the chip (and replace the motherboard). It seems that all RAM installed causes BSODs so should I now consider testing each RAM stick individually with Memtest86 to find the offender? Or are other tests recommended?
Also I hope everyone had a good Thanksgiving and I'm very grateful for all of your help thus far!
The bugcheck codes and callstacks are so random I have trouble believing it's not a memory issue. I can't vouch for the accuracy of this post but, if true, it suggests some CPU pins are dedicated to memory communication and can effect individual memory slots.Can you clarify whether or not you tried the A1, B1, C1, D1 configuration?
Nope I surely didn't and will try here shortly.
Ok I've populated A1, B1, C1, D1, the BIOS reads quadruple channle and the system boots fine. I'm going to initiate a GPU render and see what happens. Do you think this could be a CPU problem?
Quote