Memory_Management BSoD and weird shutdown error

[Skyl3r]

Hello All,

I've been working on a new build and it's been quite the ride
The hardware I'm working with is as follows:

CPU: Opteron 6276 @2.6GHz
MOBO: SuperMicro H8SGL-F
GPU: R9 390x XFX 8VR6
RAM: Crucial Ballistix 8x2 16GB CL9 1600MHz
PSU: Corsair CX600


Background Info, in case it's relevant:

• 
  ■ Disabled onboard VGA via the JPG1 jumper
  ■ Reapplied thermal paste to the NB heatsink

Now the symptoms:

• 
  ■ When I first try to install AMD GPU drivers via AMD Install Manager after a fresh Windows 7 installation, it gives me a BSoD when it arrives at "Installing Gaming Evolved." The error is "MEMORY_MANAGEMENT". It has done this 3 times now at the same location. Once I uninstall the drivers and try again, it installs fine.
  ■ I experienced these BSoD through regular operation. I'd click a button and it'd randomly BSoD ("MEMORY_MANAGEMENT" is always the error). To test this, I swapped out one stick of ram and continued experiencing it, then swapped the other and it SEEMS to have gone away.
  ■ I did a Windows Memory Diagnosis test and it returned no errors on the stick that seems okay thus far.
  ■ I thought it was now fine, so I started a download and walked away. When I returned the PC screen was black and the system seemed to have rebooted itself and got stuck at turn on. I power cycled and it came up fine... Okay.
  ■ I attempted to shutdown from windows. It pops up the little spinny circle like normal but then jumps straight to the screen seen below.

This appeared suspiciously like a GPU problem. Drivers or otherwise (wouldn't be surprising given the failing AMD installation). So I used Furmark to stress test the GPU and it operates fine. It heats up to ~64C @ 1080P and all my other temperatures look okay. CPU goes to about 44C and NB is at 59C,.

So. I don't know if I have other problems or how I can test them; but that weird screen when I shutdown is seriously concerning.

TL;DR
I'm getting the screen seen above when I try to turn off (Start menu -> Shutdown). Have you seen this and what are possible causes?


EDIT:
SFC /SCANNOW yielded no errors.

EDIT:
Just got the following screen when trying to open chrome:

I don't know where to post this, because I don't know if it's a driver issue (I have a suspicion it is) or a hardware issue.

 

[Paul NZ]

Use memtest from here. The windows one does stuff all

If any errors come up replace the stick

Is the BIOS up to date? Looks like it's up to R 3.5a

 

[Hamperking68]

Try doing an entire re-installation of windows.

 

[Skyl3r]

Hamperking68:
I'm okay with doing a complete reinstall, but I have done like 4 in the past few days. Sorry to not include that in my first post.


Paul NZ:
I have Memtest86 running and will let it go for a few hours. I'd be so happy if the problem was as inexpensive as a few new sticks of ram

I didn't see your BIOS question. The BIOS are at 3.5 running a custom firmware OCNG. This firmware should be stable; I'm nearly positive it is not the issue.

In addition, I'll add that if I install the ram in slots A1, A2 the system gives 6 beeps indicating no memory detected in system. If I install it in B1 and B2 then it works.
I'm starting to get more and more suspicious of this motherboard.

 

[Hamperking68]

In addition, I'll add that if I install the ram in slots A1, A2 the system gives 6 beeps indicating no memory detected in system. If I install it in B1 and B2 then it works.
I'm starting to get more and more suspicious of this motherboard.

I know this may sound weird but try loosening the screws closest to the RAM slots. If there is pressure on the board it can flex and cause issues.

However, Now that I have that information i'm 70% confident that the issue is the motherboard.

 

[Skyl3r]

Yes, I'm getting unfortunately suspicious of this as well. I'll try the loosening the screws tip first and perhaps cleaning the slots out.
I was hoping I wouldn't have to buy another motherboard, but it's starting to look like I might.

The first pass is almost done on the ram and I'm getting no errors thus far.


More info on the RAM slots working/not working.
Every blue slot (B) works with one module and every black slot (A) fails.
This isn't the end of the story though.

If I do one module in A1 and one in B1, it works and shows 16GB of memory, but it's underclocked it 1333MHz (this may be due to single channel or non ecc. I recall reading something in the manual about this depending on the slot population.)


Soo... To be diligent, I ran some stress tests. Each for 10 minutes.

CPU using Prime 95 Ran fine and peaked at 51C
GPU using FurMark ran fine, though I had to adjust the fan settings to not overheat. I had it throttle at 70C and no crashes.

How come I click chrome and it crashes but I can run stress tests and it handles it like a champ...


EDIT
Okay, CompuBench got to the BTC mining portion then crashed in the same way the pictures I posted previously describe.
Is it possible that my 600W PSU isn't enough?

I ran the BTC test twice more with no crash.

EDIT... meh, spoke too soon. Crashed several minutes after running.
My friend said he suspects it might be NB failure.


I've decided to contact the Mobo seller for a refund.