2 sticks of RAM in paired slots, running in single channel mode, and Memory Management stop codes

DecibleBreaker

Commendable
Jan 4, 2017
4
0
1,510
I have an ASRock 990FX Extreme mobo (only a couple months old) with a pair of 4GB Kingston HyperX Black DDR3 RAM (operating at 1600, a couple years old), and over the past few days, very literally starting on Monday 1/2/17, I have been getting Memory Management (0x1a) stop codes. It can be extremely frequent, sometimes happening within a few minutes of the computer booting up.

No system or driver updates appear to have been done shortly before these crashes started happening. I have since updated a few drivers that did have updates, with no change to the crashes.

Windows Memory Diagnostics - 5 passes, extended test - cleared with no found issues. I disabled the paging file, deleted it, and recreated it in case that was corrupted, but that wasn't the issue. I noticed in the Windows taskmanager, it reported 4GB was hardware reserved, so I checked my BIOS. It reported the RAM was working in single channel mode, but it saw both sticks properly.

I attempted to swap the position of the two sticks (a1>b1, b1>a1), no change. I also placed them in the empty slots(a1>a2,b1>b2), to no avail. I removed one of the sticks, and attempted to boot with just one(in a1), receiving the "NO RAM" beeps from my mobo. Assuming that stick was bad, I removed that, and placed the other stick in a different slot(b1). The computer booted, and appeared to be OK for a while, before reaching another Memory Management stop code.

So my question... what's most likely the issue? The motherboard, both RAM sticks failing simultaneously, or something entirely else?
 
Solution
The memory control hub could be on its way out the window, which sadly means you may need a new motherboard.
This is exactly how my last motherboard fried on me, the memory control hub, or "north bridge" got too hot and left scorch marks. My computer would still start, but it left me with frequent random restarts and Memory Management BSOD's.

DecibleBreaker

Commendable
Jan 4, 2017
4
0
1,510


Well, I'm using memtest86 from PassMark. I would like to note, its defaulting to v4.3.7, rather than using the v7.2 UEFI boot... not sure why, other UEFI bootable images don't have an issue.

Running the test currently. I've noticed that not even memtest can see both memory sticks at the same time, and even with both connected (and visible in the BIOS) it keeps saying there's only the 4GB of RAM. I'll be running the test separately on each of the two sticks.

More news, there are times where both sticks fail to boot with the "no memory" beep. It appears mostly in slots a1 and b2 that the memory isn't detected (slots are organized... a1|a2|b1|b2). Key word being "mostly". I'm leaning more towards it not being the sticks itself being something else.
 

DecibleBreaker

Commendable
Jan 4, 2017
4
0
1,510
So memtest has come back, no errors found on either stick.
1 pass on each stick using the "sequential" setting for CPU cores, so each test was run 4 times, once by each core - effectively 4 passes.

I can run it more, id get probably 7-8 full passes(x4) per day not including the time id be using the computer, but I'm still thinking it has nothing to do with the RAM itself.

I have an old mobo that I'm thinking of testing with... but... that one has a failing SATA controller, so it won't be a permanent fix. At most it'll either rule out, or single out the motherboard as the issue.
 

Tumeden

Honorable
Oct 15, 2016
449
0
11,160
The memory control hub could be on its way out the window, which sadly means you may need a new motherboard.
This is exactly how my last motherboard fried on me, the memory control hub, or "north bridge" got too hot and left scorch marks. My computer would still start, but it left me with frequent random restarts and Memory Management BSOD's.
 
Solution

DecibleBreaker

Commendable
Jan 4, 2017
4
0
1,510


Symptoms sound the same. This mobo is only 2 months old, so I was leaning less against that. There doesn't appear to be any signs of damage, and the board is supposedly capable of handling 2133, while I'm only running it at 1600... so I feel like that shouldn't be the issue.