Diagnosing the Source "Memory Related Errors" Help Appreciated

isaborders1999

Reputable
Nov 17, 2015
33
0
4,530
Okay, as described in the title, I'm having issues with diagnosing the exact source of these memory errors. I'll try to provide as many details as I can and will provide more if you guys need them. If you need the specs ask, but I think you at most you need to know I have 2x4 Trident 10cas 24000mhz sticks and replaced the psu to see if it was the issue and currently sits with a 1000w EVGA P2, so that rules out the psu. I also use Windows 10 Pro upgraded from 8.1 Pro.


So, here is what I have done on my own to try to narrow this down.

-I think an important thing to note is that every time after coming out of sleep, not on a fresh boot and always sometime after it goes to sleep, it will start producing memory errors.
-Initially before I even caught wind of these problems when my PC was first built, I had issues overclocking, I can say confidently that it wasn't my inability to use the various function in the bios, but when undergoing a stress test I would encounter Memory_Management BSODs along with irql_not_less_or_equal BSODs
-I have done multiple memtest runs, I means multiple, with at least 10 passes and here is what I encountered

  • So I down-clocked everything to make sure it wasn't because of my overclocks before I started.
    Initially I encounter 1 single error when I first built it, and tested it again later and couldn't turn up any errors at all.
    Then, my computer began acting increasingly.... well unpredictable, it would sometimes BSOD and sometimes not come out of sleep correctly if at all.
    So I started testing the RAM for errors, and lo, I turned up more errors 20 or so this time.
    At this point I figured it would be a good point to start testing the sticks individually marking the slots they came from.
    So I took stick one out and left the other stick 2 in slot 4 and it turned up errors, so then I moved stick two in slot 2 and it didn't turn up errors.
    [strike]But, when I moved stick 1 in slot 4, I get no errors.
    Dumbfounded, I put stick 2 back slot 4 and it now has no errors.
    [/strike]
    GO TO BOTTOM POST BY ME, MUST NOT HAVE DOWNCLOCKED UNTIL I WAS AT THIS POINT
    So I move the sticks around in various slots testing them individually and together even overclocking trying to augment the problem. I haven't picked up an error since.
At this point your probably wondering why am I here if the errors are gone. Well, that is the problem, they aren't gone.
I still have weird sleep related errors and even errors when shutting down that read "Memory Break-point Exception Reached" which has what looks like a memory address listed. This maybe unrelated and its not consistent enough if it isn't user generated; my display drivers like to fail and not recover. I have a Nvidia 970 card. I don't think this is that important of information, but I heard system memory errors can cause all sorts of problems.

Anyway, as you can see, I have been giving it some thought and need some reassurance. I think it is ultimately the motherboard, but don't know for sure and would like to hear your suggestions as to what to do. I'm thinking of RMAing the RAM first and if that didn't solve it then the Mobo.

Thanks, I appreciate the help
Cheers!!
 

isaborders1999

Reputable
Nov 17, 2015
33
0
4,530
I will check to see if there is a newer version out, and wait till you respond. Just noticed my specific set of RAM is not in its support pdf but I don't know if that's a good judge because most ddr3 ram sets work in these boards.
 

isaborders1999

Reputable
Nov 17, 2015
33
0
4,530
Looks like there is an F7 version now, and my bios is currently at F6. So if I update the bios and problems still persist, should I consider a fresh install of windows or do you think that has anything to do with it?
 

isaborders1999

Reputable
Nov 17, 2015
33
0
4,530
I have it at 1333 normally, I suppose I can try it if you think it will make a difference. When at stock it defaults to 1600mhz, but I usually don't have it there. I wonder, when I first put this thing together I recall having a close call with thermal compound stringing in the RAM slots but I thought I cleaned it all out. I have looked at it closely and I suppose it won't hurt to look again, but I thought I took care of it a long time ago. I know it isn't a memory controller problem on the cpu because I have given it extra voltage with no change. If it were the RAM you think it would be more consistent. I wonder if it is just a matter of shear bad luck and tiny amounts of left over compound in the slots. Besides, I kind of doubt the integrity of the board anyways because it makes high pitch sequels, although subtle, coming from what I narrowed down as the voltage regulators. Should I just start RMAing the parts first the RAM then the mobo or is there something else I am still missing?
 

isaborders1999

Reputable
Nov 17, 2015
33
0
4,530
I will change the RAM speed to a looser time at 1600mhz and report the results sometime tomorrow. I might also change the DIMM slots to see if my worst fears are correct, see if it is just one slot acting weird. Thanks for your help thus far!
 

isaborders1999

Reputable
Nov 17, 2015
33
0
4,530


I found out the plugin-container.exe error was just a web browser error that needs to be updated. I will not shutdown my computer for a while now, so if I get a blue screen or Nvidia Driver not recovering or something weird, I will report it.
 

isaborders1999

Reputable
Nov 17, 2015
33
0
4,530


[Data Error] Test: 13, CPU: 0, Address: B99E69D0, Expected: FFFFFFFF, Actual: FFFF7FFF x1
[Data Error] Test: 13, CPU: 0, Address: B99E5ADC, Expected: FFFFFFFF, Actual: FFFFFFBF x1
[Data Error] Test: 13, CPU: 0, Address: B99E5E84, Expected: FFFFFFFF, Actual: FFFFF7FF x10
[Data Error] Test: 13, CPU: 0, Address: 22FCCD668, Expected: 00000000, Actual: 00001000 x2
[Data Error] Test: 13, CPU: 0, Address: B99E69D0, Expected: 00000000, Actual: 00008000 ?
[Data Error] Test: 13, CPU: 0, Address: B99E51B8, Expected: FFFFFFFF, Actual: FFFFFFF7 ?
[Data Error] Test: 13, CPU: 0, Address: B99E69D0, Expected: 00000000, Actual: 00008000 ?

All of these errors were in the Hammer Tests

The numbers to the right are the amount of the corresponding error. I think that these errors were augmented by overclocking. I replicated the process today. Downclock none and overclock errors.
 

isaborders1999

Reputable
Nov 17, 2015
33
0
4,530
I figured out that it was one single stick with the errors. I must have forgot to downclock the RAM until that confusing part in my previous test done months ago. Apparently the one stick wouldn't come up with errors until I overclocked the RAM by tightening the timings. I guess if there is one thing to be learned about this is that a controlled environment is key for troubleshooting a computer.

Thanks for your help by the way Tradesman1
Have you tried the DRAM at 1600?
You inadvertently helped me diagnose the RAM problems :D