RAM-related BSOD, memtest86 results

April 11, 2009 at 23:26:17
Specs: Windows XP 64bit SP2
Hi there... My new rig has been suffering from the same BSOD randomly (while gaming, opening the control panel, browsing the internet... etc) (BSOD stop 0x00000050: PAGE_FAULT_IN_NON_PAGED_AREA) and my research has pointed to RAM as a likely culprit, although there are other possibilities.

System Specs:
Core i7 920 @ 2.66 (default)
EVGA 132-BL-E758-A1 X58 Motherboard
EVGA nVidia GTX 260
OCZ Gold 3 x 2GB DDR3 1600 Triple Channel SDRAM
COOLER MASTER UCP RS700 700w PSU
Running WinXP 64bit, SP2

Eventually I found out that I could consistently reproduce the BSOD by starting a virtual machine in VMware, and some quick testing revealed that the BSOD happens when I have any two or more DIMMs installed, in any order. I tested this the following way:

As per my motherboard's instructions, I used slots 1, 3 and 5 for my RAM. The serial numbers of my three DIMMs ended in 1, 2 and 3, so when I say (1,2,x) I mean that I had DIMM 1 in slot 1, DIMM 2 in slot 3, and nothing in slot 5.

Tests:
(3,2,1): BSOD on VMware virtual machine startup
(1,2,3): BSOD on VMware virtual machine startup
(3,2,x): BSOD on VMware virtual machine startup
(3,1,x): BSOD on VMware virtual machine startup
(1,2,x): BSOD on VMware virtual machine startup
(3,x,x): No BSOD - runs fine
(2,x,x): No BSOD - runs fine
(1,x,x): No BSOD - runs fine

As you can see, the BSOD seems dependent upon how many DIMMs are installed, but not on the order or _which_ DIMMs are installed.

Because of this I was tempted to say that faulty memory wasn't the problem, but I had to run memtest to be sure. When I ran memtest86 v3.5 (with DIMMs 3,2,1), it stopped after 3 seconds with the following result:

Test #0, 5%: [Address test, walking ones, no cache]

Lowest error address: 0007f800000: 2040.0mb
Highest error address: 0007fc00000: 2044.0mb
Bits in error mask: ffffffff
Bits in Error - Total: 32
Max contiguous errors: 1
Errors per Memory slot:
0:0
1:22
2:0
I've never played around much with memtest, so I'm not sure why it stopped. But, since I thought faulty memory was probably the cause, I reconfigured to (1,3,2) and ran memtest again. I got the same result, including the same error addresses, even though the DIMMs were in different slots. I got the same result for (1,3,x) and (1,2,x). However, when I ran memtest86 with only one DIMM (1,x,x) (2,x,x) and (3,x,x) it ran fine and for each found no errors.

I ran memtest86 v3.4 with the configuration (3,2,1) and received no errors. I also ran memtest86+ v2.11 for several passes with (3,2,1) and it reported no errors.

Some random facts that may help:
The BSODs that randomly occur while gaming etc cite 'disk.sys' while those occurring when I started a virtual machine in VMware don't - but they're the same stop code and same "PAGE_FAULT_IN_NON_PAGED_AREA".
While running Windows 7 64bit I haven't gotten any BSODs while gaming.

I'm pretty clueless about what's going on here. If it weren't for memtest86 3.5 reporting the errors, I'd say it wasn't a memory problem at all... The fact that memtest 86 v3.5 reports no errors for individual DIMMs and that I don't get any BSODs while using only a single DIMM makes me think... motherboard, possibly? It could also be WinXP 64bit. But if the problem is simply the OS, would memtest86 v3.5 report any errors?

Thanks for taking the time to read (or skim!) all this. Any help with this is _greatly_ appreciated.


See More: RAM-related BSOD, memtest86 results

Report •


#1
April 12, 2009 at 09:55:54
Make sure your memory settings are correct in the BIOS. Do NOT accept the defaults...manually configure the timings & voltage according to the RAM manufacturer's specs.

Also try testing the RAM with Windows Memory Diagnostic & Memtest86+

http://oca.microsoft.com/en/windiag...

http://www.memtest.org/


Report •

#2
April 12, 2009 at 10:47:24
If... possible... borrow a single known to be good stick (min of 128Meg - or does 64Bit really prefer a min of 256 or more ) and see if problem(s) persist...?

Personally I've found direct RAM substitution a better path than running mem-test utils; certainly faster... But doubtless many will disagree with me...

Most definitiely ensure you are using correct RAM for the system - specs. etc. are as they ought be and the rest...


Report •

#3
April 14, 2009 at 02:18:08
Well, with a bit of googling I came up with this:
http://communities.vmware.com/threa...

Apparently the BSOD is standard given my vmware version under XP 64bit. The random BSOD's I had gotten while doing other things (the ones mentioning disk.sys) vanished as soon as I disconnected my old 400gb seagate barracuda. Random, yes. Hopefully this will help anyone else looking for a solution. Thanks for all your answers!


Report •

Related Solutions

#4
April 14, 2009 at 02:27:22
Many thanks for the feeedback... Always appreciated here when folks come back with how they resolved their posted problem(s)...

Report •


Ask Question