Win7, 64 Bit, BSOD - mostly 0x50

[strickla]

I've been getting random BSOD's. I've installed windbg and analyzed them as much as I can to try to determine whats going on, but I think I've hit the limit of my knowledge.

I ran RealTemp with P95 stress test to check if I was having overheating issues - I'm not. Normal temps at idle are around 36-37c for all cores. Under full stress they maxxed out at ~60c.

I ran Memtest 86+ for one pass which did report 2 failures in test #7 at 2 different addresses. What does that imply? Do I need to run more passes?

I've attached a bunch of minidumps so others can look through them. Mostly they are 0x50, which is apparently PAGE_FAULT_IN_NONPAGED_AREA. The !analyze -v reports memory_corruption a lot but I have no idea what that "really" means.

This is a brand new computer I built myself from parts purchased at newegg.

Intel i7 920
6 GB Corsair TR3X6G1600C9
2 x 640 GB WD HDD
EVGA nVidia GeForce GTX 260
ASUS P6T 1366 X58 RTL MB

Thanks for all your help in advance - I'd really like to know if this is something I can fix by just updating a driver or stopping a service I don't need, etc. More importantly though, I need to know if I need to RMA this Corsair RAM as that will be a far larger PITA. Thanks again.

 

[Route44]

0x50 errors can be caused by several issues and one of the main ones is corrupted memory.

The unwritten rule with Memtest is to run it a minimum of 7 Passes (more is better).

However, another rule comes into play here and it is: If you get any errors with Memtest you have corrupted memory that must be replaced.

You need to replace your Corsair but the question becomes all sticks, one stick? Run memtest on each stick and another diagnostic step you might try is to check the voltage settings of your RAM in your BIOS and compare it to your designed RAM specs. Many times we see corrupted memory read outs and when in actuality the motherboard has set the memory voltage settings differently than what the memory manufacture designed them for.

First thing, though, is to run memtest with one stick installed at a time.

Edit Update: If this memory came in a set you'll have to RMA it as a set.

Oh! One other thing. Asus is quite notorious for being picky about which RAM can be utilized for every model motherboard they make. Make sure your Corsair is on their approved vendor list for your particular model.

 

[strickla]

Thanks for the reply and the tips.

Last night and today I did as you suggested and ran Memtest with one stick at a time for a minimum of 7 passes each, one with 12 (overnight). 2 of the 3 sticks had no errors. The third stick had multiple errors on each pass.

I ran the tests with each stick in the same mobo ram slot (slot 1) so that I would be sure it was a memory problem and not possibly confuse the issue with a bad ram slot. This also leads me to believe it's not a voltage issue as you suggest, but rather the ram itself...do you agree?

Bummer about having to RMA all 3 sticks. I'm pretty sure I have two good ones and when I get a new set I have no idea about all three. Thats pretty annoying actually.

 

[Route44]

Yes, if you got errors on your third stick that is your issue. It wouldn't hurt to check Asus' vendor list of recommended RAM for your P6T 1366 X58 RTL.

* Keep us up-to-date.

 

[strickla]

Completed the RMA with Newegg. Got the three new sticks of RAM and put them all in - the system has now been up and stable with no BSODs for two weeks.

Route44, thanks for the tips and suggestions.

 

[Route44]

Excellent to hear and thanks for getting back to us. This information helps everyone.