Sign in with
Sign up | Sign in
Your question

[Troubleshooting] bluescreens on 16GB RAM Asus Z68 platform

Last response: in Motherboards
September 12, 2012 12:14:29 PM


CPU - 2600K
Motherboard: Asus P8Z68V-Pro
VIDEO: GTX 570 + GTX 560 physx
RAM: G.SKILL Ripjaws X Series 16GB (4 x 4GB) DDR3-1333
SSD: Crucial M4 128GB
OS: Windows 7 Ultimate x64

History: Built late August, early September 2011. Functioned fine up until about a week ago, over which span of time the following issue grew more frequent and problematic:

Issue: blue screens at random intervals ranging from 15 minutes to 4 hours; afterwards, SSD disappears from post until cold boot.

Steps taken:
1) SSD swapped to different ports and controllers. issues persist. noted that mechanical drive and optical drives did not vanish. at this point, I suspected that the SSD was having issues and was considering an upgrade to Samsung 830 256GB or 512GB.
2) windows reinstallation over existing windows copy. issues persist. (also crashed once during one install attempt).
3) windows reinstallation over command prompt drive overwrite (with diskpart's clean all) . issues persist.
4) memtest on all four DIMM's. 10 errors in 15 hours. it was here that I got a sinking feeling in the pit of my stomach.
5) reset BIOS to default and reconfigured. issues persisted through the next windows session.
6) reduced memory frequency. issues persist.
7) loosened memory timing. issues persist.

I'm really hoping that it's the RAM and not the motherboard. Here's what I plan to do:
1) Remove the GTX 560. Don't think this will help at all, but worth a shot.
2) Upgrade BIOS. I DOUBT this will work, since I was functioning fine up until now, but it's worth a shot.
3) Memtest on single DIMM's. If no errors found, memtest on different combinations of DIMM pairs. If errors found, replace "problematic" DIMM with known good DIMM. If errors still found, replace motherboard.

Anyone have any input on this? I can tell ya right now that I'd much rather simply buy a new computer than spend the next few days memtesting, but whatever.

edit: I also have a 3TB storage drive, and yes, I have tried disconnecting no effect.
a b V Motherboard
September 12, 2012 12:36:39 PM

The blue screen error message and/or the information from the crash dump would be a good place to start.
September 12, 2012 2:28:29 PM

neieus said:
The blue screen error message and/or the information from the crash dump would be a good place to start.

One of the ones I got before I attempted any of the windows reinstallation involved kernel_data_inpage_error, which inspired me to check out my hard drive first. Chkdsk found no errors.

Under startup recovery, it's set to kernel memory dump at windows\memory.dmp, but none of the bluescreens post-Windows-reinstall seem to have generated one. Darn thing also disappears faster than I can snap a photo.
Related resources
September 12, 2012 8:49:28 PM

So here's the memory setup:

DIMM 1 in A1
DIMM 2 in A2
DIMM 3 in B1
DIMM 4 in B2

Again, all four DIMMs running at mfg spec gave me ten errors in 15 hours.

I've reset frequency and timings back up to mfg spec, yanked 1 and 3 out and tested 2 and 4 for roughly five hours. 8 passes. This means:

1) I got lucky.
2) DIMMs 1 or 3 or both are bad.
3) Slots A1 or B1 or both are bad.

Onto the following:

1) I'm going to play around for the remainder of the afternoon to see if it crashes.
2) Overnight, I'll memtest DIMMs 1 and 3 in slots A2 and B2. If there are errors by the time I wake up, I'll run each for a few hours in slot A2 (as recommended by the manual for single-DIMM setups). If there are no errors, I'm going to assume that one or both of the remaining slots are bad and see if I can memtest that.

If it turns out that I am indeed stuck using two DIMMs to keep it running stable, I suppose I'll satisfied. 8GB is still better than any previous computer I've ever owned. I'll gladly take the remaining two DIMMs (if they work) and stick them in another DDR3 machine. WAY better than having to replace the motherboard.

I don't do this for a living anymore. Though not much seems to have changed method-wise, there's enough new stuff that I still appreciate input, especially if I'm overlooking something derpy-obvious.
September 12, 2012 9:34:41 PM

Apparently, I "got lucky" on that last memtest. Despite no errors in 5 hours, I got a kernel_data_inpage_error after about an hour of web surfing. Well, those results go out the window.

Since that popped up again, I'm probably going to go with my original suspicion and try RMAing the drive. I'll check for firmware first, but if it's just doing it now after a year, I doubt it'll help.