BSOD/Memory Issues

[diven]

I have recently received some blue screens of death:
IRQL_NOT_LESS_OR_EQUAL and PAGE_FAULT_IN_NONPAGED_AREA, I received a third BSOD but for some reason I didn't get the dump file created for it. All times playing Starcraft II (though I believe this is due to starcraft being the only large memory application I've been using).

I ran memtest initially it froze on test #5 no errors were reported.
I ran again and it went through 10 passes overnight with no errors.

yesterday it seemed to work fine. But I noticed today that it was showing only 4GB out of the 6GB I have.

I restarted it showed 6 again, went to play starcraft ii and bsod 5 minutes into a game.

I tried memtest again individually on each stick... each passed no errors.

I changed the speed and timings from 1066 7-7-7-19 to 1600 8-8-8-24 and upped the voltage to 1.64v (what the memory is actually rated to)

Again passed for each stick individually. When I reinserted all of them it failed early on (test #3 I believe). I upped the QPI voltage from 1.20v to 1.24v. It went through one pass, but then failed on test #3 again. This time it showed it was testing 6144 - 6656 when it errored and crashed. Since I only have 6GB it is odd that it is testing an extra 512MB.

I reset to auto configuration 1066 7-7-7-19 and noticed it was showing 4GB only again. In the bios it showed dimm slots 1,3,5 occupied (which is correct) and 1 and 4 enabled (which is odd as there is nothing in 4).

I changed back to 1600 8-8-8-24 and upped the vdimm to 1.66v (was showing only 1.614v when set to 1.64v) and qpi to 1.34v. So far system has been stable, but I've only ran a short sc2 game and left it in the menus (where one of the crashes happened) while I was out.

I'm not sure if there is anything more I can do to rule out ram being the issue. And I'm not sure if the issues I'm getting are PSU or motherboard related.

RAM: 3x2GB OCZ3G1600LV6GK
Motherboard: Gigabyte X58A-UDR3
PSU: OCZ Z-series 650W
CPU: i7 930
Video: evga GeForce GTS 250 512MB.

 

[pepe2907]

Usualy this BSOD with IRQL_ blablabla is just a driver related problem

 

[diven]

I've just updated my bios, lan, audio, video drivers, and repaired the starcraft ii installation. Hopefully that will resolve the issue.

 

[marbski]

if not re-install again the system... hope it works...

 

[diven]

Had more issues. Disabled Turbo Performance for RAM. Errored with Prime95. Noticed new bios set default timings to 7-7-7-16. Up those to 7-7-7-20. Ran for 30 minutes no crash or errors. This is very short but much longer than previous.

I noticed my cpu core temperatures getting up to 85C in this test. From what I've read those temps shouldn't cause instability... but it isn't good to run them at that temp for long... and those are high temps for a stock speed i7. Thinking heat sink might need to be reseated, but too late to reseat tonight. I tried dropping the voltages from 1.25v to 1.10v Vcore as I'm told i7 at stock can do stock ok undervolted.

Ran prime95 and voltages seemed to begin to level at 70C. I went away and it ran for 1 hour going through 4 of its tests. I checked prime95 and it died on the 5th test for all threads.At some point my computer bsod as it was restarted when I got back to it.

I suppose it could have failed as a result of undervolting by too much... it could still be a heat issue if the heat crept up over time.

I will reseat the CPU tomorrow, and remove excess power draws (2 extra hard drives) see if either of those improves things. If not I'm out of ideas. I don't think it is a driver issue as the BSOD are happening for a variety of reasons with different bug check codes.

 

[diven]

It seems it was a memory stick after all. Slightly longer memtest revealed it to be one stick causing the issue. I will RMA them.