Riddled With Bluescreens - 0x0F4

npsgaming

Distinguished
Mar 31, 2009
146
0
18,710
I am out of ideas this time. For several months I have been plagued with genuinely random bluescreens. I am now looking for help from the community.

Info:

Upgrades:
#1 Bought 2 additional 120Gb SSDs from Corsair (Force GT3) for a Raid 0
#2 Bought 6 Corsair dimms to update memory (48Gb Dominator Platinum)
#3 Bought 3 GTX670's for 3 Way SLI( EVGA P/N: 04G-P4-3673-KR)


The problems began after the 3rd upgrade. Random bluescreens when gaming, browsing, or even typing documents. BlueScreenViewer points to repeated crashes of ntoskrnl.exe or nvlddmkm.sys of which I know nvlddmkm is an nvidia driver component. The bluescreens error codes were 0x0F4 and a locale ID of 1033. Everywhere I found online points to this being ram based. So I ran memtest on all 48Gb for 20 hours(3 passes) with no errors. Then I proceeded to test each stick individually for 5 hours (3 passes each). Still no errors. My next thought was that a video card was bugged. So I tried them as single cards 1 by 1 and found 2 that were bugged (repeated driver crashes and memory corruption). Replaced those. On the next set another was DOA. And now as I write this post I'm on my 4th SET of replacements. I'm an unlucky person but 4 sets of dead cards seems cosmically improbable at best. All 3 cards do great on their own now (no SLi of any kind). Unfortunately the problems are still present when I have more than 1 card installed. I saw a post on EVGA forums about updated firmware for the GTX670 class cards. I tried updating firmware but that did not seem to fix the problem.

After all this I am considering the SSDs are the problem. The drives however have not given any trouble in the past (1 was previously operating 3x9800GTX+ and another was in a rig with 2xGTX560Ti cards).


Drivers are up to date and everything seems updated and in place.


Specs:
3x Corsair Force GT 3 120Gb ssds Raid 0
3x 7200rpm hard disks
1x BD-Rom
6x 8Gb Corsair Dominator Platinum DDR3 1600Mhz
1x Intel Core i7 960 3.33Ghz (Corsair H80)
3x EVGA 04G-P4-3673-KR GTX670 4Gb FTW+ cards
1x Corsair AX1200 psu

So my question is: What in the world can be causing these problems?!?!?!?!?
 

USAFRet

Titan
Moderator
So my question is: What in the world can be causing these problems?!?!?!?!?

Multiple video cards, multiple fails.
Just a wild guess, but could it be the slot(s) on the motherboard? Try shuffling the cars around between the various slots. Is it always the same card, or always the same position?
 

npsgaming

Distinguished
Mar 31, 2009
146
0
18,710


I had thought about that but this same board has housed 3 9800GTX+ cards, 2x GTX560Ti cards, and then 2x GTX670 cards all without a problem until the 670's then a few random bsods which were driver related.
 
First - let me say I am jealous of your rig :)

To troubleshoot this, I would strip it down to the basics - 1 RAM stick and 1 GPU. Then run Prime95 on the system for at least 24 hours to see if errors are generated. If all is good, try adding in one item at a time (I would suggest the RAM sticks first), and re-run the tests. I would seriously doubt that you have so many defective cards....
 

npsgaming

Distinguished
Mar 31, 2009
146
0
18,710


Well it would only be worthy of jealousy if it worked :D . And the strangest thing is it has no problem with prime95. Memtest didn't flag any problems on the ram sticks when tested one by one or as a group. And that was with several passes too. I suppose I could try prime95 with each stick one by one. And I've never had any trouble with evga products until these graphics cards. I'm losing a little bit of faith in them as the first 2 cards were bugged 1 from the next set was bugged and the 3rd wouldn't even power on x.x .
 

onichikun

Distinguished
Nov 13, 2009
304
1
18,860
Have you monitored your rails voltage with all 3 cards installed? Do you have the proper power hookups?

Try a different nvidia driver version. It sounds to me like either a power issue or a driver issue with those cards.

GLHF.
 

npsgaming

Distinguished
Mar 31, 2009
146
0
18,710


Yes I am also in agreement. I just don't know what else could be affecting the system so much. Additionally the dump files I am getting also indicated nvidia drivers or nvidia hardware failure. I just don't know what can be causing the problem. Would a dying cpu cause these problems?
 

npsgaming

Distinguished
Mar 31, 2009
146
0
18,710


Its in the OP but I'm running a Corsair AX1200. The cards are wired with genuine 6/8 pin cables not adapters etc. and the power seems pretty clean. The AX1200 only has 1 12v rail and according to the logging its lowest value was 12.014 and highest was 12.052v Even at full load the power doesn't fluctuate much.
 
Do you have a page file, if not create one. Did you have only one ssd to start, and kill the pagefile to save space? it should be set to 1.5 times the amount of ram you have. The page file does not have to go one the primary hard drive if you have more than one, although I would recommend it does go on your fastest hdd.
If you already have one follow the directions below to remove and recreate it.
http://forums.techguy.org/windows-xp/686085-solved-bsod-stop-error-0x0f4.html

the error indicates CRITICAL_OBJECT_TERMINATION f4 indicates that a process or thread crucial to system operation has unexpectedly exited or been terminated (Pointer to an ASCII string containing an explanatory message ).



How to stop blue screen errors in windows 7
 

npsgaming

Distinguished
Mar 31, 2009
146
0
18,710


I had 1 ssd to start and when I got the 2 new ones I did a clean install of win7 at which time I did nuke the PF off the Raid 0 and setup my 3rd Corsair ssd as a dedicated "PF/Swap" drive. The Crit_obj-term is the error I get (obviously as its the f4 code) so I don't know where to go from here. I'm considering calling an exorcist or something :ouch: