Intermittent BSOD's Graphics Card on its way out?

Tavster · 16 Jul 2013 at 16:15

Hello All,

Looking for some advice really, over the last few weeks i've been experiencing random BSOD's whilst using my computer, at first it was every few days, but now I am getting atleast 2-3 a day.

Primarily happens whilst gaming but also happens when idling at desktop or browsing the web.

Specs are as follows:

i5 2500k 4.2Ghz
8GB Corsair XMS3 DDR3
OCZ Agility 120GB SSD
500GB WDC HDD
HIS 6970HD 2GB
Windows 7 64 Bit

I have used memtest and the RAM passed without any issues and have also carried out checks on both of my hard drives with no problems to report.

Some of the errors I am getting are as follows:

atikmdag.sys
dxgmms1.sys
memory_management
pfn_list_corrupt
page_fault_in_nonpaged_area
system_service_exception

I have spoken to a few people who all seem to think that my problems are more than likely down to an issue with the graphics card rather than anything else.

I'd appreciate any answers you guys can provide as I have exhausted my knowledge on this and google results bring back hundreds of possibilities.

Regards,

Tavster.

snips86x · 16 Jul 2013 at 17:51

What PSU do you have?
Have you used BSOD viewer to read the mini dump file?

Tavster · 16 Jul 2013 at 18:13

It's an OCZ Stealth extreme 750W.

And no i've not used it, what would I need to do?

snips86x · 16 Jul 2013 at 18:21

Tavster said:
It's an OCZ Stealth extreme 750W.

And no i've not used it, what would I need to do?

Download Blue Screen Viewer and once installed, you can read the mini dump file to find the cause. If the feature is still available, you can right click on the error and search using Google to find a fix.

wazza300 · 16 Jul 2013 at 18:23

quick easy test would be to takeout the gpu and use onboard graphics,if no crashes you know its graphic card

id guess at memory looking at those errors,using x.m.p profile?

what motherboard are you using? might need one/wto clicks more cpu imc voltage

Rroff · 16 Jul 2013 at 18:27

wazza300 said:
quick easy test would be to takeout the gpu and use onboard graphics,if no crashes you know its graphic card

Not necessarily - these type of errors tend to be memory related as per the 2nd part of your post but in some cases changing GPU could result in the system working around the issue and/or not the same memory operation thats causing the crash happening as frequently and not crashing while the problem memory location(s) is still there.

The GPU driver is one of the most realtime intensive parts of the system so is often the first thing to go even tho the problem is often at a lower level and I'd suspect either failing RAM or some other part of the RAM/memory controller failing - could just be insufficent cooling with the current temperatures.

wazza300 · 16 Jul 2013 at 18:31

I would put

memory
ssd
psu before the graphic card imo

esp if its randomly crashing if gaming or just web browsing

be sure to use latest ssd firmware too and try and catch the bsod codes,it will be something like 124 followed by 0000000000000000000000000x0000000

you can also downclock the gpu memory in ccc by 200mhz and see if it still crashes,gpu memory is usually the first thing to crash esp in hotter weather

latest mb bios wouldn't hurt either

Nickolp1974 · 16 Jul 2013 at 19:17

This

Rroff said:
The GPU driver is one of the most realtime intensive parts of the system so is often the first thing to go even tho the problem is often at a lower level and I'd suspect either failing RAM or some other part of the RAM/memory controller failing - could just be insufficent cooling with the current temperatures.

And this

wazza300 said:
I would put

memory
ssd
psu before the graphic card imo

I would clean your gfx drivers out and reinstall
Reseat your ram and run memtest

Rroff · 16 Jul 2013 at 19:27

Hes already run memtest tho I still find those errors are more likely RAM related - most of the BSOD errors aren't that commonly caused by GPU hardware failing tho could be GPU driver problems or corrupt files possibly.

Nickolp1974 · 16 Jul 2013 at 19:35

Rroff said:
Hes already run memtest tho I still find those errors are more likely RAM related - most of the BSOD errors aren't that commonly caused by GPU hardware failing tho could be GPU driver problems or corrupt files possibly.

So he did, i should read more carefully!!

I used to get the same error but cant remember how i solved it, pretty sure it was because i had oc'ed my ram with too tighter timings.

Tavster · 16 Jul 2013 at 19:43

Would a clean format of both my SSD and HDD be worth trying?

wazza300 · 16 Jul 2013 at 19:48

try the easy stuff first like updating drivers,tweaking memory,testing gpu

Rroff · 16 Jul 2013 at 19:49

Try reinstalling GPU drivers and monitor CPU, GPU and possibly HDD temps incase the cooling is failing or insufficent for the warmer weather.

Tavster · 16 Jul 2013 at 19:49

Have updated drivers to no avail, have even tried older ones.

A friend is bringing over a 650ti tomorrow for me to test for a few hours, just going through the paces of what I usually play, Smite, CS:GO, BF3 etc.

What are the ideal memory settings i'd be looking at?

wazza300 · 16 Jul 2013 at 19:54

whats rated speed/timings dram voltage for them?

one or two clciks more cpu/imc voltage might help even if its error free in memtest

stock imc voltage is 1.05v so maybe 1.06-1.08v and retest

still don't know what board your using,if asus its called vccio or vccsa depending on the model,if gigabyte its cpu/vtt,no idea for others