BSODs and Driver Problems

Proccy OCed?

^yes to 3.1ghz. But it is in that state froms months. And moreover msg comes that ati driver stopped responding and has successfully recovered. So i think a prob related to card.

Happens at load or idle.. ?

Temperature at idle/load.
1)at stock clock.
2)when overclocked.

Jaskanwar Singh said:

Guys these days i am getting a lot of these.

first these started when i used catalyst 11.7. i updated it to 11.8 yesterday but i got these today again. i had overclocked my card to 850/1300 from 800/1250. i was running a unigine test and found lower fps. i checked msi afterburner and found the card running at 500/1300!!!
a BSOD happened and clocks were reset to 800/1250 automatically. then i thought oc might be a problem. so i ran furmark at stock and again a BSOD happened.

i had overclocked it a long time back and no problems were there. started 3-4 days back.

some advice needed what to do?

Click to expand...

The way you have been playing with Furmark, highly probable you have ruined your accelerator.

Anyways, if the BSOD is generating any dumps post the last 5 on a shared link, and place the link here, will look into them.

Also, try a fresh GPU driver install, using Driver Sweeper method as per Guru3d.

asingh said:

The way you have been playing with Furmark, highly probable you have ruined your accelerator.

Anyways, if the BSOD is generating any dumps post the last 5 on a shared link, and place the link here, will look into them.

Also, try a fresh GPU driver install, using Driver Sweeper method as per Guru3d.

Click to expand...

asingh i installed fresh 11.8 after using sweeper.
Sorry for such ques but where are dumps generated?

@tenida
in oc the temp goes around 75 in games and idle 41~.

@megamind
sometimes on windows startup and sometimes gaming.

c:\windows\minidump\

Click to expand...

here.

Sam said:

here.

Click to expand...

thanks sam.

asingh
i am unable to upload. there are 4 files. when i upload it says i dont have sufficient permission to open it.

and now i am unable to play any game or whatever. just recovered from another hang up. the screen fills with white squares or sometime dashes

Jas, reseat the power connectors... Get everything to stock settings(CPU n GPU) n give a try...

yes everything is at stock megamind.

i reseated the card only after that game crash. at first i tried booting from IGP after removing card but it stuck at windows logo. but after installing card its booted. didnt try any game now. shall i?

^^U mean With IGP, PC didnt bootup?

MegaMind said:

^^U mean With IGP, PC didnt bootup?

Click to expand...

Yes, it stuck at win logo.

And before removing card, card was giving dashes even in safe mode. After reinstalling card, i used net. But didnt try game.

Try with IGP first,
1. Remove the GPU,
2. Restore bios to default settings,
3. Boot with IGP, remove ATI drivers
4. Reboot,
5. Install IGP drivers, Try any game with lowest setting for sometime..

post back...

megamind and others an update. today i played civ5 and it ran without problem. this is my first game on gpu after re-installing yesterday. ran heaven too fine.

So what you say? i think its working fine.

^Go-on.. I hope its solved...

@jassy, don't run any heavy benchmark. your GPU has already taken a good lot of beating. just keep your system safe.

PS: i think you remember somone's GPU got fried here after running furmark in high setting. you are lucky

ok sammy.

guys i spoke too early yesterday. another BSOD yesterday evening and today a crash of civ5.

asingh here you go -

082611-34476-01.dmp
082511-29889-01.dmp
082411-55973-01.dmp
082411-28594-01.dmp
082211-29733-01.dmp

jassy all these errors are cause of just 1 single module: atikmpag.sys
and the error is Timeout Detection and Recovery error (TDR).

did a bit of googling & found these:

The following hardware issues are examples that can cause a TDR event:
- Failing overclock on CPU or GPU
- Bad sector in memory resulting in corrupt data being communicated between GPU and the system (either video or system memory)
- Corrupt hard drive/windows install resulting in corruption to the system registry or the page file
- Over heating of GPU or CPU again resulting in corrupt data being communicated.
- GPU failure due to any sort of issue from insufficient power(VERY common) to heat.

The following software issues are examples that can cause a TDR event:
-Incompatible drivers of any sort (either GPU, sound, etc)
-Messy registry
-Bad direct x files
-bad system files
-Bad driver

Click to expand...

first & 4th option looks most likely.

Well, I'm a sufferer of this problem and there are two possible reasons for it:-

Corrupted DirectX file
Corrupted GPU Memory

JAs, do a fresh installation of the OS and then install the Gfx driver. Make sure everything is running @ stock speed.
If reinstalling OS doesn't solve the problem then I'm 99.5% sure it is Gfx card hrdware issue, most probably corrupted memory module.
And why the hell you are running future mark tests all the time? Futuremark is a bad piece of software to measure GPU's capability as it creates unrealistic type of workload over the GPU and stress it in such a high level which is nearly impossible in any real life scenario.
That's why all the review sites actually dropped futuremark tests to measure load power consumtion.

^^Does Futuremark software means (3D mark 06,3D Mark 11,3D mark Vantage and PC mark 7)??

Tenida said:

^^Does Futuremark software means (3D mark 06,3D Mark 11,3D mark Vantage and PC mark 7)??

Click to expand...

yes. but don't mix furmark & futuremark.

i formatted, reinstalled win7 but still issue persists. Time for RMA i guess?

^^Maybe you can get the Dual-fan one HD6950 if replaced.

Jaskanwar Singh said:

i formatted, reinstalled win7 but still issue persists. Time for RMA i guess?

Click to expand...

If possible, pay extra n get 2GB version..

so what anil told at the start is most probably right. the GPU or the memory chips have got corrupted due to excessive benchmarking.

ok guys. Btw contact smc or aditya directly?

Contact Aditya Infotech, Chandigarh...directly.

I looked at the BSOD dumps..and got this...

VIDEO_TDR_FAILURE (116)
Attempt to reset the display driver and recover from timeout failed.
Arguments:
Arg1: fffffa80043594e0, Optional pointer to internal TDR recovery context (TDR_RECOVERY_CONTEXT).
Arg2: fffff880040969c8, The pointer into responsible device driver module (e.g. owner tag).
Arg3: 0000000000000000, Optional error code (NTSTATUS) of the last failed operation.
Arg4: 0000000000000002, Optional internal context dependent data.

Debugging Details:
------------------


FAULTING_IP:
atikmpag+79c8
fffff880`040969c8 4883ec28 sub rsp,28h

DEFAULT_BUCKET_ID: GRAPHICS_DRIVER_TDR_FAULT

CUSTOMER_CRASH_COUNT: 1

BUGCHECK_STR: 0x116

PROCESS_NAME: System

CURRENT_IRQL: 0

STACK_TEXT:
fffff880`03c669c8 fffff880`0413bcbc : 00000000`00000116 fffffa80`043594e0 fffff880`040969c8 00000000`00000000 : nt!KeBugCheckEx
fffff880`03c669d0 fffff880`0413b9c6 : fffff880`040969c8 fffffa80`043594e0 fffffa80`03847930 fffffa80`038da410 : dxgkrnl!TdrBugcheckOnTimeout+0xec
fffff880`03c66a10 fffff880`0400ff07 : fffffa80`043594e0 00000000`00000000 fffffa80`03847930 fffffa80`038da410 : dxgkrnl!TdrIsRecoveryRequired+0x1a2
fffff880`03c66a40 fffff880`04039b75 : 00000000`ffffffff 00000000`000003fd 00000000`00000000 00000000`00000002 : dxgmms1!VidSchiReportHwHang+0x40b
fffff880`03c66b20 fffff880`040382bb : 00000000`00000102 00000000`00000000 00000000`000003fd 00000000`00000000 : dxgmms1!VidSchiCheckHwProgress+0x71
fffff880`03c66b50 fffff880`0400b2c6 : ffffffff`ff676980 fffffa80`038da410 00000000`00000000 00000000`00000000 : dxgmms1!

VidSchiWaitForSchedulerEvents+0x1fb
fffff880`03c66bf0 fffff880`04037e7a : 00000000`00000000 fffffa80`03847930 00000000`00000080 fffffa80`038da410 : dxgmms1!VidSchiScheduleCommandToRun

+0x1da
fffff880`03c66d00 fffff800`03532166 : 00000000`02db0959 fffffa80`0376eb60 fffffa80`018bb040 fffffa80`0376eb60 : dxgmms1!VidSchiWorkerThread+0xba
fffff880`03c66d40 fffff800`0326d486 : fffff800`03407e80 fffffa80`0376eb60 fffff800`03415c40 fffff880`01254534 : nt!PspSystemThreadStartup+0x5a
fffff880`03c66d80 00000000`00000000 : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : nt!KxStartSystemThread+0x16


STACK_COMMAND: .bugcheck ; kb

FOLLOWUP_IP:
atikmpag+79c8
fffff880`040969c8 4883ec28 sub rsp,28h

SYMBOL_NAME: atikmpag+79c8

FOLLOWUP_NAME: MachineOwner

MODULE_NAME: atikmpag

IMAGE_NAME: atikmpag.sys

DEBUG_FLR_IMAGE_TIMESTAMP: 4e166fa8

FAILURE_BUCKET_ID: X64_0x116_IMAGE_atikmpag.sys

BUCKET_ID: X64_0x116_IMAGE_atikmpag.sys


*****************************************************************************************

Debugging Details:
------------------


CUSTOMER_CRASH_COUNT: 1

DEFAULT_BUCKET_ID: VISTA_DRIVER_FAULT

BUGCHECK_STR: 0xA0000001

PROCESS_NAME: System

CURRENT_IRQL: b

LAST_CONTROL_TRANSFER: from fffff88004854591 to fffff800034bbf00

STACK_TEXT:
fffff880`02ffb668 fffff880`04854591 : 00000000`a0000001 00000000`00000005 00000000`00000000 00000000`00000000 : nt!KeBugCheckEx
fffff880`02ffb670 00000000`a0000001 : 00000000`00000005 00000000`00000000 00000000`00000000 00000000`00000000 : atikmdag+0x4d591
fffff880`02ffb678 00000000`00000005 : 00000000`00000000 00000000`00000000 00000000`00000000 fffffa80`00000000 : 0xa0000001
fffff880`02ffb680 00000000`00000000 : 00000000`00000000 00000000`00000000 fffffa80`00000000 00000000`00000000 : 0x5


STACK_COMMAND: kb

FOLLOWUP_IP:
atikmdag+4d591
fffff880`04854591 ?? ???

SYMBOL_STACK_INDEX: 1

SYMBOL_NAME: atikmdag+4d591

FOLLOWUP_NAME: MachineOwner

MODULE_NAME: atikmdag

IMAGE_NAME: atikmdag.sys

DEBUG_FLR_IMAGE_TIMESTAMP: 4e31d49f

FAILURE_BUCKET_ID: X64_0xA0000001_atikmdag+4d591

BUCKET_ID: X64_0xA0000001_atikmdag+4d591

Followup: MachineOwner
*******************************************************************************************
*******************************************************************************
* *
* Bugcheck Analysis *
* *
*******************************************************************************

Unknown bugcheck code (a0000001)
Unknown bugcheck description
Arguments:
Arg1: 0000000000000005
Arg2: 0000000000000000
Arg3: 0000000000000000
Arg4: 0000000000000000

Debugging Details:
------------------


CUSTOMER_CRASH_COUNT: 1

DEFAULT_BUCKET_ID: VISTA_DRIVER_FAULT

BUGCHECK_STR: 0xA0000001

PROCESS_NAME: CivilizationV_

CURRENT_IRQL: a

LAST_CONTROL_TRANSFER: from fffff8800486e591 to fffff8000347df00

STACK_TEXT:
fffff880`08f8f788 fffff880`0486e591 : 00000000`a0000001 00000000`00000005 00000000`00000000 00000000`00000000 : nt!KeBugCheckEx
fffff880`08f8f790 00000000`a0000001 : 00000000`00000005 00000000`00000000 00000000`00000000 00000000`00000000 : atikmdag+0x4d591
fffff880`08f8f798 00000000`00000005 : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000100 : 0xa0000001
fffff880`08f8f7a0 00000000`00000000 : 00000000`00000000 00000000`00000000 00000000`00000100 00000000`00000000 : 0x5


STACK_COMMAND: kb

FOLLOWUP_IP:
atikmdag+4d591
fffff880`0486e591 ?? ???

SYMBOL_STACK_INDEX: 1

SYMBOL_NAME: atikmdag+4d591

FOLLOWUP_NAME: MachineOwner

MODULE_NAME: atikmdag

IMAGE_NAME: atikmdag.sys

DEBUG_FLR_IMAGE_TIMESTAMP: 4e31d49f

FAILURE_BUCKET_ID: X64_0xA0000001_atikmdag+4d591

BUCKET_ID: X64_0xA0000001_atikmdag+4d591

Followup: MachineOwner
---------

Click to expand...

Can you update your Dx flavor to the latest. Just overlay it. Also do a clean sweep, yes again of the CCC.

Let us know.

guys had a pm discussion with asingh. updated every driver including chipset, gpu.
and played civ5 for 1 hour 15 min without problems.

so does every benchmark affects the gpu?

^^ excess of everything is bad. you benchmarked your GPU a bit too hard. first of all you overclocked it & then stretched it to its limit.

BTW GPU still overclocked or running at stock now?

its running at stock now sam.

^^ Are you getting more bsod's? If yes, then RMA seems to be the best option. Cilus faced similar problems and found out that some of his memory modules were corrupt.
Sapphire has some excellent service in kolkata and he got a new 6870 within 3 days. Same must be in your place and 6950 is a high end card and therefore has high priority in the replacement queue.

P.S- After you get your new card, don't stress it anymore , whether benchmarking or overclocking. Use it at stock because 6950 is a very powerful card at stock & you don't need to overclock.

When games start becoming heavy, then you can overclock a bit. But i think all new game releases this year will run flawlessly in a 6950 at fullhd. So your gaming needs are completely covered.

thanks batman. no more probs (fingers crossed)

as my card is gone for RMA, i was playing civ5 at low-med settings at 1366x768 at non full screen on my IGP.

yesterday it started to lag and i checked GPU-Z. showed this -
*img713.imageshack.us/img713/421/capturegpn.png

memory is running at low speed.

but afterburner showed this -
*img17.imageshack.us/img17/6728/capture2sf.png

and what is GPU-1 and GPU-2 in afterburner?

and no i didnt run any benchmark on my IGP

Doubt after burner will realize and IGP properly.

card sent for RMA?!?!
what happened?

baba bsods and crashes during gaming.

asingh IGP problem gone away but still gpu z shows less memory frequency. so these softwares must be optimized of discreet cards. what say?

Jaskanwar Singh said:

baba bsods and crashes during gaming.

asingh IGP problem gone away but still gpu z shows less memory frequency. so these softwares must be optimized of discreet cards. what say?

Click to expand...

OK....was reading the thread
now why did you do OCing and running testing s/w like furmark so frequently?
heck...even those hardware reviewers don't do this stuff that frequently

btw from where did you bought this card?

Piyush said:

OK....was reading the thread
now why did you do OCing and running testing s/w like furmark so frequently?
heck...even those hardware reviewers don't do this stuff that frequently

btw from where did you bought this card?

Click to expand...

no baba. i didnt run furmark like you mentioned. but i did make a mistake to run it whenever i ran it.

smcinternational.

.GUYS GOT THE CARD
and u know what, i have got a dual fan sapphire cooler now. The vram is 1gb. May they changed the cooler. The box also same.

Will post pics soon.

^^ Congrats jas. Now don't push the card too much. Play at stock because 6950 is a pretty powerful card out of the box.

Enjoy.

U're lucky Jas...

*img26.imageshack.us/img26/3706/image0181f.jpg

more images -

ImageShack Album - 8 images

thanks megamind and batman.

Congratulations JS bro the card looks great..enjoy

thanks max buddy.

guys i found the product -
*www.sapphiretech.com/presentation/product/?cid=1&gid=3&sgid=1041&pid=0&psn=000101&lid=1&leg=0

its SAPPHIRE HD 6950 1GB GDDR5 PCIE
*www.sapphiretech.com/presentation/product/?cid=1&gid=3&sgid=1041&pid=1271&psn=&lid=1&leg=0

AND IT HAS A DUAL BIOS SWITCH!!

Newegg.com - SAPPHIRE 100312-1GDP Radeon HD 6950 1GB 256-bit GDDR5 PCI Express 2.1 x16 HDCP Ready CrossFireX Support Video Card with Eyefinity

to be more specific -
current - SAPPHIRE 100312-1GDP
Newegg.com - SAPPHIRE 100312-1GDP Radeon HD 6950 1GB 256-bit GDDR5 PCI Express 2.1 x16 HDCP Ready CrossFireX Support Video Card with Eyefinity
newegg mentions a dirt3 box though.
previous - SAPPHIRE 100312-1GSR
Newegg.com - SAPPHIRE 100312-1GSR Radeon HD 6950 1GB 256-bit GDDR5 PCI Express 2.1 x16 HDCP Ready CrossFireX Support Video Card with Eyefinity

installed pics -
*img20.imageshack.us/img20/4448/027fdf.jpg
ImageShack® - Online Photo and Video Hosting
ImageShack® - Online Photo and Video Hosting

and thanks to everyone who helped

^^ congrats - this thing will remain much more cooler and will offer more performance headroom

Thanks TP. The cooler is nice. During deus ex human rev at highest settings with mlaa etc i got a max temp of 72c at default fan profile (at that temp it was 54%).

No more fooling around with Card for you

^well i will try after long time

Now enojoy gaming jassy.

thanks dark night.

Piyush said:

OK....was reading the thread
now why did you do OCing and running testing s/w like furmark so frequently?
heck...even those hardware reviewers don't do this stuff that frequently

btw from where did you bought this card?

Click to expand...

Only high overclocking would results in such problems.

@Jas what was your stock and overclocked speeds?

stock - 800/1250
oc - 875/1300 (stock volt)