BFG 285 GTX OC problems

FuriaRi0T

FuriaRi0T

n00b
Joined
Nov 10, 2006
Messages
40
So, I have a rather large issue with Driver failures with my 285 GTX, and I'm not sure if anyone else is having the same problem, or if it's just a power issue.

  • What I have:
  • i7 920 @ 3.8GHz
  • EVGA E758-TR 3-Way SLI Mobo
  • BFG 285 GTX oc
  • 6GB Corsair Dominator ram
  • f4tality Creative Sound-card (PCi)
  • SeaSonic m12 700W Power supply (Link here)

So, what happens:
While playing a game, or at very random times (usually whenever certain graphics come on the screen) my Drivers will fail, then (sometimes) recover.

In certain instances of WoW it will fail, then recover... And it will happen a few times in a row. When it repeatedly fail's/recovers, it will end up failing completely and I will have to manually force-close the program via task manager. Other times, it will just fail then recover, then I won't see the problem again for a while (sometimes hours).

I've also played Battlefield: Bad Company 2 and the drivers will fail, then the program will have to be forced closed after playing a while. They do not recover - it goes to an all white screen.

This same problem happens with Crysis.

In all games, except WoW, there never seems to be a trackable cause for the failure.


Now, It's done this sense WQHL 190.99 drivers (currently using 260.99 - the most up to date & DX-11). I'm unsure if it's a power issue due to my large increase in voltage for the CPU Overclock, but being as the drivers are what fails, and not a BSOD... I doubt it.

Anyone smarter than me have any ideas? :confused:
 
I'm well aware. DirectX 11 is a super-set (DX 10.0/10.1/11). It automatically utilizes the best option based upon the hardware.

*EDIT*
Oh, however I did forget to mention... My system is running Windows 7 Ultimate 64-bit edition.
 
First things first,

Use driver sweeper in safe mode to completely remove all existing driver files and then install the latest drivers again.

Let us know if the problem persists.
 
I did this (uninstalled all drivers via 3Dguru's Driver Sweeper... Rebooted back into safe mode... Installed 260.99 Nvidia drivers) - and the problem indeed persists. I was able to go into Event Viewer (using Crysis as my test subject) and get the data from the crash.

(I apologize in advance for the large amount of data)

Underneath 'System'
Code: 
Display driver nvlddmkm stopped responding and has successfully recovered.

Underneath 'Application'
Code: 
Log Name:      Application
Source:        Windows Error Reporting
Date:          12/23/2010 11:37:07 PM
Event ID:      1001
Task Category: None
Level:         Information
Keywords:      Classic
User:          N/A
Computer:      RagERaiD-PC
Description:
Fault bucket X64_0x117_Tdr:2_IMAGE_nvlddmkm.sys, type 0
Event Name: LiveKernelEvent
Response: http://wer.microsoft.com/responses/resredir.aspx?sid=17274&Bucket=X64_0x117_Tdr:2_IMAGE_nvlddmkm.sys&ID=8e84006d-96ce-4172-97d9-e99e6cd15207
Cab Id: 0

Problem signature:
P1: 
P2: 
P3: 
P4: 
P5: 
P6: 
P7: 
P8: 
P9: 
P10: 

Attached files:
C:\Windows\LiveKernelReports\WATCHDOG\WD-20101210-1833.dmp
C:\Windows\Temp\WER-183890046-0.sysdata.xml
C:\Windows\Temp\WERF10B.tmp.WERInternalMetadata.xml

These files may be available here:
C:\ProgramData\Microsoft\Windows\WER\ReportArchive\Kernel_0_0_cab_0f7b13ae

Analysis symbol: X64_0x117_Tdr:2_IMAGE_nvlddmkm.sys
Rechecking for solution: 0
Report Id: e9c47103-04b5-11e0-b761-001fbc016d17
Report Status: 0
Event Xml:
<Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
  <System>
    <Provider Name="Windows Error Reporting" />
    <EventID Qualifiers="0">1001</EventID>
    <Level>4</Level>
    <Task>0</Task>
    <Keywords>0x80000000000000</Keywords>
    <TimeCreated SystemTime="2010-12-24T04:37:07.000000000Z" />
    <EventRecordID>20689</EventRecordID>
    <Channel>Application</Channel>
    <Computer>RagERaiD-PC</Computer>
    <Security />
  </System>
  <EventData>
    <Data>X64_0x117_Tdr:2_IMAGE_nvlddmkm.sys</Data>
    <Data>0</Data>
    <Data>LiveKernelEvent</Data>
    <Data>http://wer.microsoft.com/responses/resredir.aspx?sid=17274&amp;Bucket=X64_0x117_Tdr:2_IMAGE_nvlddmkm.sys&amp;ID=8e84006d-96ce-4172-97d9-e99e6cd15207</Data>
    <Data>0</Data>
    <Data>
    </Data>
    <Data>
    </Data>
    <Data>
    </Data>
    <Data>
    </Data>
    <Data>
    </Data>
    <Data>
    </Data>
    <Data>
    </Data>
    <Data>
    </Data>
    <Data>
    </Data>
    <Data>
    </Data>
    <Data>
C:\Windows\LiveKernelReports\WATCHDOG\WD-20101210-1833.dmp
C:\Windows\Temp\WER-183890046-0.sysdata.xml
C:\Windows\Temp\WERF10B.tmp.WERInternalMetadata.xml</Data>
    <Data>C:\ProgramData\Microsoft\Windows\WER\ReportArchive\Kernel_0_0_cab_0f7b13ae</Data>
    <Data>X64_0x117_Tdr:2_IMAGE_nvlddmkm.sys</Data>
    <Data>0</Data>
    <Data>e9c47103-04b5-11e0-b761-001fbc016d17</Data>
    <Data>0</Data>
  </EventData>
</Event>
 
Next step is to underclock your card and see if it persists under those conditions.
 
I was actually in the middle of doing this -- It turns out the newer nVidia drivers changed fan-speeds to defaults (rather than taking OC cards into account). I started by manually turning the GPU fan 100%. This helped (some). Rather than fully crashing, it would stutter for a while.

I took the GPU's stock clocks (I have never overclocked it... It was a BFG OC card) and changed them to default based upon nVidias website.

This solved it. After reading [H]jimmy's post about replacing TIM on GPU heatsinks, I'm gonna give this a shot tomorrow and see if I can turn it back 'out of box' clocks that it was origionally at.


Thank you for your help, lostroc888.
 
I have this exact same issue with my BFG GTX275. It's been only happening for me for a couple weeks. My first thought was driver issue, but in the last week I've put together a new system and moved from a Core 2 Quad to a Core i7 and the issue did not resolve. I didn't think to check the fan speed settings. Guess I'll be tweaking a bit now.
 
What are the specifics of your overclock, such as the BCLK and multiplier? When I tried to do 200BCLK×20 multiplier I would get nvidia driver crashes all over the place and it seems that x20 multiplier can be a problematic multiplier on i7's (sounds stupid). I went to 200×21 and I've been great since.
 
djoye,

This card was NOT overclocked by me. It's a BFG 285 GTX OC (came from BFG already overclocked!) It's a factory issue, not a user issue in this case.
 
FuriaRi0T said:
Now, It's done this sense WQHL 190.99 drivers (currently using 260.99 - the most up to date & DX-11). I'm unsure if it's a power issue due to my large increase in voltage for the CPU Overclock, but being as the drivers are what fails, and not a BSOD... I doubt it.

Anyone smarter than me have any ideas? :confused:
Click to expand...

Just how much stress testing have you done with that overclock? Are you sure that it's stable?
 
We've already solved the problem - it's not power at all.

The CPU (i7 920 Stock @ 2.4GHz) is running at 3.8GHz. Been running stable sense the week the i7 was launched (I've had it that long). Prime-95 tested for 48 hours when it was originally done.
 
I have a BFG 280GTX OC2 and had some similar driver errors early on. Reducing the factory memory clock slightly fixed the problem for me. The card could tolerate increasing the GPU clock significantly, but the factory memory clock was not stable. Eventually I re-did the heatsink and the problem went away, I was able to run at the factory overclock settings. I'd suggest:

-Dump 260.99. They are terrible drivers. 197.45 worked perfectly for me. I think the 258.xx drivers are supposed to be good to, but I never tried them. The 260.99 driver feedback thread at nVidia has almost 400k views. I don't think that happens when drivers are stable.

-Double check that the card is properly seated and getting enough power. I had a few problems when I used a splitter/PCIe power converter cable to power the card, solved by using the proper PCIe cables.

-Make sure the fan is running properly and not impeded by lint.

-Consider taking the heatsink off and redoing the TIM. I suspect my poor memory timing tolerance was due to incomplete contact with the memory chips, as it seemed to have gone away after replacing the heatsink.

-If all else fails, you could try baking it. That could work especially if it worked fine before. I wouldn't do that untill you addressed all the above tips, especially getting rid of the 260.99 completely. Worst drivers I have ever had in >10 years with nVidia.

Good luck and merry Xmas.
 
I recently had to suffer through this issue with the nvlddmkm error. Would consistently balk after just a few minutes of running any 3D app. Thought my card (new g470 rec'd from an RMA) was bad also. Did some research and found that the PowerMizer feature in the drivers may be causing the problem. The fix was to make a registry edit in the nVidia video control. Applied the fix last night and my problems seemed to have gone away.

http://www.nvlddmkm.com/?p=56

http://forums.nvidia.com/index.php?showtopic=100800&st=100

1.Open up Regedit and go to \HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\ Control\Video\********\0000\
The *'s will be random numbers. There will be a few of them, pick the one with the most stuff in it.
2.Create the 4 Powermizer keys (these will all be DWORD):
PerfLevelSrc
PowerMizerEnable
PowermizerLevel
PowermizerLevelAC
3.Change the values (double click on them) to:
PerfLevelSrc 3322 (I changed to 2222 to completely disable PowerMizer)
PowerMizerEnable 0
PowermizerLevel 1
PowermizerLevelAC 1
4.Reboot and Powermizer will be disabled!
Click to expand...
 
You must log in or register to reply here.
Back
Top