CPU speed, BSOD and backlit keyboard problem - Optimus V 15"

[mrducking]

i would go for the windows reinstall, less headaches more time efficient

 

[kizcat]

Yeah, I think I'm going to have to consider it

 

[kizcat]

Okay, new problem!

My dGPU will no longer run any games or applications. The Nvidia Control Panel reverts everything back to Auto-Select: Integrated, even launching a game through the NVIDIA GeForce Experience console doesn't seem to work. Everything is running on the iGPU. and I have pressed the VGA button a few times just to make sure it's not that. GeForce Drivers running on version 344.11, just updated the geforce experience thing as well. and it worked fine yesterday on these graphic drivers.

Any ideas on what to do? (other than complete reinstall of OS/RMA - because they're very very last resorts)

 

[SpyderTracks]

Okay, new problem!

My dGPU will no longer run any games or applications. The Nvidia Control Panel reverts everything back to Auto-Select: Integrated, even launching a game through the NVIDIA GeForce Experience console doesn't seem to work. Everything is running on the iGPU. and I have pressed the VGA button a few times just to make sure it's not that. GeForce Drivers running on version 344.11, just updated the geforce experience thing as well. and it worked fine yesterday on these graphic drivers.

Any ideas on what to do? (other than complete reinstall of OS/RMA - because they're very very last resorts)

A fresh windows install is the first thing, it's possible it's a faulty graphics card, but until you disprove any other driver conflicts etc, it's impossible to tell with certainty.

 

[Boozad]

I've looked at your dump file but it's hard to get the full picture from one dump. Anyway, it's a 0x116 VIDEO_TDR_ERROR (Timeout Detection Recovery). This can be down to the driver, overheating or the GPU itself. Normally video errors on laptops with switchable graphics blame both the iGPU and the dGPU, but yours just blames the nVidia driver.

*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

Use !analyze -v to get detailed debugging information.

BugCheck 116, {fffffa8007191010, fffff8800f3e45c8, 0, 2}

Unable to load image \SystemRoot\system32\DRIVERS\nvlddmkm.sys, Win32 error 0n2
[B][COLOR="#FF0000"][B]*** WARNING: Unable to verify timestamp for nvlddmkm.sys
*** ERROR: Module load completed but symbols could not be loaded for nvlddmkm.sys[/B][/COLOR][/B]
Probably caused by : nvlddmkm.sys ( nvlddmkm+1575c8 )

Followup: MachineOwner

6: kd> lmvm nvlddmkm
start             end                 module name
fffff880`0f28d000 fffff880`0ff61000   nvlddmkm T (no symbols)           
    Loaded symbol image file: nvlddmkm.sys
    Image path: \SystemRoot\system32\DRIVERS\nvlddmkm.sys
    Image name: nvlddmkm.sys
    Timestamp:        [B][COLOR="#FF0000"]Sat Sep 13 20:59:50 2014[/COLOR][/B] (5414A236)
    CheckSum:         00C8E3A6
    ImageSize:        00CD4000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4

Looking at the timestamp you have the latest driver installed, it may be worth trying an earlier driver (331.65 is probably the most stable driver working back) as the latest drivers aren't always the best choice or most stable option. Follow this for the cleanest possible PU driver install. Uninstall GeForce Experience too.

If you haven't run CCleaner at all since receiving the BSODs I can provide you with a tool that will collect all of your dump files together so I can take a look, it's completely up to you. Also have you checked your temperatures when your GPU is under load? Sorry if you've mentioned that, I'll admit to skipping lots of the thread.

 

[kizcat]

Hmm, well (before your post) i did just blanket uninstall anything nvidia because the dGPU just wasn't working at all, I then installed driver 340.52 as that was the last working one I had previous to this one that seemed stable. So far, there have been none of the original issues with it.

The BSOD only happens under very specific circumstances, being when 1 core is maxxed out at 100% by what appears to be the 'System Interrupts' process and I/an application try to use the dGPU. This isn't happening when the dGPU is under any load at all.

and I have no recorded temperatures because it has never happened under any load or stress because I've never managed to get that far in this very specific circumstance.

I will keep using my laptop as normal and if the issue persists then I will upload any further dumps/information I get. As it stands there are no dumps at all of the previous BSOD(s), even the one I have already uploaded has disappeared, I think possibly when I did a (failed) system restore.

 

[SpyderTracks]

Hmm, well (before your post) i did just blanket uninstall anything nvidia because the dGPU just wasn't working at all, I then installed driver 340.52 as that was the last working one I had previous to this one that seemed stable. So far, there have been none of the original issues with it.

The BSOD only happens under very specific circumstances, being when 1 core is maxxed out at 100% by what appears to be the 'System Interrupts' process and I/an application try to use the dGPU. This isn't happening when the dGPU is under any load at all.

and I have no recorded temperatures because it has never happened under any load or stress because I've never managed to get that far in this very specific circumstance.

I will keep using my laptop as normal and if the issue persists then I will upload any further dumps/information I get. As it stands there are no dumps at all of the previous BSOD(s), even the one I have already uploaded has disappeared, I think possibly when I did a (failed) system restore.

The latest update only had compatibility additions and new feature sets for the upcoming 900m series and 800m series, so there's absolutely no benefit to having it. If it's stable on the previous release, then at least that gives more scope to find out the underlying issue. Keep us posted with any progress.

Bravo, Boozad for processing the dmp file.