New GeForce drivers issue!(blank screen)

I have this irritating issue
whenever i try to update my drivers(whql/beta)
i always goes blank screen on installation(clean/overwrite)

the only stable version i found for me is the 314.14( the 314.22 works but 'breaks' the 3D)
https://forums.geforce.com/default/...-with-the-314-22-drivers-i-ve-lost-3dtv-play/

every time i have to boot into safe mode, remove drivers, again boot into safe mode, and again install 314.14 and then i'll be able to get to the desktop

i would also like to mention that i have flashed a modded vbios from tech inferno, could this be the root of all problems?
Downloads

it is really irritating because, i really cant benefit from these updated drivers for newer games(tomb raider, bioshock infinite, etc)

would really appreciate any help!
Thanks in advance!

 

R4 Dell card makes matters worse, it's not a Clevo vbios problem as I thought but probably modded vbios and 3D Vision certification, even w/ nvidia stereo controller disabled in device manager . Mine M17XR3 Clevo card 80.04.33.00.10 'OCedition' revised_01 vbios, Win 8 http://forum.notebookreview.com/alienware-m17x/672230-m17x-r3-680m-compatability-80.html#post9196483.

Try here [MOBILE] Modded nvdmn.inf v.320.14 Alienware M17XR3/M18XR1

 

thanks for the reply!

but it doesnt have modded inf for M17XR4
will the M17XR3 680m do?

and i cant download from techinferno, not enough posts!
thanks for the help anyway, i'll check it out

 

I'm having the same problem, R4 doesn't need a modded .inf the card is officially supported but you can post there as it's also driver release related or here NVIDIA Kepler VBIOS mods - Overclocking Editions, modified clocks, voltage tweaks

 

flashing stock bios does not help

still getting a blank screen when installing drivers

staying on 314.14 for now

really cant figure out what the problem is

 

Actually that's good news, we can keep using current unlocked OC vbios. Nvidia must fix this, try to post here https://forums.geforce.com/default/...hread-released-5-13-13-/post/3809245/#3809245

Obs. Clevo P370EM 3D w/ GTX 680M, same issue https://forums.geforce.com/default/...hread-released-5-13-13-/post/3809487/#3809487

Nvidia didn't test 320.xx drivers properly with the marketed 3D Vision laptops or simply didn't give much attention to this, very busy optimizing the TITAN.

 

hahahaha! so true! after that gtx titan, all efforts are going into that, no love for us 3D supporting people!

and thanks for the thread, will post there soon

 

Bug Filed https://forums.geforce.com/default/...hread-released-5-13-13-/post/3810004/#3810004

Depending on the cut bug fix could be available with the next release.

 

that sure is exciting news!
its good to hear that the problem has been acknowledged!

 

Hey peeps long time no post. Don't worry guys, I have the same. 3D + Clevo with custom rom. The driver recognises any external monitors as primary therefore our display is blanked out. I'm still on 314.22

 

No external display attached and clean installed (manually deleted registry keys/folders, no PnP external monitor hidden driver) R4 3D gtx 675m installs fine. Clevo P370EM 3D + gtx 680m...same issue; specific 3D Vision + Kepler gtx 680m bug.

..........

 

never really tried an external monitor when installing the new drivers

do all the displays go blank, or the laptop only?

 

Laptop only.

 

Problem persists with v. 320.18 WHQL (didn't meet the cut) it would be wise to post here https://forums.geforce.com/default/...hread-released-5-23-13-/post/3815589/#3815589 so they can give priority to this, system spec. all stock... It seems that DELL is completely unaware of it.

 

*Self Removed*

 

i didnt have much hope for 320.18
it was just a modified 320.14 to just include the new GTX 780 in various games and in GeForce experience as they wouldnt recognize it

 

It's only Microsoft WHQL certification and GeForce experience optional install, .inf files 780m no entries. More reporting welcomed https://forums.geforce.com/default/...hread-released-5-23-13-/post/3816054/#3816054 Clevo user https://forums.geforce.com/default/...hread-released-5-23-13-/post/3816123/#3816123

"We are having a hard time reproducing this issue. We are continuing to try different combinations" comprende' more....

 

well, its good that they are working on it

but i was referring to the desktop GTX 780 which was released this week
AnandTech | NVIDIA GeForce GTX 780 Review: The New High End

 

Just included @MickyD1234 as an example M17XR4 3D Fermi GTX 675M successful driver 320.18 installation https://forums.geforce.com/default/...hread-released-5-23-13-/post/3816777/#3816777 your reporting will help us climb Nvidia's Priority List which at this moment is endless thanks to 32x.xx. Nvidia acting RED.....

 

Same black screen for me, plugged in tv with hdmi and rolled back to 314, disappointing huh!

 

good to hear that someone worked out a solution for the issue,
did anyone try to install the way @MickyD1234 did?

 

No, that's to prove specific 'Kepler' GTX 680M / 3D Vision bug, @MickyD1234 R4 'Fermi' GTX 675M / 3D Vision, installs fine https://forums.geforce.com/default/...hread-released-5-23-13-/post/3816777/#3816777

 

sorry, needed some time to get my head around it
had forgotten that 675m is a re branded 580m

is it just specific to kepler 680m with 3D?
some people have 660m with 3D, do they also report the same problem?

and yes, nvidia does seem to be turning a bit 'red'

 

https://forums.geforce.com/default/...hread-released-5-23-13-/post/3817251/#3817251

and #217

 

Different chip GK107 GPU, haven't seen any reports yet.

rjtnag ?

 

Yup guys, the only problem I had was the lockup (no input) that the FF acceleration was causing. I have tested most titles (Inc. BF3) and feel like I dodged a bullet. I can even roll back easilly.

Shame on NV, it was looking so good for the last year or so. If AMD come out with a slick, killer driver before NV fix this it's over and out for their market advantage

 

Well, driver was released two days ago; 16 pages already. :thumbsup:........

 

i meant the posts by the user 'chtan' #212
also reporting the same problem
too bad he re installed windows
could have resolved it with a safe mode boot

 

Yeah I know, thought it might be you; wrong.

 

It's taking on the shape of a disaster with cards having HW problems now: https://forums.geforce.com/default/...read-released-5-23-13-/25/?offset=364#3818426

I was never happy seeing a .92v boost on my card since 3xx release since it has been known to burn my card with this OV, but just a few seconds seemed OK. Looks like cards that are already factory OC'd are cracking up, nooooo

 

well that is quite dangerous, jumping voltages are not nice
do you use a modified vbios?

nvidia really messed up their previous gen cards for gpu boost 2.0
hope that it gets fixed up in the next release

even the desktop people with 3D are experiencing problems #369

 

I do have a modded vbios, but only for P1 to match P0 so the voltage drop does not occur if it power throttles. No increases in clocks or voltage (and P8 still drops correctly to 0.80V). I use nvidiaInspector to control the P0 and P1 clocks but this occurs at stock and was noted by stock vbios users as well

 

Remember http://forum.notebookreview.com/ali...o-need-modify-system-bios-25.html#post8438863
Nvidia Drivers v. 320.18 Microsoft WHQL certification..... ..... .....

 

Oh yes, the bane of my life back when I had the R3/580. The issue I still had with this 675m was that if I did not use a modded vbios, the voltage would drop at P1 and make it unstable with an OC - and P1 occurs a lot more often when overclocking P0

On the plus side I note that as a driver gets optimised for a specific game P1 throttle occurs a LOT less in that game (and temps are lower)

 

"We continue to investigate the causes of people's issues. From what I can gather, the problems *may* only exist on factory-overclocked and user-overclocked GPUs. The driver passed our normal QA checks before launch, and we are working tirelessly to understand what, if anything went wrong, and why some users are experiencing problems.

If a user believes that their GPU has been damaged by the driver, they should contact the seller of the GPU and/or the manufacturer."

Nvidia Facebook / Twitter

https://forums.geforce.com/default/...hread-released-5-23-13-/post/3821376/#3821376

Nvidia forums, silence.

 

This is ridiculous! They are only listening to complaints or issues which does not blame them!
Blaming 3rd party manufacturers, overclocks etc is just not acceptable.
I think they should own up and not keep saying, "We are not able to reproduce the same conditions you are experiencing". Well then, how do you explain 100s of other people who ARE experiencing!
Nvidia really messed up big time, not once but TWICE!

 

And the wiggling, dis-information and downright lies begin! Of course all we want to hear is that they really messed up and are putting all their resources into fixing it. The blame game is just stupid at this point.

Driver A works. Driver B1 is released for testing (beta) it has a bunch of problems. Driver B2 is released, same issues. It then goes to WHQL and the world as a stable new release - it still has the same issues as the two provious betas. Whatever the cause these drivers should have been pulled or at the least a warning regarding the specific cards already known to be failing. We know here a 680/3D AW gets messed up with this release!

But the marketing/public Relations (forum auto-correct puts 'Google Page Ranking department' in place of P R??) department takes control of these things. Damage limitation mode. Say whatever we can to not look like complete as*es or risk a lawsuit.

Such a shame as I often advised people that were not enthusiasts to go with NV as the drivers are solid and easy to use. The jury is out now.

Edit: Oh the irony, the NV newsletter just arrived pushing this driver

 

Nvidia QA checks + Windows Hardware Quality Lab (WHQL) testing = 320.18 WHQL then Microsoft releases Nvidia drivers v. 311.44 http://download.windowsupdate.com/m..._43383f9ea8c8725009c658052b9a98f33e6f2e19.cab (GTX 780M included) :thumbsup:.........

 

#589 https://forums.geforce.com/default/...hread-released-5-23-13-/post/3822575/#3822575 accuracy ? @MickyD1234 have you checked DPC latency ?
Nvidia killed post linking.....

 

Not sure which thread you are pointing at but I did research DPC latency this morning and it seems to be the cause of the lockups for sure. I've not tested mine since it's working (with acceleration off in FF) and I'm sure a new driver will be here very soon - hope!

Extract from wikipedia: When working with streaming audio or video that uses interrupts, DPCs are used to process the audio in each buffer as they stream in. If another DPC (from a poorly-written driver)[citation needed] takes too long and another interrupt generates a new buffer of data, before the first one can be processed, a drop-out results.

I think we have a citation now!!

 

There are many reports of high DPC latency with 320.18, post #589 damaged GTX 580M (Alienware)

DPC latency checker (standalone, Win 7) http://www.thesycon.de/dpclat/dpclat.exe

http://www.thesycon.de/deu/latency_check.shtml

Pro tool http://www.resplendence.com/latencymon (Installer Win8 / 7) http://www.resplendence.com/download/LatencyMon.exe free home edition

 

That's what prompted me to investigate.

Just responded to that one (#589)

 

Edited previous post, tried it out?

 

Just run it, not sure what it is telling me but I'm not seeing the spikes that others have posted. I should check it out with the acceleration back on but it locks within minutes

 

Those are very good results for 320.18 (max 264 µs), let it run for at least 60 seconds (anti-virus off).

That's bad coding( Nvidia), don't bother.

 

Turned off MSE. After about 60 seconds it started to climb slightly...

 

That's more like it, probably nvlddmkm.sys - NVIDIA Windows Kernel Mode Driver 459 µs spike, still acceptable for 320.18

LatencyMon gives detailed results to pinpoint drivers / services....

 

Seems you are right.

 

MSE still disabled (hard pagefaults) ? Look for Stats then Processes / Drivers.

 

Doh! You got me, you sure know your stuff

I turned it back on after the first test. I wondered about those hard page faults, didn't fit anything related to the NV driver. Here is a better one: