Nvidia 335.23 driver

Well guys here we go again. I'm gonna wait out Experience prompting me - I want to test the 'silent' upgrade.

Nvidia has released GeForce 335.23 WHQL drivers.


Notebook drivers

NVIDIA DRIVERS GeForce 335.23 Driver WHQL


NVIDIA DRIVERS GeForce 335.23 Driver WHQL

Good luck and please post results

 

I guess this is for the Titanfall release tomorrow?

 

damn it and i want dell to confirm my order like now already! LOL

 

That is what it says. What bugs me is that all the published stuff with this driver is profile updates. These can be updated as separate components silently, and do not need a complete new driver.

NV: FESS UP ON YOUR DRIVER DOCUMENTATION. TELL US WHAT BUGS IT IS SUPPOSED TO FIX!

 

780M


1. 3Dvision installer bug:

Code:

Problem signature:
  Problem Event Name:	APPCRASH
  Application Name:	nvStInst.exe
  Application Version:	7.17.13.3489
  Application Timestamp:	52f6580b
  Fault Module Name:	ntdll.dll
  Fault Module Version:	6.1.7601.18247
  Fault Module Timestamp:	521ea8e7
  Exception Code:	c0000005
  Exception Offset:	0003332f
  OS Version:	6.1.7601.2.1.0.256.1
  Locale ID:	1033
  Additional Information 1:	0a9e
  Additional Information 2:	0a9e372d3b4ad19135b953a78882e789
  Additional Information 3:	0a9e
  Additional Information 4:	0a9e372d3b4ad19135b953a78882e789

2. 3dVision Bug:

3Dvision unchecked:

View image: bug0


3Dvision checked:

View image: bug



Same bugs since the release of the 700 series drivers.



Both do not happen on 670M.

 

I agree, but it looks like some other stuff has been fixed with this driver according to this link, but don't know where they got their info from:

GeForce WHQL Driver 335.23 - News - LaptopVideo2Go Forums

(2nd post in that link) (maybe they got the info from Geforce forum - Manuelg - but just supposition on my part).

 

Thanks man. I just found that stuff buried in a link on the full NV document page so that's NV's official fix list - nowhere near the list of things that the last WHQL introduced and reported

 

Rushed release...see open issues in version 335.23 page 19 http://us.download.nvidia.com/Windows/335.23/335.23-win8-win7-notebook-release-notes.pdf and compare with fixed issues...:hi2:

Changes and Fixed Issues.
• [NVIDIA Inspector]: V-SYNC does not work if FPS Limiter is enabled. [1467316]
• [3D Vision][G-SYNC capable displays]: With stereoscopic 3D enabled, screen issues occur when running some DirectX 11 applications in windowed mode on G-SYNC capable displays. [1451395]
• [3D Vision]: The NVIDIA Control Panel->Set Up Stereoscopic 3d->View Compatibility with Games page does not show updated game ratings. [1452179]
• [Surround]: In Surround mode, the process of moving windows or minimizing/maximizing windows is sluggish. [1449016]
• [Surround][GeForce GTX TITAN]: Windows maximize across all three monitors in 2D Surround mode instead of one. [1421034]
• [SLI][Diablo III Patch 2.0.1]: The game frame rate drops with SLI and the Reflections game setting enabled. [1469204]
• [SLI][GeForce GTX 690][DOTA 2]: The game chat area flickers with SLI enabled. [1399587/1456817]
• [SLI][Surround][GeForce GTX 660 Ti]: The Surround accessory monitor desktop color and player color settings get adjusted instead of monitors in the Surround configuration. [1396022


I'm keeping Quadro 334.95 QNF 100% stable & best image quality, GF game optimization + unified drivers making wonders...

 

Hmm, seems like they are still struggling with the newer cards. Already a few negative comments in the last driver thread. NV haven't even put up the new one yet!
That looks like vanishing physX objects but I wouldn't think they would have any need for physX properties on them, strange

Guess you're on a break manuelG?

 

Too funny . None of that stuff stands up to examination of the current issues but hey-ho at least they are trying. Keeps me occupied anyway haha.

 

j95, is it a big difference in image quality, where do you notice it most? Do you find performance to be the same, or lower? (I might give the Quadros a go if the performance is not lower).

 

Tried this driver, no problems so far, identical results to previous whql in all benchmarks and game benchmarks: 3DMark11, 3DMark 2013, Metro Last Light, Bioshock Infinite, Batman Arkham Origins, Grid 2, and Tomb Raider. Works fine on Skyrim too.

 

Thanks, good to hear no probs. I didn't expect any game increases since they have enough work with just trying to get it to work for a lot of people. At least it hasn't got slower which sometimes happens when it's a fix release.

Still no GFE prompt yet for me, probably coincide with a new thread @ NV?

Edit: https://forums.geforce.com/default/...-335-23-whql-driver-feedback-thread-3-10-14-/

Wow, that got lively pretty quick!

 

Just a quick aside to the main title of this thread, I just tried the Quaddro 334.95 driver that was mentioned earlier in this thread. I didn't notice much difference in IQ, it's possible that the edges were sharper in general, which helped with clarity, but might have made jaggies a bit more noticeable. Interestingly I got the best performance I've ever seen in a couple of PhysX titles: Metro Last Light, Batman Arkham Origins, got a 2% and 4% increase in fps on both of those in comparison to whql 335.23 (otherwise performance identical). The Quaddro driver was a bit of a pain to install as I had to modify a different inf file (nvdmwi.inf), and during installation there were more component option choices to install (like nview, and NVidia WMI, both of which I chose not to install), it also didn't give the option to install PhysX, but I already had that installed, and PhysX works fine. Would be interesting to hear anyone's experiences with Quaddro drivers & if and what difference you think they make.

EDIT: 2% increase on Batman Arkham City with Quaddro driver too. Looks like it works well with PhysX titles.

 

Thanks man. +rep (when I can!).

Sounds good, I'm still waiting on GFE but I expect it to happen today or tomorrow.

Interesting what you say about the quality. It's trick to identify just what IS better but when I moved to the last WHQL, the last test I did on the previous driver was Dirt Showdown and the first on the new driver. I could immediately see it looked 'better'. The only part I could say was obvious was the boot of your car in the benchmark. It was always kinda washed out but not distracting. Now it's crystal clear .

As soon as I update and get some testing in I'll post...

 

Are you gonna try the Quaddro or the Geforce (or both!)? (My last post was about the Quaddro). Yeah, the IQ is so subjective, hard to tell what's placebo and what's not! (The power of placebo has been proven to be stronger than some drugs!)

 

I'm waiting out GFExperience prompting me. I'm then going to try quiet upgrade. Risky on past experience (sic) but I keep hammering away, pushing NV to get this stuff working. There's a lot of new PC gamers and all this juggling about it ends up with them here!

Doing my bit for the community

 

Fair dos! I used to have GFExperience installed, but I always found the new drivers before GFE did! (But that's not surprising considering I go on tech sites everyday, and it's always listed there where a new driver comes out!). Yes, I can see that you're testing it like an average user, will be interesting to see how long it takes for GFE to update!

 

That's the plan. It's not looking good right now. I just caught up on the NV thread and it seems people are still having GFE hang-up at 'Preparing to install...').

This happened to me last time. Luckily I have seen this behaviour before so I just left it until all disk access dropped off. I then killed GFE in task manager and restarted GFE, - without a reboot. It said I was running the new version so I benchmarked it. Not a single issue

I'm pretty sure those people having a black screen that never returns are having the same thing, only they don't see any message about installing .

 

Well, GFE just popped up telling me there is a new driver. I had another scan of the NV forum and another 3 pages of problems - many with the installer hang. Too much knowledge is a bad thing. Since it happened to me last time and now I've seen three people with messed up windows not starting I can't be doing with that right now - I need to evaluate further.

They probably killed the installer with a power down before it had finished so I'm hopeful it will be OK. Or I could run the download outside if GFE (cop out ).

Let's see what the next hour or two brings - I've already put my 2c in on the forum...

 

Driver for Titanfall....?? Nice..... Just Installing the game & play it for finishing 1 campaign.... gameplay is good, but big problem in balancing team during multiplayer issue and surprisingly is not eaten your laptop spec too much, considering it took 50GB for a hard disk space..... lol..... just turn off the AA and you got solid 60 FPS, for most system I think.....

I saw j95 already post his magic .inf file, so I will give it a try..... hope is better than the last driver update which give me constantly crash back then so I need to refers back into 332.21....

 

OK, so I bit the bullet and went for easy upgrade option in GFE. As I expected it just stuck at 'Preparing for installation'. I could see it was still working and the only strange bit was when it gets to the 3D components and a file copy window pops up. This time it took ages and I saw every name of a dozen or so files. Once I was sure it had finished I did the same as last time. Killed it in TaskManager and restarted it.

All looked good so I did a few quick benchmarks. Basically exactly the same last the one I tried:

Batman AO went down by 2fps. F12011 went up 2fps and Dirt Showdown went up by 2fps. All normal day-to-day variations.

And I appear to have reached the FPS that Dirt benchmark works best at. I win every time now!

Upshot for anyone reading, use the download and run it outside of GFE. Select Clean Install for best chances of a good install.

One day NV, one day....

 

Same problem for me. Tried to update driver and would just hang at preparing to install. After third try it made it past preparing to install and got stuck at installing. Powered down computer and rebooted and then loaded Gfe again and it was updated with new driver .
What's up with that .
It did it to me on another driver update before this one .
A big company like nvidia and they can't even make their update process work correctly.
I'm just glad it didn't screw up my computer trying to update a driver.

 

I have it working on my Alienware M17x R1 . I have attached the 2 files I modified to get it working . The first file is the list devices that is in the main folder . Then I have the NVDM that you replace . I have only configured it to run the GTX260M Sli . I could get the 9400 working later .

View attachment Nvidia 260m.zip

 

Yeah version 334.95 100% stable, no power throttle NVIDIA GeForce GTX 780M video card benchmark result - Intel Core i7-3940XM,Alienware M17xR4

Previous versions M17xR4/780m (M18xR2 no improvements) Modded Quadro nvdmq.inf v332.21 AW M17XR2/R3/R4 3D + M18XR1/R2 GTX 680M / 780M
NVIDIA GeForce GTX 780M video card benchmark result - Intel Core i7-3940XM,Alienware M17xR4

@Robbo nvdmwi.inf gtx 670mx post edited install instructions/section.


Quadro Notebook Driver 334.95 [Alienware] Aftermarket Upgrades Driver Support - Modded INF

 

Yeah, I continue to loose faith in NV to the point of being disgusted with their driver support for the last year or so.

This problem started with the last WHQL and was widely reported, not a mention from NV. Then they do it again with this release and their mod (Hi manuelG :hi2:, or should I call you ..., nah not going to 'out' you - yet ) just addresses single game issues ignoring people that have had serious problems caused by this installer.

There have been at least two NV posters that found their windows would not start and had to resort to repairing it. I think that they got fed up with watching that 'preparing...' message and powered down. The driver install was actually still in progress so their machine gets messed up .

Just to repeat: If you see this behaviour just wait it out. Mine took about five mins for all the disk access to stop. I waited a further five mins then killed it.

 

You should be getting a much higher score than that. The 780M scores between 7400 and 7900, out of the box.

 

FIRE STRIKE... (1.1V) 3DMark - test your graphics card with the gamer's benchmark

 

Oh, lol. I didn't notice that.

I only use 3DMark 11 for a baseline, so...

 

Windows 8.1 dearest...

 

i have the alienware m17x with i7-4700mq and gtx780. What did you have to do to your machine to get the clock up that high? Stock clocks for me and I only get 4900 on fire strike.

 

flash the unlocked vbios i would think.

 

Anyone experiencing random freeze with these? - I updated via GFE when prompted, since then I have had a few lock-ups when just browsing the WWW which require a hard reboot. Nothing is responsive when it happens, just completely locks up and freezes.......

 

I don't advise installing GFE. It's an incomplete program that probably is never going to be completed like the rest of nvidia's products. The freeze use to happen on my 670M. Nvidia is discontinuing support for lower cards because they don't know how to fix the issue for those cards. Your card may be one of them but you can always check their list on their site.

 

Hi Stevie, I'm seeing a lot of this over in the NV driver forum. Mostly on 7xx desktop cards.

NV fixed the majority of the interrupt driven problems (TDR) but the hiccups remain. They are now stating the issue with the 5xx/6xx cards was resolved two driver releases ago - actually they fixed the overvolt that was burning a lot of cards on earlier drivers but they will never admit to that!

Anyhoo it seems that it's simply a case of using one of the earlier drivers until you find one that works for you - great huh?



Fun list, the dates between directly correlate to the number of messes it caused . Only one rock solid one in there; 314.22, even damaged cards will sometimes work on that one...

 

I think what they did was added a throttle point that did not exist before because the last two drivers introduced a throttling condition at extreme overclocked levels (above 1135 on core) with enough voltage to support it. That issue did not exist before and I cannot get rid of it. It typically happens around 130% of TDP. Going to older drivers, flashing stock vBIOS and modded vBIOS does not eliminate it. It causes this on all four of my GTX 780M cards and Johnksss has run into the same thing with his GPUs. They seems to be monkeying around with stuff in drivers that makes permanent changes to the system (maybe vBIOS) and I'm not appreciating it. Might be with good intentions to keep people from pushing them too hard, but that's none of their business IMHO.

 

That is bad for a new card model!
I think they are struggling with two root-cause issues that are incompatible (so far).

1. Getting the best driver out there for the new 7xx cards.
320.18 was a panic WHQL release as the cards were on sale and NV didn't have a driver published. Everyone had to use the OEM supplied modified driver that came with their card.

2. maintain compatibility with the older 5/6 series (fermi) cards.
The boost feature was first implemented at the 3xx release. The only issue to begin with was a brief overvolt on fermi cards, which it should have no effect on. On mine it jumped to 0.92v, a voltage overclockers had already tried and it burned a lot of 580m's! I didn't worry about it initially and just reported it to NV (ignored as usual). Once 320 came out fermi cards were dieing all over the place. Since the most affected desktops were factory overclocked fermi cards the OEM's were in the spotlight. I can't say for sure since I don't have one but if the OEM has already made a slight voltage increase then using a percentage in the calculation will result in a much higher voltage, and subsequent failure.

Add in the fact MS changed the HAL (again, vista caused all these TDR's as well ). MS recently published a new set of calls specific to video card manufactures. TDR changes in Windows 8 (Windows Drivers) This allows selective reset of the card without producing an error or log-file entry. It seems NV that used some of this as I have seen quite a few reports that games are still pausing like a TDR reset has occurred but no sign of it.

Now NV claim all the fermi issues are resolved I can see the mods actually helping people, although they focus on single game issues at least they are there for a change.

 

That's interesting to know Mr Fox. I haven't noticed such a problem with mine, supposedly the TDP on mine is 75W, but it uses close to 100W at my overclock at 1.05V, which is about 130% TDP. I wonder if there's a condition on the throttling associated with the voltage used and/or temperature (both my voltages & temperatures are low), or perhaps an absolute maximum Wattage for the MXM slot that has appeared in the last 2 drivers.

 

it's saying my gpu is not supported

EDIT: Reseated the battery and now it's detecting in the BIOS....proceeding with install.

 

I was having some problems with these drivers....no problems before hand but afterwards I had some serious FPS lag in Loadout which went away after rolling back the drivers. Didn't spend any more time then that with them.

 

Yeah, seeing a lot of this with the 7xx desktop cards. Seems to me they have added in the new 8xx stuff and broke it?

One to avoid for 7xx users, no gain, risk pain

 

On the topic of NVidia Drivers this sounds quite exciting!:
NVIDIAs TBA DX11 Driver will be better than Mantle

Some big performance improvements on the way, and a rival to Mantle in terms of efficiencies, but all while doing it on DX11! Maybe with those kind of efficiencies MickyD, that might push your 675M over the edge into oblivion - but at least you'd get to upgrade sooner if it did kill your card with it's blinding efficiency and perhaps corresponding increase in heat production!

 

Haha, I'm sure it will tempt me.

I'm still finding this one just fine for high to med gaming and am hoping I can hold out until something very new has been checked out.

What I want to see is Tessellation applied to physics objects along with AO. It's always great to watch the movement of a bunch of physics objects but they often have 'hard' edges and never quite interact right with solid surfaces, as if they are touching them. Physics smoke/fog is perfect now IMO.

I think that would be where a super powerful card will be a must.

 

Sounds like you're after that 20nm Maxwell high end GPU! I wonder when this new ultra efficient driver will be released? Anyone heard anything?

 

Yeah, I've had the cash ready since the AW17 came out but has not tempted me at all - yet!

'Ultra efficient' eh? Nothing mentioned in the NV forum, they are till smarting at this release causing tons of rollbacks!

Edit: I missed the link @ NV - reading up now....

Nvidia’s Next Driver Will Bring Massive Performance Improvements In CPU Heavy Games | DSOGaming | The Dark Side Of Gaming

 

Cheers for that new link, different to the one I posted. They say in that link that they hope the new driver will improve titles across the board rather than just a couple of games. I think they will improve across the board, due to the contents of the graph entitled 'Continuous API Efficiency Improvements'. Looks like that's some generic improvement of efficiency that would help all games, especially ones that are more CPU bound. Fingers crossed, hope that new driver releases soon!

 

i am hopeful for this too

 

Been thinking about this (sad eh?) and it could be a game-changer (sic ). If they are rebalancing the driver then they will have been working on this for a while. Might explain the lack of care with recent releases? Could go either way?

Gotta hope the fermi core is part of the deal

Seems NV are not setting expectations too high: https://forums.geforce.com/default/...eedback-thread-3-10-14-/post/4165468/#4165468

 

Ah, good find MickyD. Looks like ManuelG is indeed being very conservative or very truthful when he says that 'some games' will see a performance boost with the new driver. I tend to think that the optimisations will apply to Fermi too, because it's related to CPU optimisations, and I don't see why that couldn't extend to Fermi too. We'll see though! ManuelG has put a bit of a damper on my enthusiasm for that new driver now.

 

i don't think this driver will fix the crappy performance of COD ghosts which was an utter pile of crap to being with. it'll be a miracle if it does