Nvidia 191.07 drivers causing outright hardware failure?

I guess I will have to test more... I run eVGA precision with my cards at 696/1450/1150 and my fans at 70% and they didn't seem to get any hotter than normal with these drivers (load in the 50s) but I only played l4d2 for like 30 mins last night.. I'll need to hammer on them with some full day gaming and see what happens..

edit: the other caveat seems to be, most if not all the accounts of this happening I read were with windows 7.. I am on vista64 still.
 
Last edited:
Hooligan said:
No, cuz it happens to me as well on an Intel mobo
Click to expand...
NF200 are not limited to NV-chipset.
X58 has it as well and that is Intel-mobo and this is only something related to NF200 and 191.07 driver.
 
This confirmed NF200?

I just wanna make sure. I've been running it fine on my system for a day or two.
 
Just installed the 191.07 drivers. I have a gtx 295 and a gigabyte g31 motherboard. Works fantastic so far in cod5, cod4, dirt, batman, and resident evil 5.
 
After installing 191 drivers mine got worse and worse... at first it would crash out of games... now windows wont even recognize it.... and it wont even load windows unless its from a cold boot

read:

http://hardforum.com/showthread.php?t=1470834

anyways...

i did a fresh install of WinXP 64 and went with the 182 drivers. same thing... wouldn't load windows unless it was in safe mode, or without ANY drivers installed or from a cold boot.

I just reinstalled Vista64 again loaded the 186 drivers and same shit is happening... now the drivers wont even recognize the video card

so yes. 191 drivers caused my videocard to fail outright. on an intel mobo.

gx2.jpg
 
Do the 195 beta drivers cause this?

I must of missed the 191.07 train, and I saw the 195 beta drivers last week or so.

I'm getting a 9800gx2, so I'm now scared as heck. Maybe it's just a fluke, but still...
 
Hooligan said:
After installing 191 drivers mine got worse and worse... at first it would crash out of games... now windows wont even recognize it.... and it wont even load windows unless its from a cold boot

read:



anyways...

i did a fresh install of WinXP 64 and went with the 182 drivers. same thing... wouldn't load windows unless it was in safe mode, or without ANY drivers installed or from a cold boot.

I just reinstalled Vista64 again loaded the 186 drivers and same shit is happening... now the drivers wont even recognize the video card

so yes. 191 drivers caused my videocard to fail outright. on an intel mobo.

ouch i really feel for you Manm. I cant believe Nvidia released these. trhey should be sued. I uninstalled them today myself.
Click to expand...
 
Last edited:
played quite a bit of L4D2 on these drivers this weekend with no issues.. *knocking on wood* I might just bounce up to the 195 betas though.. I am a little freaked out now. :p
 
My previous 8800gt borked itself possibly because of the driver, my replacement card from Gigabyte currently only works with the default 182 driver on win7 64bit rc. Will reinstall windoze soon to test further.
 
I find this hard to believe, since it seems there is no correlation between the video card and the chipset and the video card and the drivers. Also I've used basically every driver made and my card hasn't died if the driver itself was at fault then everyone who used that driver would have dead cards. So there must be some underlying issue that is unknown.
 
flak-spammer said:
I find this hard to believe, since it seems there is no correlation between the video card and the chipset and the video card and the drivers. Also I've used basically every driver made and my card hasn't died if the driver itself was at fault then everyone who used that driver would have dead cards. So there must be some underlying issue that is unknown.
Click to expand...

Typically I wouldn't think drivers would cause hardware to fail, but the timing was entirely too coincidental for me. My card had zero issues for 8 months and within the space of 3 days it went from perfect to being unable to run any 3D applications.

It seemed that anything that pushed the VRAM would cause it to fail and the screen corruption I received on system freeze was indicative of VRAM failure. This would happen after about 2 -15 seconds in game by the end of things, though desktop applications were fine.

Here is the screenshot at Nvidia forums (you have to login to view though) http://forums.nvidia.com/index.php?act=attach&type=post&id=13837

The thread related to my issue can be found here: http://forums.nvidia.com/index.php?showtopic=109904&st=0&p=605057&#entry605057

Whats really interesting is that all of the stickers on the cooler had peeled off and rolled up indicating excessive heat (causing the glue to melt). I'm thinking the drivers were not throttling the fan properly and as a result causing the components to overheat under load. Its likely this would not be so much an issue for stock to very mildly OC'd cards so perhaps this is why not everyone is having issues. As I have an OCX the cards clocks are fairly significant over Nvidia stock specs. More clock = more heat = more quick to fail when the above comes into play.
 
i ran them for a while and had no trouble, but i just upgraded for peace of mind.
GTX 260
 
My friend just told me his video card crapped out (new GTX285) using 191.07 drivers.. he was on them for about a month and suddenly during a game he got a BSOD and the code referenced an Nvidia display driver (his system is an all intel chipset mobo on vista64).. well, now his system is all messed up.. any 3D apps run like it's in slow motion, if it works at all.. according to him.. I haven't seen it myself.. but now I am kind of freaked out. going to bounce up to the 195 betas tonight. :p

edit: my logical mind still doesn't want to believe a driver could do this but the conspiracy theorist in me says anything is possible.
 
i really don't see how drivers could possibly cause such a catastrophe, but who knows

i mean from one version to another that has this and doesn't, i can't see how they would change anything in the code base that would even come close to touching something delicate / magically killing systems.
 
wow this is some big BS from nv

im never updating my drivers ever again for real


The only time I might think bout upgrade is when the games I play start acting weird or something
 
Had this same issue just happen to me this week. I've been running a 8800GT for almost 2 years now (came with my Dell XPS 420). I upgraded to Windows 7 64-bit, and have been playing lots of Dragon Age the past 2 weeks. A few days ago, I got my first BSOD. After getting it again while playing DA, I manually cranked the fan up on my card with RivaTuner. This allowed me to play again for a good 6-7 hours. The next day, the BSOD's were back, and have not gone away. Any game will cause a lockup, some (DA) faster than others. The error is always the same - a crash in the nvlddmkm.sys file, with the error "attempt to reset the display driver and recover from timeout failed." Often the lockup will cause graphical corruption, and sometimes it'll get stuck there rather than BSOD.

Outside of games, the system is working fine. I am convinced my 8800 GT is dead, although I didn't think to blame the drivers until I saw threads like this. So long NVidia, I have a Radeon 5850 coming to replace it.

Edited to add: My motherboard is pure Intel, no NVidia chipset.
 
My BFG 8800 GTX just took a dump tonight with the exact same problems as everyone else. 2 years no problems, installed windows 7 no problems for 3 weeks, then updated to 191.07 and got the code 43 error message after about a week of use. Nothing helps " rollbacks, reformat, Win XP, Bios resets." Going to RMA card, hopefully the motherboard is not dead. BTW motherboard is a Gigabyte w/ P35 chip-set.
 
Sorry for your misfortune kevin... but join the club.

This is some serious bullshit from nvidia if you ask me :(
 
I ran these drivers for a week or so about a month ago and got NOTHING but crashes in all my games. Black screen followed by a driver reset that required a reboot or all games would run at 40% speed until a reboot was done. I rolled back a few versions and haven't had any issues for the most part. I see the new WHQL drivers came out today but honestly I don't want to even go there since everything is working now and there are no games out there anyway that I am interested in. I agree that nvidia needs to get their shit together and soon. As it stands now, they are going to be looking at major lawsuits for destroying hardware with these drivers.
 
If the motherboard's PCI lane is fried, does anyone know if BFG or Nvidia can or should be held culpable for the repair or replacement? I know that my board is out of warranty. Oh well I was looking for a reason to upgrade but not so soon as my Q6600 still puts out more than I need. I hope BFG doesn't give me a hard time about the RMA.
 
Chromatik said:
my machine will be ok for anywhere from 0 to 30 seconds and then it looks like high res rainbow colored bubble wrap across the screen. Hard boot is the only thing that will clear it.
Click to expand...

I'm having the same problem on 191.07. I just downloaded 195.62 to see if it will fix it. It was happening in TF2 until I installed rivatuner and cranked the fan up to 100%. I recently purchased MW2, borderlands, and L4D though and they all crash within 5 minutes of playing. I hope my hardware isn't effed.

P5Q Pro, E8400, 4 Gigs ram, XFX GTX260 Black Edition

update: New drivers didn't help. If anything the games are crashing more quickly. Looks like it's time to start my RMA process.
 
Last edited:
Hmmm, upgraded to the 191 driver with no issues. Played some CoD4 for about 30 minutes, and everything putted along fine. Running a EVGA 260 card on Vista 64. Temps are the same. Nothing out of the ordinary.

EDIT - Just to error on the side of caution, I'm going uninstall and reinstall my 182's until this SNAFU is corrected.
 
Last edited:
.....this whole "driver issue" is getting out of hand. I went almost a year without updating my drivers and for whatever reason decided to download the new 19xxx drivers when they were released.

At the time the only game I played was Cod 4 mp and as soon as a installed the drivers I started getting hardlocks, driver not responding errors......I tried rolling back the drivers with no luck. No matter what version I rolled back to I continued to get the same errors.

I could play cod4 for 15 minutes at most then the game would freeze and I would get the driver not responding error. I pretty much gave up and just stopped playing the game.

This is where it gets weird. I purchased Borderlands with hopes that I could get this driver issue resolved and once again enjoy some pc gaming.......and guess what......using the 190.71's I never even had a hiccup, and this is after some 8-10 hour gaming sessions.

Well with the latest Steam sale I decided to jump all over Dead Space for 15 bucks and guess what.......the same ol' cycle has started. I can play Dead Space for 15 minutes max and then it freezes and I get the driver not responding error.

Though i'm quite burnt out on Borderlands I decided to fire it up and see if I had any problems.......I played for at least a couple of hours and the game ran just fine.

I'm really scratching my head over this one.......
 
I used these drivers on my gtx 285 on a 790i Ultra and had no problems over 3 days. I just rolled up to the beta drivers last night. :/ So far so good.

EDIT: I rolled back to 180's The new drivers are hellish on vista 32 bit. Over 8 of my friends + myself have all had framerate drops / problems with the new drivers.
 
Last edited:
Actually, if what I have been reading is correct, only people with seriously overclocked above nVidia spec cards are failing.

What this points out is that the previous drivers did not max the cards potential, the 191.07s do, thus more heat. With that added performance AND the extreme overclock of OCX/FTW/XXX/etc versions of the cards, they are failing.

This, honestly, is not nVidias problem. If they produce a driver that finally maxes the hardware, GREAT! It is just that only those that ran thier cards at nVidia 'spec' are the ones that aren't having hardware failures.

People should be pissed at themselves (if they are the ones that majorly overclocked thier stock cards), or at the manufacturer (eVGA/BFG/XFX/Asus/etc) if they bought a retail-overclocked card.

Honestly, it is sad it took nVidia this long to produce a set of drivers that are maxing out the hardware. But, we, the end users, are the ones to gain by them continueing to develop them to get more out of the hardware investment we have already made. To bitch and moan when your 'knowlingly' overclocked above nVidias own spec cards are frying is absurd.

Now, if the company you bought from has a good warranty, this shouldn't be an issue at all. As long as that company doesn't screw you over, you are set. Hopefully people are smart enough to know they can't run a new revision driver set on a massively overclocked card.

It is like people that max overclock thier systems and never run Prime95 or any other serious burn-in test to make sure the system is stable, and when it starts corrupting due to excessive CPU usage they start blaming everyone else. Fact of the matter is you have a part that is run out of manufacturer spec and those specs exist for the safe useage of that part. Run it beyond that and how dare you bitch and moan when it dies. Again, people don't take responsibility for thier actions and blame nVidia for finally producing a driver that does what it is supposed to.

For the record I have a XFX GTX280, plain vanilla, and a eVGA GTX280 FTW edition, which I run in SLI, but I underclock the FTW and slightly overclock the XFX to only around 644/1401/1200 and have yet to have any issues with any drivers. I was also running on a Foxconn Destroyer 780a motherboard until recently and have switched to a Asus Crosshair III Formula running the SLI hack that only works with the 191.07 drivers and have zero issues. I was running Vista Ultimate 64-bit on the Foxconn Destroyer before and am running Windows 7 64-bit on the Asus Crosshair III Formula now. Zero issues in games and no dead cards at a reasonable overclock.

Now, theone data-point I will add is that the Destroyer/GTX280 SLI setup did have the cards running hotter, but I believe that is due to a BFG ES-800 power supply more than anything. On the Crosshair III Formula setup I am running a Corsair TX950 power supply and have noticeably cooler temps. I will be replacing the ES-800 powersupply on the Destroyer system tomorrow with a Corsair TX650 unit and will be able to tell right away if the system runs cooler.

And to clarify, I am not sayaing that maybe nVidia doesn't have some change in the drivers that make cards run hotter, or that it migth not affect non-overclocked cards, but I just have not been able to see where anyone with a stock clocked card is complaining that drivers killed thier card.
 
dbphelps said:
Actually, if what I have been reading is correct, only people with seriously overclocked above nVidia spec cards are failing.

What this points out is that the previous drivers did not max the cards potential, the 191.07s do, thus more heat. With that added performance AND the extreme overclock of OCX/FTW/XXX/etc versions of the cards, they are failing.

This, honestly, is not nVidias problem. If they produce a driver that finally maxes the hardware, GREAT! It is just that only those that ran thier cards at nVidia 'spec' are the ones that aren't having hardware failures.

People should be pissed at themselves (if they are the ones that majorly overclocked thier stock cards), or at the manufacturer (eVGA/BFG/XFX/Asus/etc) if they bought a retail-overclocked card.

Honestly, it is sad it took nVidia this long to produce a set of drivers that are maxing out the hardware. But, we, the end users, are the ones to gain by them continueing to develop them to get more out of the hardware investment we have already made. To bitch and moan when your 'knowlingly' overclocked above nVidias own spec cards are frying is absurd.

Now, if the company you bought from has a good warranty, this shouldn't be an issue at all. As long as that company doesn't screw you over, you are set. Hopefully people are smart enough to know they can't run a new revision driver set on a massively overclocked card.

It is like people that max overclock thier systems and never run Prime95 or any other serious burn-in test to make sure the system is stable, and when it starts corrupting due to excessive CPU usage they start blaming everyone else. Fact of the matter is you have a part that is run out of manufacturer spec and those specs exist for the safe useage of that part. Run it beyond that and how dare you bitch and moan when it dies. Again, people don't take responsibility for thier actions and blame nVidia for finally producing a driver that does what it is supposed to.

For the record I have a XFX GTX280, plain vanilla, and a eVGA GTX280 FTW edition, which I run in SLI, but I underclock the FTW and slightly overclock the XFX to only around 644/1401/1200 and have yet to have any issues with any drivers. I was also running on a Foxconn Destroyer 780a motherboard until recently and have switched to a Asus Crosshair III Formula running the SLI hack that only works with the 191.07 drivers and have zero issues. I was running Vista Ultimate 64-bit on the Foxconn Destroyer before and am running Windows 7 64-bit on the Asus Crosshair III Formula now. Zero issues in games and no dead cards at a reasonable overclock.

Now, theone data-point I will add is that the Destroyer/GTX280 SLI setup did have the cards running hotter, but I believe that is due to a BFG ES-800 power supply more than anything. On the Crosshair III Formula setup I am running a Corsair TX950 power supply and have noticeably cooler temps. I will be replacing the ES-800 powersupply on the Destroyer system tomorrow with a Corsair TX650 unit and will be able to tell right away if the system runs cooler.

And to clarify, I am not sayaing that maybe nVidia doesn't have some change in the drivers that make cards run hotter, or that it migth not affect non-overclocked cards, but I just have not been able to see where anyone with a stock clocked card is complaining that drivers killed thier card.
Click to expand...



for reference, i have a bone stock bfg gtx260 that i never overclock, and these drivers caused numerous issues with me playing fallout3. i had to run the cards fan at 100% just to HOPE to get stability.

i moved to the newest drivers 195's i believe, and this crashing issue did stop. the game still has plenty of stutters however, and i am looking like others in this thread to roll back to an older driver in hopes of resolving this.

good luck everyone with your burned out cards. i hope my card is ok as well

BTW im running GTX260 OC on win7 64 asus P5B mobo
 
johnzm said:
BTW im running GTX260 OC on win7 64 asus P5B mobo
Click to expand...

So your card is factory overclocked correct? All soapboxing aside, I think dbphelps is on to something...

I'm just lucky that so far XFX has been very active in helping me troubleshoot and I'll let you all know how that concludes.
 
My 8800GT that died recently (just after upgrading to 191.07) was running completely stock. It was an NVidia reference board bundled with a Dell.
 
My evga 9800 GX2 vanilla was never overclocked since it was plenty fast for all my gaming needs stock. And it still got fucked up by the drivers.
Its getting RMA'd now as we speak.
 
Interesting, I have been running these drivers on this older (by comparison to most on here) system and all has been smooth as gravy for quite some time now.

P5N32-SLI Premium (590i chipset)
C2E X6800
BFG 8800GTX vanilla (bought 1 month after release of these cards) reflashed to an "OC2" card.
Win XP x64
 
I will add as a datapoint that my sons machine has been running the 191.07s since thier release with no issues... His config:

BFG 590a SLI motherboard
Athlon 5400+ overclocked to 3.2GHz
Dual eVGA 8800GT video cards overclocked to SSC specs
(one is a factory SSC, the other a regular vanilla one with an Akimbo cooler)
WinXP 32-bit

Now, maybe the 32-bit is the reason he doesn't have a problem? His system has been smooth and fast...
 
One thing I've seen consistently in reports of dying cards is Vista/W7-64 bit. Maybe that version of the driver is the only bad one?
 
Vesper said:
One thing I've seen consistently in reports of dying cards is Vista/W7-64 bit. Maybe that version of the driver is the only bad one?
Click to expand...
I was also thinking along those lines as I read through this thread, knowing that on XP x64 at least, I had encountered no such problems; and most everyone who posted issues was on 7.
 
Hmm... I've been using 191.07 since October 27th. However, I updated through WU (first time I ever did).

Edit: Over the past 2-3 weeks my system has frozen on me a few times... hopefully this isn't the cause.
 
Last edited:
I use MSI Afterburner for fan control and keeping an eye on temps, and one weird thing I've noticed is the card operates at higher temps sometimes for no apparent reason. Normally I'm around 42 C for 2D, but there's times that the temps will climb to 50-55 C +. Under gaming, that same thing will happen, with the card at times operating at a much higher temp than normal, causing some artifacting or system freezing.

When this happens, all I need to do is switch user and log back into the same account and the temps will start dropping immediately back to normal. Logging out and restarting the computer both work also. I don't know what the cause is, but it's definitely different behavior than I've ever dealt with before.
 
You must log in or register to reply here.
Back
Top