EVGA

Interesting 980 FTW/SC stability find

Author
DnBrn
New Member
  • Total Posts : 50
  • Reward points : 0
  • Joined: 2011/10/11 21:04:01
  • Status: offline
  • Ribbons : 1
2015/06/13 16:08:48 (permalink)
So we've seen many threads on the issues involved with the 980 FTW card and in some cases the Superclocked versions or even the 970 as well.  All having the similar issue of TDR crashes, aka black/white screen, etc.  Today someone posted yet another thread to the Elite: Dangerous forum regarding crashes he was having with his 980 SC.  What we have ended up finding is that, along with the usual under clocking that most of us have had to do, if you enable 4k DSR in the game this also alleviates the crashing.

The theory with this game in particular is the following written by their support:

We are aware of such problems but have not been able to pinpoint the cause exactly. It seems to occur mostly with cards that are factory over clocked.

Please note, the following is by no means a statement of absolute fact, it is our current understanding of the situation.

The issue seems to be that these cards are optimised for constant high load. Elite however tends to cause phases of maximum load interspersed with lower usage. Especially events like exiting from hyperspace near a star for example cause the load to rise from very low to very high. This seems to cause the cards in question to fail to adjust power levels quickly enough, their processor in turn briefly suffers from low voltage and behaves undefined, causing a crash.

Down-clocking the card in question lowers its power draw and alleviates the problem. It would be conceivable that small variations in production quality make a percentage of the affected card models more resistant to the issue, that would explain why having a replacement solved the problem for some people while it did not for others.

I am afraid as of now lowering the clock rate is the only solution at this point, we will continue to look into fixing the problem of course. I am sorry that I cannot be of greater assistance in this matter.

Fly Safe and Kind Regards,
CMDR Vega, Support Wing


Another thread on the forum was started and new information has been found:

I suspect that it it GPU boost that caused frequency and voltage fluctuations. It seems to me that turning on DSR (3840x2160) has solved the issue for me by increasing the GPU load, which lead to stable GPU frequency and voltage. No crashes today for 2 hours at much higher +70MHz comparing to yesterday without DSR.

Another option is to down-clock the GPU.


I can confirm that so far, this has made the card stable while playing this game and has definitely increased the usage of the card.  What I have noticed by downclocking the card before is that, for whatever reason, boost was disabled and it actually only down clocked the card by -1 even though I told it -51.  This of course is odd, but this is what Precision X is saying.  I do not know if it is a display bug or what it is.  By enabling DSR, while in the game, the performance has been basically locked to it's full stock potential at 1455 mhz due to it now requiring a full time high performance due to the increased load.

So what we see here is an example of forcing stability by "locking" it to a performance level, either below stock or at max potential.  I say locking because in either case, there is no fluctuations involved, it is either at 1277 or 1455 full time, which also means no voltage fluctuations as well. This may be why the card has never failed me in benchmarks.

Obviously, this would require much further testing that I'm sure no one wants to do as everyone is probably fed up and just in the 980Ti queue at this point.  This also is not to say that it's suddenly a good batch of cards either.  Even with this find, I remain in the 980Ti queue.  I found it interesting though and wanted to share with everyone.
#1

24 Replies Related Threads

    Samsander
    Superclocked Member
    • Total Posts : 113
    • Reward points : 0
    • Joined: 2015/05/21 06:09:28
    • Status: offline
    • Ribbons : 1
    Re: Interesting 980 FTW/SC stability find 2015/06/14 00:39:18 (permalink)
    DnBrn
    So we've seen many threads on the issues involved with the 980 FTW card and in some cases the Superclocked versions or even the 970 as well.  All having the similar issue of TDR crashes, aka black/white screen, etc.  Today someone posted yet another thread to the Elite: Dangerous forum regarding crashes he was having with his 980 SC.  What we have ended up finding is that, along with the usual under clocking that most of us have had to do, if you enable 4k DSR in the game this also alleviates the crashing.

    The theory with this game in particular is the following written by their support:

    We are aware of such problems but have not been able to pinpoint the cause exactly. It seems to occur mostly with cards that are factory over clocked.

    Please note, the following is by no means a statement of absolute fact, it is our current understanding of the situation.

    The issue seems to be that these cards are optimised for constant high load. Elite however tends to cause phases of maximum load interspersed with lower usage. Especially events like exiting from hyperspace near a star for example cause the load to rise from very low to very high. This seems to cause the cards in question to fail to adjust power levels quickly enough, their processor in turn briefly suffers from low voltage and behaves undefined, causing a crash.

    Down-clocking the card in question lowers its power draw and alleviates the problem. It would be conceivable that small variations in production quality make a percentage of the affected card models more resistant to the issue, that would explain why having a replacement solved the problem for some people while it did not for others.

    I am afraid as of now lowering the clock rate is the only solution at this point, we will continue to look into fixing the problem of course. I am sorry that I cannot be of greater assistance in this matter.

    Fly Safe and Kind Regards,
    CMDR Vega, Support Wing


    Another thread on the forum was started and new information has been found:

    I suspect that it it GPU boost that caused frequency and voltage fluctuations. It seems to me that turning on DSR (3840x2160) has solved the issue for me by increasing the GPU load, which lead to stable GPU frequency and voltage. No crashes today for 2 hours at much higher +70MHz comparing to yesterday without DSR.

    Another option is to down-clock the GPU.


    I can confirm that so far, this has made the card stable while playing this game and has definitely increased the usage of the card.  What I have noticed by downclocking the card before is that, for whatever reason, boost was disabled and it actually only down clocked the card by -1 even though I told it -51.  This of course is odd, but this is what Precision X is saying.  I do not know if it is a display bug or what it is.  By enabling DSR, while in the game, the performance has been basically locked to it's full stock potential at 1455 mhz due to it now requiring a full time high performance due to the increased load.

    So what we see here is an example of forcing stability by "locking" it to a performance level, either below stock or at max potential.  I say locking because in either case, there is no fluctuations involved, it is either at 1277 or 1455 full time, which also means no voltage fluctuations as well. This may be why the card has never failed me in benchmarks.

    Obviously, this would require much further testing that I'm sure no one wants to do as everyone is probably fed up and just in the 980Ti queue at this point.  This also is not to say that it's suddenly a good batch of cards either.  Even with this find, I remain in the 980Ti queue.  I found it interesting though and wanted to share with everyone.




    This is exactly what I said here:  http://forums.evga.com/FindPost/2350365
    #2
    Samsander
    Superclocked Member
    • Total Posts : 113
    • Reward points : 0
    • Joined: 2015/05/21 06:09:28
    • Status: offline
    • Ribbons : 1
    Re: Interesting 980 FTW/SC stability find 2015/06/14 11:08:27 (permalink)
    I also tested this theory on 3DMark Skydiver. That always crashes in minutes.

    Run custom Demo mode at 1080P, triliniar filtering, tessellation set to 5. It crashes in minutes with black screen. BSOD 116 is logged after restart.

    Now Run custom Demo mode at 4K DSR resolution, anisotropic filtering set to x16, tessellation leave at 5 and it will loop for hours without crashing.

    So power fluctuations is a real problem with GTX980 FTW.
    Maybe this can be fixed with a BIOS update, maybe not. I guess we can only wait and see what EVGA does about this.
    #3
    Red46
    New Member
    • Total Posts : 61
    • Reward points : 0
    • Joined: 2014/02/04 10:23:55
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/14 13:36:54 (permalink)
    Samsander
    I also tested this theory on 3DMark Skydiver. That always crashes in minutes.

    Run custom Demo mode at 1080P, triliniar filtering, tessellation set to 5. It crashes in minutes with black screen. BSOD 116 is logged after restart.

    Now Run custom Demo mode at 4K DSR resolution, anisotropic filtering set to x16, tessellation leave at 5 and it will loop for hours without crashing.

    So power fluctuations is a real problem with GTX980 FTW.
    Maybe this can be fixed with a BIOS update, maybe not. I guess we can only wait and see what EVGA does about this.


    This is what happens to me too, 1080p crashes within minutes, 4k DSR AF 16x passes without a hitch
    #4
    Samsander
    Superclocked Member
    • Total Posts : 113
    • Reward points : 0
    • Joined: 2015/05/21 06:09:28
    • Status: offline
    • Ribbons : 1
    Re: Interesting 980 FTW/SC stability find 2015/06/14 13:45:13 (permalink)
    Red46
    Samsander
    I also tested this theory on 3DMark Skydiver. That always crashes in minutes.

    Run custom Demo mode at 1080P, triliniar filtering, tessellation set to 5. It crashes in minutes with black screen. BSOD 116 is logged after restart.

    Now Run custom Demo mode at 4K DSR resolution, anisotropic filtering set to x16, tessellation leave at 5 and it will loop for hours without crashing.

    So power fluctuations is a real problem with GTX980 FTW.
    Maybe this can be fixed with a BIOS update, maybe not. I guess we can only wait and see what EVGA does about this.


    This is what happens to me too, 1080p crashes within minutes, 4k DSR AF 16x passes without a hitch




    This proves that all those black screen crashes we have with the FTW model are not due to stock overclock being to high or due to voltage being locked at 1.2v and might be to low for the stock clock of the FTW but due to instant high to low and low to high fluctuations in the power draw. The component on the card just cannot adjust fast enough and that's why everything crashes and shuts itself down to protect the system.
     
    I can even overclock the card to 1450Mhz with the fan set to auto and it will be stable as long as I keep the power draw more or less at 100%.
    But as soon as you run a game or a benchmark not at crazy high settings and the power draw is jumping around 100% to 60% to 90% to 30% then back to 100% and so on. The card cant handle it.
    post edited by Samsander - 2015/06/14 13:53:40
    #5
    sectek
    New Member
    • Total Posts : 15
    • Reward points : 0
    • Joined: 2015/05/16 07:58:28
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/15 08:28:40 (permalink)
    This is good information.  Thank you for sharing and hopefully EVGA will use this information productively. 
    #6
    DnBrn
    New Member
    • Total Posts : 50
    • Reward points : 0
    • Joined: 2011/10/11 21:04:01
    • Status: offline
    • Ribbons : 1
    Re: Interesting 980 FTW/SC stability find 2015/06/15 12:24:49 (permalink)
    I'm glad this has helped a couple people and I hope it helps EVGA pin down whatever the issue is. 

    In my experience and research, it does seem to be a combination of software and hardware.  Some software causes the card to change frequency and voltage often.  The more often and sharper the fluctuations the more unstable the card seems to become.  Some software seems to keep the card at a certain level on its own. I have noticed this in testing as in the more stable games, the card will stay at a frequency and voltage regardless of load.  I think this is why both under clocking and increasing the load to as high as the card can handle creates stability.  Under clocking seems to disable boost and as long as you have it set to max performance for your power setting, there's no fluctuations there, same with something like 4k DSR.  There's no room to fluctuate because the card needs to run at full power in order to keep up with the load.
    #7
    Methodical2
    SSC Member
    • Total Posts : 562
    • Reward points : 0
    • Joined: 2012/03/23 18:37:46
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/15 12:50:52 (permalink)
    Question.  Can the same be said if you use Precision X boost mode?  Curious.

    "Never be afraid to try something new. Remember, amateurs built the Ark, professionals built the Titanic"
     
    BlackE
    Case:  Caselab Magnum STH10
    MB:  Asus X670E Hero
    CPU:  Ryzen 9 7950x (EK-Quantum Vector²)
    GPU:  Asus 4090 TUF OC (EK-Quantum Vector²)
    Memory:  G.Skill Trident Z5 Neo DDR5 (6000 @64gb)
    Storage:  WD M.2 NVMe 2TB (OS), 2x4TB (Photography), 1TB (Games)
    Fan Controllers: Aquacomputer Aquaero 6 XT & Octo
    PSU:  EVGA 1200 P2
    Monitor:  LG 48" UltraGear OLED 4k

     
    #8
    DnBrn
    New Member
    • Total Posts : 50
    • Reward points : 0
    • Joined: 2011/10/11 21:04:01
    • Status: offline
    • Ribbons : 1
    Re: Interesting 980 FTW/SC stability find 2015/06/15 13:27:04 (permalink)
    Methodical2
    Question.  Can the same be said if you use Precision X boost mode?  Curious.



    I haven't tried it, but theoretically, it should since K-boost locks the card to max performance.
    #9
    the_Scarlet_one
    formerly Scarlet-tech
    • Total Posts : 24581
    • Reward points : 0
    • Joined: 2013/11/13 02:48:57
    • Location: East Coast
    • Status: offline
    • Ribbons : 79
    Re: Interesting 980 FTW/SC stability find 2015/06/15 14:18:29 (permalink)
    Just remember to turn kboost off every single time you stop playing. Kboost keeps your card at max clock even when games are turned off, and can cause very high idle temps and frustration with the fans running constant.

    The dsr setting should be able to be set per game, so when you exit your gpu down clocks properly. This would obviously extend the life of the card and cause less noise and heat.
    #10
    Red46
    New Member
    • Total Posts : 61
    • Reward points : 0
    • Joined: 2014/02/04 10:23:55
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/15 14:34:51 (permalink)
    Those workarounds are only helpful to provide the evga techs with as much info as possible. You shouldn't be required to use third party programs or, play around with extremely taxating graphic options that in graphically demanding games are not really feasible to use to begin with, just to get the card to run as it should be doing out of the box (EG not crashing and bringing the whole system down with it)
    #11
    Samsander
    Superclocked Member
    • Total Posts : 113
    • Reward points : 0
    • Joined: 2015/05/21 06:09:28
    • Status: offline
    • Ribbons : 1
    Re: Interesting 980 FTW/SC stability find 2015/06/15 14:49:37 (permalink)
    DnBrn
    Methodical2
    Question.  Can the same be said if you use Precision X boost mode?  Curious.



    I haven't tried it, but theoretically, it should since K-boost locks the card to max performance.




    I don't think you understood your findings. This is not going to work.
    The problem is not core clock fluctuating. It's the Power Draw (wattage) fluctuation what is making the card unstable. K-boost does not lock power draw. It is impossible to draw power without putting the card under actual load.
     
    Core clock fluctuation is not an issue at all. You can test this by overclocking your core to about 1420Mhz-1440Mhz and then lowering the Power Limit in Afterburner or PrecisionX for example.
    The clock will start jumping all over from as low at 1100Mhz and up to 1440Mhz boost clock, depends on how low you set the Power Limit.
    But as long as you run the games and benchmarks at ultra high settings and at 4K DSR resolution and keep the Power Draw (wattage) close to 100% without dropping to much the card will not crash.
    post edited by Samsander - 2015/06/15 14:56:06
    #12
    DnBrn
    New Member
    • Total Posts : 50
    • Reward points : 0
    • Joined: 2011/10/11 21:04:01
    • Status: offline
    • Ribbons : 1
    Re: Interesting 980 FTW/SC stability find 2015/06/15 19:49:33 (permalink)
    Samsander
    DnBrn
    Methodical2
    Question.  Can the same be said if you use Precision X boost mode?  Curious.



    I haven't tried it, but theoretically, it should since K-boost locks the card to max performance.




    I don't think you understood your findings. This is not going to work.
    The problem is not core clock fluctuating. It's the Power Draw (wattage) fluctuation what is making the card unstable. K-boost does not lock power draw. It is impossible to draw power without putting the card under actual load.
     
    Core clock fluctuation is not an issue at all. You can test this by overclocking your core to about 1420Mhz-1440Mhz and then lowering the Power Limit in Afterburner or PrecisionX for example.
    The clock will start jumping all over from as low at 1100Mhz and up to 1440Mhz boost clock, depends on how low you set the Power Limit.
    But as long as you run the games and benchmarks at ultra high settings and at 4K DSR resolution and keep the Power Draw (wattage) close to 100% without dropping to much the card will not crash.



    I did mention power.  All I've done is put out information based on what I've noticed.  As long as it's good enough information to actually work or help EVGA determine the root causes, I'm good.  I appreciate you trying to help me understand what's actually going on though.  I've actually been just under clocking for games that have the problem, I didn't really get anything very noticeable from 4k DSR aside from lower frames and Kboost just doesn't sound appealing.  This card is pretty much overkill for every game I have so I don't need it to run at max.   I'm hoping that I don't even need to deal with any of this by the time I get the 980 Ti.

    Thanks again.
    #13
    DnBrn
    New Member
    • Total Posts : 50
    • Reward points : 0
    • Joined: 2011/10/11 21:04:01
    • Status: offline
    • Ribbons : 1
    Re: Interesting 980 FTW/SC stability find 2015/06/15 19:52:07 (permalink)
    Red46
    Those workarounds are only helpful to provide the evga techs with as much info as possible. You shouldn't be required to use third party programs or, play around with extremely taxating graphic options that in graphically demanding games are not really feasible to use to begin with, just to get the card to run as it should be doing out of the box (EG not crashing and bringing the whole system down with it)



    I agree, we shouldn't have to.  Until various parties that may be involved pinpoint and fix the issue though, not much choice if you want to use the applications where the crashes are more common.
    #14
    rsabatino
    FTW Member
    • Total Posts : 1169
    • Reward points : 0
    • Joined: 2010/02/04 14:00:46
    • Location: new jersey usa
    • Status: offline
    • Ribbons : 6
    Re: Interesting 980 FTW/SC stability find 2015/06/16 15:50:00 (permalink)
    I have had problems only with the latest drivers I go back to 350.12 no tdrs

    Operation System   Microsoft Windows 11 Pro    Processor   AMD Ryzen 9 5950X 16-Core Processor    Memory   24.2 GB Free (31.9 GB Total)    Graphics Card   NVIDIA GeForce RTX 3070 Ti (8.0 GB)    Monitor   HP 32 Display (1920 x 1080 @ 60 Hz)    Disk Storage   5240.5 GB Free of 6542.2 GB    Audio   NVIDIA High Definition Audio    Motherboard   ASUSTeK COMPUTER INC. (ROG CROSSHAIR VIII HERO (WI-FI))    Mouse   Razer BlackWidow V3 Tenkeyless    Keyboard   Razer Viper
     
    #15
    Shelledfade
    New Member
    • Total Posts : 90
    • Reward points : 0
    • Joined: 2014/09/24 07:49:06
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/17 02:58:47 (permalink)
    this happens on my overclocked gtx 980.

    nvidia seems to be ignoring the issue on their forums.
    post edited by Shelledfade - 2015/06/17 03:01:08
    #16
    VGAMASTER198
    Superclocked Member
    • Total Posts : 206
    • Reward points : 0
    • Joined: 2015/06/10 14:10:49
    • Location: CLASSIFIED
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/17 03:28:36 (permalink)
    GTX 980 FTW cards, from what I've seen are being ignored and usually the problems won't be solved unless EVGA and nVidia take immediate action.

        "Like car accidents, most hardware problems are due to driver error."
     
     
     
    #17
    Red46
    New Member
    • Total Posts : 61
    • Reward points : 0
    • Joined: 2014/02/04 10:23:55
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/17 03:35:06 (permalink)
    We aren't talking about TDRs here though. We are talking about black screens that cause the whole system to crash with no other option than rebooting. Something that happens regardless of the driver version used and that points at it being some flaw within the card.
    #18
    stalinx20
    CLASSIFIED Member
    • Total Posts : 4977
    • Reward points : 0
    • Joined: 2009/01/03 08:56:23
    • Location: U.S., Michigan
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/17 03:57:16 (permalink)
    rsabatino
    I have had problems only with the latest drivers I go back to 350.12 no tdrs


    Are you overclocking?
    If so, not every driver is going to be stable with your overclocks. You might need to lower a little more to get the latest drivers stable. The latest drivers are not a problem with everyone, just a few amount of people.

    EVGA X79 Dark
    2080 Black edition
    980
    EVGA 1000 gold PSU (Gold)
    4820K CPU
    16x G-skill
    #19
    squeets
    New Member
    • Total Posts : 42
    • Reward points : 0
    • Joined: 2007/09/24 03:33:14
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/17 08:08:40 (permalink)
    stalinx20
    rsabatino
    I have had problems only with the latest drivers I go back to 350.12 no tdrs


    Are you overclocking?
    If so, not every driver is going to be stable with your overclocks. You might need to lower a little more to get the latest drivers stable. The latest drivers are not a problem with everyone, just a few amount of people.


    I don't overclock anything outside of EVGA's factory overclock and I was getting TDRs on most of the new drivers.
     
    On 350.12 I got two of them randomly over a two week period on two different games.
     
    On everything above 350.12 I was getting random driver crashes just on the desktop/using chrome.
     
    I rolled way back to 347.88 and for several weeks now haven't had any TDRs/Driver crashes in any game other than the Witcher 3.  No matter what driver I use, what settings I use, etc... The Witcher 3 eventually freezes/gives me a black screen and I need to restart my computer.  Whether 20 minutes in or after 15 hours of gameplay spread across three days, eventually it happens.
     
    -------
     
    Do you think EVGA will let us exchange our cards when they fix the problem?  I have long since passed the return/exchange date for my card at the store I bought it at.
    #20
    stalinx20
    CLASSIFIED Member
    • Total Posts : 4977
    • Reward points : 0
    • Joined: 2009/01/03 08:56:23
    • Location: U.S., Michigan
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/17 08:38:14 (permalink)
    squeets
    stalinx20
    rsabatino
    I have had problems only with the latest drivers I go back to 350.12 no tdrs


    Are you overclocking?
    If so, not every driver is going to be stable with your overclocks. You might need to lower a little more to get the latest drivers stable. The latest drivers are not a problem with everyone, just a few amount of people.


    I don't overclock anything outside of EVGA's factory overclock and I was getting TDRs on most of the new drivers.
     
    On 350.12 I got two of them randomly over a two week period on two different games.
     
    On everything above 350.12 I was getting random driver crashes just on the desktop/using chrome.
     
    I rolled way back to 347.88 and for several weeks now haven't had any TDRs/Driver crashes in any game other than the Witcher 3.  No matter what driver I use, what settings I use, etc... The Witcher 3 eventually freezes/gives me a black screen and I need to restart my computer.  Whether 20 minutes in or after 15 hours of gameplay spread across three days, eventually it happens.
     
    -------
     
    Do you think EVGA will let us exchange our cards when they fix the problem?  I have long since passed the return/exchange date for my card at the store I bought it at.


    What kind of GPU do you have? PSU?

    EVGA X79 Dark
    2080 Black edition
    980
    EVGA 1000 gold PSU (Gold)
    4820K CPU
    16x G-skill
    #21
    stalinx20
    CLASSIFIED Member
    • Total Posts : 4977
    • Reward points : 0
    • Joined: 2009/01/03 08:56:23
    • Location: U.S., Michigan
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/17 12:02:13 (permalink)
    Check out So GTX 980 FTW are very bad cards?
     
    one of the techs has a Beta BIOS you all can try. PM him.

    EVGA X79 Dark
    2080 Black edition
    980
    EVGA 1000 gold PSU (Gold)
    4820K CPU
    16x G-skill
    #22
    NDiamond
    New Member
    • Total Posts : 9
    • Reward points : 0
    • Joined: 2013/04/04 15:05:48
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/17 14:48:50 (permalink)
    Using NVidias control panel to set DSR on (under manage 3d settings if anyone wishes to know), then changing in game resolutions on SWTOR, TSW, and ETS2, played for 90 minutes each, max settings in game, at 2560x1440, with factory settings on my 980ftw and no crashes, where as before I stumbled over this thread, I would crash anywhere from 5 to 30 minutes in.

    i7-4790k
    GSkill TridentX(2133) 4x 8GB
    Asus Maximus VII Hero (Motherboard)
    1 EVGA GTX 980Ti
    Corsair 760 Platinum
    #23
    BornKill
    New Member
    • Total Posts : 10
    • Reward points : 0
    • Joined: 2014/03/27 05:46:09
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/18 02:56:09 (permalink)
    Hi guys, I have a similar issue with my 980 SC ACX 2.0 - random crashes while gaming, mostly when nothing spactacular is happening on screen.

    I have to admit i have only tested this on GTAV as i didn't have the time to try out other titles.

    I can sometimes play for few hours but somteimes for 20-30 mins before my driver stops responding.
    I've tried different drivers 353.06 - 350.12 all to similiar result - everything is working fine for about 2 days then crashing starts, with and without GF Experience installed. I only crash to desktop though, no BSODs or anything else.
    I have checked my memories, my PSU and my system (Win7 Ultimate), and it all appears to be fine.

    As many before I'have noticed sudden power changes right before the driver crash.

    DSR didn't do the trick for me, crashed within 20 mins in game.

    However when i've changed the power setting in geforce driver (can't recall how it's named at the moment) to "prefer maximum performance" in global settings tab this might have improved my experience - i was able to play 5hrs straight without a driver crash.

    I know many have tried that (from nvidia's forum) and had no positive effect, but I still decided to give it a shot.
    Perhaps it was just this weird "moment" when the card/driver decided to work as intended but i'll need to test it further.

    All in all I dont know if it's a driver, card's hardware or card's BIOS :/
    #24
    stalinx20
    CLASSIFIED Member
    • Total Posts : 4977
    • Reward points : 0
    • Joined: 2009/01/03 08:56:23
    • Location: U.S., Michigan
    • Status: offline
    • Ribbons : 0
    Re: Interesting 980 FTW/SC stability find 2015/06/18 03:05:45 (permalink)
    NDiamond
    Using NVidias control panel to set DSR on (under manage 3d settings if anyone wishes to know), then changing in game resolutions on SWTOR, TSW, and ETS2, played for 90 minutes each, max settings in game, at 2560x1440, with factory settings on my 980ftw and no crashes, where as before I stumbled over this thread, I would crash anywhere from 5 to 30 minutes in.


    YOu should provide this information to the Tech/Dave. It should not be needed to where you have to go at a higher resolution (albeit better image for you) just to make the GPU stable. It shouldn't matter what settings you put on your GPU, your gpu should not crash.

    EVGA X79 Dark
    2080 Black edition
    980
    EVGA 1000 gold PSU (Gold)
    4820K CPU
    16x G-skill
    #25
    Jump to:
  • Back to Mobile