Display driver AMD has stopped responding and has recovered

Discussion in 'Videocards - AMD Radeon' started by classick, Jun 18, 2015.

  1. classick

    classick Guest

    Messages:
    38
    Likes Received:
    0
    GPU:
    r9 280x
    @PrMinister

    I flashed the Bios all good, restarted and tested....the issue its still there... this time lost signal with 58Cº xD Dunno what to do more..still ty for the help
     
  2. afaque

    afaque Member Guru

    Messages:
    153
    Likes Received:
    3
    GPU:
    x1300/g210/hd5750/r9280x
    if at 1.256v u are getting the crash more frequently now then i should say that your gpu cant handle those clocks anymore,:bang:
     
  3. primetime^

    primetime^ Master Guru

    Messages:
    909
    Likes Received:
    17
    GPU:
    Sapphire RX Vega 64
    I thought for sure PrMinisterGR bios mod would have helped your card. For what its worth he gave you solid advise and you did what you could but not every card is user fixable.

    Now one thing he can still do is modd all the clocks to a lower number. There's still a good chance he can get it running by doing this. The only time my card does what yours is doing is if i try overclocking to high like (1150mz gpu) and at some point it will give the driver crash like yours is doing. Just a matter of bringing the perf clocks down to a stable level.

    So my suggestion is see if he can modd the clocks in another attempt to get it stable:) In the mean time you can try using msi app or afterburner to lower the 3d clocks down to 850mhz and see if it crashes at that speed. Honestly if that works then he can mod your clocks to a working freq. Just a question of finding whats stable hopefully.

    If that doesn't work then you have done most everything possible
     
    Last edited: Jun 19, 2015
  4. PrMinisterGR

    PrMinisterGR Ancient Guru

    Messages:
    8,125
    Likes Received:
    969
    GPU:
    Inno3D RTX 3090
    Or it is his PSU that is killing itself because it cannot handle the extra voltage.


    I did what you said and made a new bios. This bios has the same 1.256V for the core, but both the 3D and Turbo clocks are now 925MHz for the Core, and I lowered the memory to 1450MHz.

    You have to install it with the same procedure as above. The name of the rom is Tahiti-extraV-LOWER-CORE-VRAM.rom.

    Try it, you have nothing to lose.

    I HONESTLY believe you should try to see if it does the same with another PSU, before you decide to buy another GPU.
     

  5. raclimja

    raclimja Guest

    Messages:
    28
    Likes Received:
    0
    GPU:
    EVGA GTX 970 SC
    It might be due to CPU or RAM.

    for RAM set it to the lowest possible speed and loosen the timing and see if there would be any difference.

    If problem still persist, try to use the default setting for your CPU.

    If problem still persist make a CMOS reset.

    If problem still persist then make a dual boot (to isolate ALL potential software problem).

    If problem still persist then it is 100% a faulty hardware (either your GPU or PSU failing).

    Also note driver crashes on defective graphics card are 90% VRAM related (while artifacts are GPU core related).
     
  6. primetime^

    primetime^ Master Guru

    Messages:
    909
    Likes Received:
    17
    GPU:
    Sapphire RX Vega 64
    you must have missed this post otherwise you wouldn't have said this
    i get it...its easy to miss posts:wanker:
     
  7. raclimja

    raclimja Guest

    Messages:
    28
    Likes Received:
    0
    GPU:
    EVGA GTX 970 SC
    even if you reinstall or uninstall the drivers, there would still be some dlls, registry lingering on your system.

    As an example a straight swap from NVIDIA GPU to an R9 290 or even just changing the PCIE slot makes the AMD card perform very poorly without completely cleaning everything.

    I also have a bad memory that could pass memtest86, AIDA64 and prime95 for more than 6hours without problem but would cause my regular programs to crash.
     
  8. classick

    classick Guest

    Messages:
    38
    Likes Received:
    0
    GPU:
    r9 280x
    @PrimeTime...Did with that new Bios and the issue its still there :S..

    Just a note i do those Bio flashes in Switch number 2..And with those Bios the issue happens at 55-65Cº....In the Switch Number 1 the issue goes to 70-75Cº

    Raclimja@ I already did all that.
     
  9. primetime^

    primetime^ Master Guru

    Messages:
    909
    Likes Received:
    17
    GPU:
    Sapphire RX Vega 64
    Did you verify with gpu-z the correct clocks in sensor tab? If so she's done unfortunately
     
    Last edited: Jun 20, 2015
  10. PrMinisterGR

    PrMinisterGR Ancient Guru

    Messages:
    8,125
    Likes Received:
    969
    GPU:
    Inno3D RTX 3090
    Before you get rid of the GPU, find a way to check with another PSU.
     

  11. classick

    classick Guest

    Messages:
    38
    Likes Received:
    0
    GPU:
    r9 280x
    Can be a lower psu? Lets say a 500 Nox?..i dnt like Nox but its the only i can test it i gues xD
     
  12. primetime^

    primetime^ Master Guru

    Messages:
    909
    Likes Received:
    17
    GPU:
    Sapphire RX Vega 64
    if its one of those generic supplys that self destruct at less than half there rated supply then i might not take the chance. So it depends on what on the inside the power supply really, and to be honest i have never heard of that one. You could easily pull 400 watts or more running furmark.
     
  13. PrMinisterGR

    PrMinisterGR Ancient Guru

    Messages:
    8,125
    Likes Received:
    969
    GPU:
    Inno3D RTX 3090
    Unless it's made out of tin, try with it and make sure you don't have anything overclocked. It should be enough for a run or two.
     
  14. classick

    classick Guest

    Messages:
    38
    Likes Received:
    0
    GPU:
    r9 280x
    My bad guys..the PSU is this i can test Nox ATX urano 2 630w
     
  15. PrMinisterGR

    PrMinisterGR Ancient Guru

    Messages:
    8,125
    Likes Received:
    969
    GPU:
    Inno3D RTX 3090
    Just test with something else, anything really, before you get rid of the card.
     

  16. primetime^

    primetime^ Master Guru

    Messages:
    909
    Likes Received:
    17
    GPU:
    Sapphire RX Vega 64
    Quote:
    Originally Posted by classick View Post
    My bad guys..the PSU is this i can test Nox ATX urano 2 630w
    Like PrMinister said....couple of runs should be safe enough just to verify what we already suspect and if you get the same results and you might as well store it away for emergency back up use.


    PrMinister i have a question for you regarding modding the bios. I modded mine earlier and set the clock speed 20mhz higher just to make it an even number of 1100mhz. but in my testing the clocks only went to 1080mhz...even though gpu-z main tab showed 1100mhz but in the sensor tab it only went to the black bios freq of 1080.
    If your up to it i might try and see if you have any better luck, or if you heard of any bios like xfx being a bit hard to modd correctly. VBE7 and gpuz acted like i did it correctly but its like theres a part of the bios that doesnt change after a flash....Not a problem by any means but just an unanswered question. My card has no switch but i do have a spare card if needed and a bootable usb with atiflash as well on standby
     
  17. PrMinisterGR

    PrMinisterGR Ancient Guru

    Messages:
    8,125
    Likes Received:
    969
    GPU:
    Inno3D RTX 3090
    It has nothing to do with the modding. It's the driver/CCC settings. After you change the frequencies of your cards this might happen. The best idea is to uninstall the driver using DDU and clean install it again.
     
  18. primetime^

    primetime^ Master Guru

    Messages:
    909
    Likes Received:
    17
    GPU:
    Sapphire RX Vega 64
    Well i'll be dang...That did not occur to me and it might explain the abnormal boot times after flashing the modded bios (at least it seemed a bit longer)

    Edit: much appreciated that worked fine . btw any way to get full gpu clock speed with an open browser? Funny things is my xfx 7970 would go to full clocks whether a browser was open or not But my replacement 280x has to have all browser closed or it will only go to 850mhz. but i can say i have yet to get the power play browser flicker my old 7970 had so in that respect im cool with it. 3dmark likes to get locked into 850 as well (something to do with flash?)
     
    Last edited: Jun 21, 2015
  19. classick

    classick Guest

    Messages:
    38
    Likes Received:
    0
    GPU:
    r9 280x
    Hi guys, did the test with another psu and the results are the same..well i guess its time for a new gpu xD time to see the market for one xD

    ty for all the help
     
  20. classick

    classick Guest

    Messages:
    38
    Likes Received:
    0
    GPU:
    r9 280x
    Guys i got a second hand 280x..same as mine..and im facing problems too LOL!!!!

    Its not the crash....when i start FurMArk..the temps jump...They skip all the rest...Idle is 40 and when i start the FurMark it goes to 75 in 1 sec

    Do i need change the thermal paste or its other issue?
     

Share This Page