R9 290x Black Screen Crash

Discussion in 'Videocards - AMD Radeon' started by nickrorschach, Nov 4, 2013.

  1. nickrorschach

    nickrorschach Member

    Messages:
    43
    Likes Received:
    0
    GPU:
    AMD R9 290x 4GB
    Thing is, this seemed like the perfect stop gap, 4GB of VRAM etc

    Has anyone ever had a response from an AMD support ticket? Am i kidding myself i will ever get a reply?

    I just wish it worked.
     
  2. DesGaizu

    DesGaizu Ancient Guru

    Messages:
    3,712
    Likes Received:
    74
    GPU:
    AORUS 3060TI
    I feel your pain. I want to love this card I really do. I really don't care for nvidia vs amd.

    The card performs great has 4gb of fast vram lots of bandwidth and all in all is a good price. Hell even the noise and heat doesn't bother me (Headphones most of the time for me) but the bloody thing refuses to play nice.

    also I recorded one of the crashes too

    http://www.youtube.com/watch?v=WAQoiblBJwE
     
  3. Fox2232

    Fox2232 Ancient Guru

    Messages:
    11,809
    Likes Received:
    3,366
    GPU:
    6900XT+AW@240Hz
    [​IMG]

    This is Frame I got from DesGaizu YT video. Interesting how upper part is cloned down in pattern.
    Looks like some buffer overflow at wrong place. But does not mean it may not be caused by bad memory timing, since 290(x) has new memory controller and that may require a lot of testing.

    Have you tested higher memory clocks if it happens more often? And since it does not happen on desktop, have you tried limiting memory clocks?
     
  4. Falkentyne

    Falkentyne Master Guru

    Messages:
    514
    Likes Received:
    55
    GPU:
    Sapphire HD 7970 Ghz Ed.
    If you notice, right before the screen turns black, the entire scene becomes a mess of vertical color bars. Looks like either a GPU crash or some failsafe protection shut off the card. But GPU 'crashes' that come from an unstable GPU (overclocking) are not color bars, they're application freezes, followed by an alt-tabbable black screen (VPU recovery). Here, the card just like shuts down or something...
     

  5. DesGaizu

    DesGaizu Ancient Guru

    Messages:
    3,712
    Likes Received:
    74
    GPU:
    AORUS 3060TI
    Wow great spot, I went back and watched the original 1080p file and screened the last 2 frames before the crash

    [​IMG]

    [​IMG]
     
  6. kcuestag

    kcuestag Master Guru

    Messages:
    904
    Likes Received:
    0
    GPU:
    Gigabyte GTX980Ti G1
    That's interesting, might want to report that to AMD so they can have a look.

    I should get my PC up and running again tomorrow evening, if I get the blackscreens again I'll make sure I record it to see if I get the same behavior.
     
  7. Fox2232

    Fox2232 Ancient Guru

    Messages:
    11,809
    Likes Received:
    3,366
    GPU:
    6900XT+AW@240Hz
    @DesGaizu: It looks like part of GPU is unable to change ON/OFF state of transistors, but other part tries to finish what is in buffer.
    I guess there is one more real frame between 2 which is not on video. This may be serious issue possibly resolvable by bios update if it's so.

    Do you have CPU base clock changed even by few Hz? If yes can you try to set it to 10kHz (my MB does not post if I change it by more than 80Hz and may cause stability issue for card).
    Other thing is your Sound Card. It is on PCIe too. Can you test with it removed? So GPU has all 16 lanes for itself?
    If that does not help, can you try bios default values so it resets all remembered data moved between runs? Maybe The Postman hasn't fixed it by switching to PCIe 3.0 but something else there got reset along the way.
     
  8. Osjur

    Osjur Member

    Messages:
    34
    Likes Received:
    0
    GPU:
    Radeon Fury X
    Also haunted by those black screens, though in my case it never freezes.
    The monitor just goes black for about 5 seconds and then comes back and might go out again in couple of seconds.

    It might happen in the first minute you enter the game or after 2h of gaming so there's no pattern to it. Also happens when you idle on desktop.

    I have also noticed that when it does happen, gpu utilization takes a big drop and core voltage drops to either uvd or low performance state before jumping back to high performance state.

    My card is watercooled so temps are not the problem, also my card has Hynix memory and most of the guys having problems have Elpida but they seem to hardlock instead.

    Downclocking memory has no effect in my case.

    Powertune playing tricks with us?
     
    Last edited: Nov 12, 2013
  9. DesGaizu

    DesGaizu Ancient Guru

    Messages:
    3,712
    Likes Received:
    74
    GPU:
    AORUS 3060TI
    Hey, cpu base glock is default. My sound card is on pci not pcie but ill try without it if you like.

    I've flashed a brand new bios update a day ago after having these problems so everything is pretty much default.
     
  10. eclap

    eclap Banned

    Messages:
    31,495
    Likes Received:
    4
    GPU:
    Palit GR 1080 2000/11000
    did you disable ulps in msia? also, I would try downclocking the vram 100-200mhz, or up the vram voltage. I have a feeling it's the vram being unstable. I know it should run fine out the box, but worth trying just to eliminate my theory.

    Good spot Fox btw, mas orli zrak :)
     

  11. Fox2232

    Fox2232 Ancient Guru

    Messages:
    11,809
    Likes Received:
    3,366
    GPU:
    6900XT+AW@240Hz
    Sorry, for Sound card misunderstanding. Saw STX. Then it should not be issue, PCIe has it's own dedicated lanes.

    @Eclap: hh, no I wear glasses. But I have good reaction thanks to being used on 120Hz.
     
    Last edited: Nov 12, 2013
  12. DesGaizu

    DesGaizu Ancient Guru

    Messages:
    3,712
    Likes Received:
    74
    GPU:
    AORUS 3060TI
    No I've not not disabled ULPS should I try?

    I really don't want to mess with vram voltages as I can't get any program to read them properly. I can downclock it though.
     
  13. nickrorschach

    nickrorschach Member

    Messages:
    43
    Likes Received:
    0
    GPU:
    AMD R9 290x 4GB
    So, please go easy on me, i'm a developer, not great with hardware. My clocks will sit at about 1000.0 for core and 1250.0 for memory (GPUZ), temps are at about 86 - 90 always, i have good airflow in my case. This is normal for the 290x, yeah?

    But sometimes in games i'll see very brief drops core drops down to 340.1 core \ 150.0 memory . Very rare, but for one line in the logs. Or sometimes the core will go something like 998.0 or 993.6 and the memory seems to stay static. I think this is normal, but just checking.

    In 2D, where i have zero crashed of any kind (left machine on for 3 days straight, writing a big document), my core clock hovers about like mad, so:

    300.0
    350.8
    356.3

    etc, always those sort of numbers, but my memory will go between just two numbers

    1250.0
    &
    150.0

    again and again

    The GPUZ graphs look like they are doing this in a pretty rhythmic pattern. When gaming my memory mostly stays at 1250.0, with those very brief drops to 150.0 .

    Just to be clear i am running the uber BIOS, but the crashes happen in both stock or uber. Short of changing BIOS with the switch, I have changed nothing with the card, no voltage or clocks etc.

    Temps look totally fine.

    There is some guy in this thread:

    http://battlelog.battlefield.com/bf...065218069487298/last/post_2955065218199968194

    ... who i think is claiming he fixed the crashes by locking the clocks at their highest rates? If i have read it correctly?

    Perhaps this helps keeps the discussion moving a bit.

    Thanks very much for all your help, i really hope we get to the bottom of this somehow or AMD make *some* kind of statement.
     
    Last edited: Nov 12, 2013
  14. IcE

    IcE Don Snow Staff Member

    Messages:
    10,692
    Likes Received:
    73
    GPU:
    Zotac GTX 1070 Mini
    It's a shame you guys have this issue, it's one of the reasons I stayed away.
     
  15. mitzi76

    mitzi76 Ancient Guru

    Messages:
    8,722
    Likes Received:
    19
    GPU:
    MSI 970 (Gaming)
    am hoping my return to amd is going to be successfull. but tbh am kinda regretting the purchase...maybe am being negative..
     

  16. eclap

    eclap Banned

    Messages:
    31,495
    Likes Received:
    4
    GPU:
    Palit GR 1080 2000/11000
    try disabling ulps. also try downclocking the vram by some 100-200mhz.
     
  17. The Postman

    The Postman Ancient Guru

    Messages:
    1,773
    Likes Received:
    0
    GPU:
    MSI 980 TI Gaming
    I ve seen those crashes before when testing overclocks with different cards in the past. Especially vram or low voltage on core.

    Btw do you use hdmi?
     
  18. DesGaizu

    DesGaizu Ancient Guru

    Messages:
    3,712
    Likes Received:
    74
    GPU:
    AORUS 3060TI
    Will do, I'll test some bf4 in a bit.

    Everything is stock I don't even have any overlay/oc software installed.

    Yes I am on HDMI but this is not just a picture corruption sound goes as well most of the time.
     
  19. The Postman

    The Postman Ancient Guru

    Messages:
    1,773
    Likes Received:
    0
    GPU:
    MSI 980 TI Gaming
    Sorry I didnt mean to say you overclock the card.

    If ULPS doesnt solve your problem, could you try downclocking your core to 900mhz and vram to 1100? This is just to check there is no lack of juice for those parts.

    Btw in CCC overdrive, do you have maximum fan speed set at 100%? If you set at that doesnt mean it will run at that speed.
     
  20. DesGaizu

    DesGaizu Ancient Guru

    Messages:
    3,712
    Likes Received:
    74
    GPU:
    AORUS 3060TI
    Just wanted to make it clear that I wasn't running some silly oc on anything :)

    This psu is very high quality I've had 2*680's running fine I honestly can't see that been the problem unless it's died at the same time me getting this 290x.

    I will try downclocking next.

    overdrive is not even on at the moment, i'll set it to 100% before i jump into bf4.
     

Share This Page