Unstable OC in recent drivers

Discussion in 'Videocards - NVIDIA GeForce Drivers Section' started by Terepin, Dec 7, 2017.

  1. Terepin

    Terepin Guest

    Messages:
    873
    Likes Received:
    129
    GPU:
    ASUS RTX 4070 Ti
    Dunno what happened, but recent drivers made my OC highly unstable. In games like Evolve, or Serious Sam Fusion, everything goes fine, until GPU usage spikes to 100% without reason and it gives "device hung" error. Both games were fine before with my OC. Is this a known bug or something?
     
  2. cryohellinc

    cryohellinc Ancient Guru

    Messages:
    3,535
    Likes Received:
    2,974
    GPU:
    RX 6750XT/ MAC M1
    Off topic but, good lord, how do you even play that?

    On topic - well perhaps your components are getting "bad" due to previously too high overclock? Have you considered this? What temps are you running?
     
  3. Terepin

    Terepin Guest

    Messages:
    873
    Likes Received:
    129
    GPU:
    ASUS RTX 4070 Ti
    By having fun, duh.
    The card is only few months old and temps are fine. In fact, Evolve barely stress the card.
     
  4. cryohellinc

    cryohellinc Ancient Guru

    Messages:
    3,535
    Likes Received:
    2,974
    GPU:
    RX 6750XT/ MAC M1
    Considering all the controversy around it, highly doubtful.
    Define "fine". Secondly if you suggest same settings worked before, and issue started appear recently -have you made a barebones driver installation, with a DDU cleanup?
    [​IMG]
     

  5. Terepin

    Terepin Guest

    Messages:
    873
    Likes Received:
    129
    GPU:
    ASUS RTX 4070 Ti
    Fine as in fans are doing their job. I did use DDU, but it didn't help.
     
  6. cryohellinc

    cryohellinc Ancient Guru

    Messages:
    3,535
    Likes Received:
    2,974
    GPU:
    RX 6750XT/ MAC M1
    Try to use DDU, followed by a barebones installation approach.
     
  7. fantaskarsef

    fantaskarsef Ancient Guru

    Messages:
    15,693
    Likes Received:
    9,573
    GPU:
    4090@H2O
    Just to mention that, later drivers changing overclocks has been here before, this is not the first time. Did you try to step down a single bin of your GPU? That usually helps, if you can't push in any more voltage.
     
  8. Terepin

    Terepin Guest

    Messages:
    873
    Likes Received:
    129
    GPU:
    ASUS RTX 4070 Ti
    I had to go from +150 to +115. That's freaking huge step.
     
  9. fantaskarsef

    fantaskarsef Ancient Guru

    Messages:
    15,693
    Likes Received:
    9,573
    GPU:
    4090@H2O
    Well that's something you didn't meantion ;)
    What's the voltage difference for that OC drop with your card? At what voltage could you reach +150 and +115 before?
    Can you install older drivers to see if the OC is still stable?
     
  10. Terepin

    Terepin Guest

    Messages:
    873
    Likes Received:
    129
    GPU:
    ASUS RTX 4070 Ti
    Voltage is at +100% the whole time.
     

  11. MrBonk

    MrBonk Guest

    Messages:
    3,385
    Likes Received:
    283
    GPU:
    Gigabyte 3080 Ti
    I have had to redo my oc with driver changes before. Nothing new to me.
     
    -Tj- likes this.
  12. Terepin

    Terepin Guest

    Messages:
    873
    Likes Received:
    129
    GPU:
    ASUS RTX 4070 Ti
    Well, now it works fine again. I hope it will stay this way.
     
  13. IceVip

    IceVip Master Guru

    Messages:
    914
    Likes Received:
    224
    GPU:
    MSI 4090 Liquid X
    Oddly enough I've noticed this myself. I had my 1080 at +200, reaching 2114 in every game. But ever since the 388.x I started dropping it down until i found a stable +160 which does around 2038-2050
    but even then in some cases games don't even boost the card normally even though gpu utilization is at max. Like NFS Payback and Battlefront 2. In older titles since my voltage is at +100 it goes up
    to 1.091v, but in newer titles it does 1.014v max and the core does 1876. Even if i change clocks in whilst watching them in real time in the game nothing changes. GTA 5 started crashing everytime
    the core goes pass 2050 for some reason. But the card is stable as hell. I can do almost 2.2Ghz at synthetic benchmarks, temps are fine, around 75c at 2.1ghz but still, voltage spikes around and thats
    what i think makes everything unstable. Older drivers didn't do this though.
     
  14. Terepin

    Terepin Guest

    Messages:
    873
    Likes Received:
    129
    GPU:
    ASUS RTX 4070 Ti
    Damn, it crashed again. I lowered voltage increase by 10% and so far it is stable.
     
  15. Dragondale13

    Dragondale13 Ancient Guru

    Messages:
    1,527
    Likes Received:
    244
    GPU:
    GTX 1070 AMP! • H75
    Try setting Pci-E slot config in bios to "auto"...let Windows determine the port speed and try again with your previous voltage setting.
     
    Last edited: Dec 20, 2017

  16. CK the Greek

    CK the Greek Maha Guru

    Messages:
    1,316
    Likes Received:
    37
    GPU:
    RTX 2060S
    yeah,good call for lowering it, but I think you may be able to lower more. Doesn't need that much.
     

Share This Page