1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

GeForce GTX series 980 Ti and below will not get Adaptive Sync Support

Discussion in 'Frontpage news' started by Hilbert Hagedoorn, Jan 16, 2019.

  1. user1

    user1 Maha Guru

    Messages:
    1,339
    Likes Received:
    417
    GPU:
    hd 6870
    normal Displayport supports direct Drive mode, they are electrically identical. the only difference is in the data packet information. whether the display has a scaler or not makes no difference if the display supports adaptive sync.
    im not saying there can't be limitations, since switching refresh in realtime constantly is alot different than clocking down the display temporarily,( which is probably why thehd 7000 series "freesync" only supported clocking down to 24hz to watch movies at native framerate.) but when the hardware clearly supports doing adaptive sync in this manner, it is not a satisfactory explanation.
     
  2. Astyanax

    Astyanax Ancient Guru

    Messages:
    2,471
    Likes Received:
    612
    GPU:
    GTX 1080ti

    Except it does.
     
  3. user1

    user1 Maha Guru

    Messages:
    1,339
    Likes Received:
    417
    GPU:
    hd 6870
    if you could be more specific, i think we all might learn something.
     
  4. Astyanax

    Astyanax Ancient Guru

    Messages:
    2,471
    Likes Received:
    612
    GPU:
    GTX 1080ti
    pasting VESA documentation is against copyright law.
     

  5. user1

    user1 Maha Guru

    Messages:
    1,339
    Likes Received:
    417
    GPU:
    hd 6870
    I there is such a thing as paraphrasing , dont think thats illegal. everything i've read shows that eDP uses the AUX data link to do display control ,standard DP 1.2 uses the AUX channel for general data transfer(can be used for things like USB hubs) and as far as I can tell the optional adaptive sync feature uses this aux channel aswell, seems like there is virtually no reason that you could not transmit slightly different data over a general data link that both standards support. or maybe I am crazy.


    edit: I' not against being wrong, I like learning new things, if there is a reason, even a "it doesn't do this _vagueterm_ because _vagueterm_ would be far more useful, than " except it does"
     
    Last edited: Jan 17, 2019
  6. K.S.

    K.S. Maha Guru

    Messages:
    1,298
    Likes Received:
    225
    GPU:
    RTX 2080 GAMING OC
    Ref. #115 cite - "NV/support VESA Adaptive-Sync tech/GSYNC Compatible"

    "DisplayPort version 1.2a was released in January 2013. After a proposal from AMD, VESA later adapted the Panel-Self-Refresh feature for use in standalone displays and added it as an optional feature of the main DisplayPort standard under the name "Adaptive-Sync" in version 1.2a"

    Note: Support for Adaptive-Sync is not required for a display to be DisplayPort 1.2a-compliant.
    EDIT: Support for Adaptive-Sync is included among requirements for a display to be DisplayPort 1.3-compliant.
     
    Last edited: Jan 17, 2019
    BlackZero likes this.
  7. LesserHellspawn

    LesserHellspawn Master Guru

    Messages:
    627
    Likes Received:
    4
    GPU:
    3x GTX980ti
    Well, as long as my triple 980ti and my 1440p gsync screen plod along, I'm fine. I have no plans to upgrade anytime soon, at least until Nvidia releases cards that are affordable again, or until AMD actually get competitive.
     
  8. fantaskarsef

    fantaskarsef Ancient Guru

    Messages:
    10,555
    Likes Received:
    2,775
    GPU:
    1080Ti @h2o
    First of all, what did we expect from Nvidia? Like @Chastity named the thread, hell freezes, and suddenly Nvidia supports Freesync at all.
    The truth is probably somewhere in between things, something like this could be what an Nvidia rep would say when they'd talk to somebody in private, off the records:

    "Yeah, we could support it. But to be honest, we don't want to push our money into engineering hours to support Freesync monitors and old cards. We don't earn on the monitors directly, only with graphics card sales. We only even did the Freesync support thing because we want to sell some more of our stock... if we wouldn't be able to keep up with current demands on cards, if people would grab them off the trucks already, we would only smile and rather put our efforts into promoting Gsync further.
    So... we don't want to push engineering hours into 4, 5 year old cards which you bought back then, since you already got your money out of them after this time, and... well, we would be silly to give you a reason to not buy a new graphics card, right?
    So I hope you understand, we would much more like you to buy a new RTX card, and if that's too expensive for you, maybe a Pascal card?"

    ;)
     
    BlackZero likes this.
  9. Fox2232

    Fox2232 Ancient Guru

    Messages:
    9,469
    Likes Received:
    2,082
    GPU:
    5700XT+AW@240Hz
    HD7970=r9-280X => GCN1.0
    HD7790=r7-260X => GCN1.1

    Those were exactly same chips, just with different vBIOSes enabling additional features.

    Therefore HD7790 should be as capable to do freesync via vBIOS update as r7-260X is.
    But r9-280X (HD7970) is one generation older. And does not even have that mentioned TrueAudio.
     
    warlord likes this.
  10. DiceAir

    DiceAir Maha Guru

    Messages:
    1,349
    Likes Received:
    13
    GPU:
    Galax 980 ti HOF
    Luckily I have a true Gsync display (PG279q) So I can still use adaptive sync on my 980 ti. I feel no need to upgrade my gpu just yet. Was thinking to upgrade to next gen after 1080 ti but then they decide to launch RTX series so now i will wait for maybe next gen as RTX is still new.

    The other reason is that Gsync is one reason why I don't upgrade. I get 90 - 110+ FPS on BF5 medium settings and it still feels smooth enough for me to not worry about upgrading my gpu yet.

    I'm still pissed at Nvidia for not supporting this on 980 ti.
     

  11. Denial

    Denial Ancient Guru

    Messages:
    12,217
    Likes Received:
    1,390
    GPU:
    EVGA 1080Ti
    Nvidia's argument is that it can't do Adaptive on Maxwell because Maxwell cards don't support DP1.2a, only DP1.2 - they aren't claiming that the architecture itself is preventing them.

    My argument is that flashing a HD7790 won't enable Freesync because the same reason, it's DP1.2 not DP1.2a. Sure the architectures are the same but the physical connector isn't.
     
  12. Miguel73

    Miguel73 New Member

    Messages:
    3
    Likes Received:
    0
    GPU:
    GTX970 / 4GB
    Hi all
    Saw this video in the internet. Is it real???
     
  13. Miguel73

    Miguel73 New Member

    Messages:
    3
    Likes Received:
    0
    GPU:
    GTX970 / 4GB
  14. -Tj-

    -Tj- Ancient Guru

    Messages:
    16,294
    Likes Received:
    1,423
    GPU:
    Zotac GTX980Ti OC
    And the funny thing is dp1.2a was released in 2013 according to some posts here, while 980ti in 2015..
     
  15. Fox2232

    Fox2232 Ancient Guru

    Messages:
    9,469
    Likes Received:
    2,082
    GPU:
    5700XT+AW@240Hz
    Ah, you mean that one of wires required is not physically wired? Well that can be argument. But then we would have to go into signaling. And see if Adaptive-Sync uses some other than standard data communication to handle this.

    But as far as original AMD's demonstration went, they used existing screens and updated its firmware to do Freesync. Therefore it either had some extra wire (unlikely), or it is just some kind of datagram on regular DP1.2.

    ...
    And mere look at DP connector wiring datasheet. There is no special wire. Physically is is same thing since DP inception till now.
     

  16. user1

    user1 Maha Guru

    Messages:
    1,339
    Likes Received:
    417
    GPU:
    hd 6870
    https://www.vesa.org/wp-content/uploads/2010/12/DisplayPort-DevCon-Presentation-eDP-Dec-2010-v3.pdf
    https://www.vesa.org/wp-content/uploads/2011/01/ICCE-Presentation-on-VESA-DisplayPort.pdf

    https://www.hardwaresecrets.com/introducing-the-panel-self-refresh-technology/2/

    these are some relevant still publicly available documents,

    for edp panel self refresh appears to be what is used, where the panel stores the last image it recieved and uses it until a command is issued over aux. by the gpu

    unfortunately there is very little information about what exactly is used for the dp 1.2a's adaptive sync,

    I'm assuming the aux data link is used to send a command to the display to trigger an update(not exactly the same as psr), which would make sense since alot of displays only need a firmware update to support adaptive sync.

    now if maxwell can issue commands fast enough ( which it appears it can) the question is whether or not it can issue what i am assuming is a "new" command, other than the standard dp 1.2 ones.

    but until someone spills the magic beans as to what is actually going on, thats about as far as can be speculated.
     
    carnivore, Embra and Fox2232 like this.
  17. Astyanax

    Astyanax Ancient Guru

    Messages:
    2,471
    Likes Received:
    612
    GPU:
    GTX 1080ti
    GPU and board designs are set in stone years before the final product is put to production.
     
  18. ManofGod

    ManofGod Maha Guru

    Messages:
    1,328
    Likes Received:
    38
    GPU:
    Sapphire R9 Fury Nitro
    So? Stop being cheap and support Nvidia again! MNvGA! ;) :D
     
  19. vbetts

    vbetts Don Vincenzo Staff Member

    Messages:
    14,510
    Likes Received:
    1,060
    GPU:
    RTX 2070FE
    Can you source this? You've posted a few things already claiming things with no source.
     
    warlord and -Tj- like this.
  20. Astyanax

    Astyanax Ancient Guru

    Messages:
    2,471
    Likes Received:
    612
    GPU:
    GTX 1080ti
    i'm sure one of the asic creators that lurk around here can confirm, the board is designed on paper with components existing at the time, not components that might exist when the card is put into production (unless its something custom specifically obtained by the board designer).

    nvidia would have had Turings board designs ready for well over a year running tests on it before the gpu was even finalized, you don't need the finished silicon to test power delivery, component combinations or the amount of layers the pcb may need.
     

Share This Page