Nvidia Turing Architecture [2018]

Discussion in 'Architecture and Products' started by pharma, Sep 13, 2018.

Tags:
  1. Malo

    Malo Yak Mechanicum
    Legend Veteran Subscriber

    Joined:
    Feb 9, 2002
    Messages:
    6,953
    Likes Received:
    3,032
    Location:
    Pennsylvania
    It was :) Probably end up $2k plus
     
  2. ToTTenTranz

    Legend Veteran Subscriber

    Joined:
    Jul 7, 2008
    Messages:
    9,780
    Likes Received:
    4,431
    Keep in mind that these new cards have the added value of having actually passed through all the proper quality assurance steps.
    The probability of having to return these will be way lower than a 2080 Ti. :yep2:
    :thumbsup:
     
    nnunn likes this.
  3. DavidGraham

    Veteran

    Joined:
    Dec 22, 2009
    Messages:
    2,718
    Likes Received:
    2,454
    OCASM, nnunn, Lightman and 3 others like this.
  4. Malo

    Malo Yak Mechanicum
    Legend Veteran Subscriber

    Joined:
    Feb 9, 2002
    Messages:
    6,953
    Likes Received:
    3,032
    Location:
    Pennsylvania
    What's the customer for a Titan like that vs the Quadro range? You get better than a Quadro 5000 for more money but no access to Quadro drivers?
     
  5. DavidGraham

    Veteran

    Joined:
    Dec 22, 2009
    Messages:
    2,718
    Likes Received:
    2,454
    They are catering for mainly the data scientists, AI, ray tracing content creators and rich gamers with this release.
    A Titan RTX is essentially a slightly faster Quadro RTX 6000, but without Quadro drivers. Which explains the price difference.
     
    A1xLLcqAgt0qc2RyMz0y likes this.
  6. vipa899

    Regular Newcomer

    Joined:
    Mar 31, 2017
    Messages:
    922
    Likes Received:
    354
    Location:
    Sweden
  7. ToTTenTranz

    Legend Veteran Subscriber

    Joined:
    Jul 7, 2008
    Messages:
    9,780
    Likes Received:
    4,431
    I think after the Vega FE opened up pro driver optimizations to "prosumers", nvidia followed suit with the Titan range.
    IIRC the Titan drivers now carry the optimizations but not the proper certifications from professional software makers.

    In practice, I think this means people who buy Titan/FE add-in boards get about the same performance as the Quadro/Radeon-Pro counterparts, but those cards can't go into new graphics workstations from Dell, HP, etc.

    But this is for Quadro, not Tesla. nvidia has forbidden (or tried to forbid) compute farm builders from using Titans and Geforces in their systems, through the driver EULAs, as far as I know.
     
  8. Voxilla

    Regular

    Joined:
    Jun 23, 2007
    Messages:
    708
    Likes Received:
    279
    Why would anyone buy this, if you can have two RTX2080Ti for that price ?
    At least not for gaming graphics.
    It's clearly intended for the NN training folks only (as can be seen from the removal of half speed mixed precision training limitation on Geforce).
    This is not the kind of Titan that used to be. :-(
     
    #208 Voxilla, Dec 3, 2018
    Last edited: Dec 3, 2018
    CarstenS likes this.
  9. vipa899

    Regular Newcomer

    Joined:
    Mar 31, 2017
    Messages:
    922
    Likes Received:
    354
    Location:
    Sweden
    Probally for people that want even faster then 2080ti. 24gb gddr6 also enables higher resolutions and its RT performance is better.
    2070 is a nice mid range alternative, 2080 high end, Ti ultra, titan elite range.
     
  10. 3dilettante

    Legend Alpha

    Joined:
    Sep 15, 2003
    Messages:
    8,119
    Likes Received:
    2,864
    Location:
    Well within 3d
    Micron's datasheet for GDDR6 includes a clamshell mode. The individual channels get split into 8 per channel.
     
    iamw and Kaotik like this.
  11. Ryan Smith

    Regular

    Joined:
    Mar 26, 2010
    Messages:
    609
    Likes Received:
    1,036
    Location:
    PCIe x16_1
    You get full FP32 accumulate on the tensor cores for a fraction of the price of a Quadro RTX 6000. That alone will sell it to droves of data scientists.

    Tesla: Servers
    Quadro: Content creation & CAD
    Titan: Developers & data scientists; develop on Titan now, then sell these users some Teslas when they need to scale out
    GeForce: Consumer gaming & everyone who can't afford a better card.
     
  12. fellix

    fellix Hey, You!
    Veteran

    Joined:
    Dec 4, 2004
    Messages:
    3,485
    Likes Received:
    396
    Location:
    Varna, Bulgaria
    First glimpse into the TU102 die (IR photograph):

     
    Lightman, pharma and vipa899 like this.
  13. vipa899

    Regular Newcomer

    Joined:
    Mar 31, 2017
    Messages:
    922
    Likes Received:
    354
    Location:
    Sweden
    Very impressive indeed what RTX series are doing, like DF mentioned extremely impressive graphics with RT and now possible @60FPS, you basicly get 2080TI performance with a 2070 after the new patch. Deep learning, DLSS can only improve things, as do new game patches and nvidia drivers.



    Variable rate shading seems intresting too. Damn these new RTX gpu's are loaded with new features, and on top of that normal rasterisation performance is much faster then the 10 series.
     
    sonen, OCASM and pharma like this.
  14. Shifty Geezer

    Shifty Geezer uber-Troll!
    Moderator Legend

    Joined:
    Dec 7, 2004
    Messages:
    40,510
    Likes Received:
    10,886
    Location:
    Under my bridge
    Is there any raytracing in that video?
     
  15. vipa899

    Regular Newcomer

    Joined:
    Mar 31, 2017
    Messages:
    922
    Likes Received:
    354
    Location:
    Sweden
  16. DavidGraham

    Veteran

    Joined:
    Dec 22, 2009
    Messages:
    2,718
    Likes Received:
    2,454
  17. pharma

    Veteran Regular

    Joined:
    Mar 29, 2004
    Messages:
    2,884
    Likes Received:
    1,591
    December 14, 2018
     
    vipa899, sonen and OCASM like this.
  18. pharma

    Veteran Regular

    Joined:
    Mar 29, 2004
    Messages:
    2,884
    Likes Received:
    1,591
    Using Turing Mesh Shaders: NVIDIA Asteroids Demo
    December 18, 2018

    https://devblogs.nvidia.com/using-turing-mesh-shaders-nvidia-asteroids-demo/

    Edit: Clarification.
     
    #219 pharma, Dec 21, 2018
    Last edited: Dec 22, 2018
    Jupiter, Lightman and vipa899 like this.
  19. Ethatron

    Regular Subscriber

    Joined:
    Jan 24, 2010
    Messages:
    856
    Likes Received:
    260
    Why do they have such a problem with attribution? This statement basically sh*ts on the guys bringing cluster-culling into practice.
     
    milk likes this.
Loading...

Share This Page

  • About Us

    Beyond3D has been around for over a decade and prides itself on being the best place on the web for in-depth, technically-driven discussion and analysis of 3D graphics hardware. If you love pixels and transistors, you've come to the right place!

    Beyond3D is proudly published by GPU Tools Ltd.
Loading...