2080 Ti (11GB ONLY!) - Is this going to be a worthwhile upgrade?

PlacidFalconPlacidFalcon Posts: 16
edited August 2018 in The Commons
I know it's going to be a guessing game for now, but I'm getting a queasy feeling about this generation. For the last 3 generations, our VRAM has increased linearly by a factor of 2x. It is my humble opinion that a large chunk of our gains were from that VRAM increase alone...
Post edited by PlacidFalcon on

Comments

  • MescalinoMescalino Posts: 436
    edited August 2018

    The bigger VRam will let you render larger scenes in iray on the GPU (Or having to do less optimizing) But from what i saw so far from the numbers it was a bit underwhelming.

    With any luck the 1060/1070/1080 will drop in price with the new cards making them far more interesting looking at the value per dollar/euro

    Personally i have doubts if its really worthwhile. But time will tell what they can do in terms of optimizing drivers etc. to boost its performance.

    Post edited by Mescalino on
  • Ghosty12Ghosty12 Posts: 2,080
    edited August 2018

    One you have to work out what you have now, two are the new cards worth the upgrade (baring in mind that you will want to wait for the new tech to mature a bit more, ie: next year) so for now just a case of waiting and seeing what the numbers are and also it all depends on how long it takes nvidia to update iRay and for Daz to implement it..

    Post edited by Ghosty12 on
  • nicsttnicstt Posts: 11,715

    If you mainly render, or it's the main reason you're considering the ti version, then from a 1080ti say, a tough call; I'd wait for reviews. From earlier cards, I'd say yes it would be a good upgrade, but worth it - personal reasons.

    Just remeber, 10 series was a long wait for driver support for Iray. Daz were great at getting it out of the door, Nvidia were not great - it took time.

    We were'nt really told anything quantifiable; it was all hyperbole and marketing; far more so than previous releases I understand; such a presentation worries me (It may turn out that my worries are groundless.), because of their style. Sure I get the new tech, and am looking forward to it - I love new tech. What concerns me is that all these claims of needing a new set of benchmarks whilst quite possible true, doesn't mean they can't be related to what has gone before. Why didn't that happen?

    There are various reasons, some of which I don't like.

    We're speculating at best, and won't know anything until the various sites get their hands on some kit to test; hopefully the kit won't be specially chosen to present exceptionally good figures, I don't believe that would happen, but can't help wondering.

  • MescalinoMescalino Posts: 436
    nicstt said:
     

    We're speculating at best, and won't know anything until the various sites get their hands on some kit to test; hopefully the kit won't be specially chosen to present exceptionally good figures, I don't believe that would happen, but can't help wondering.

    For this reason i only watch non sponsered reviews/sites/youtube channels instead of the numbers presented by Nvidia in hopes of getting an as much as possible objective review. Bt given the fact i bougt my 1070 6 months ago i will not be looking to upgrage.

  • ebergerlyebergerly Posts: 3,255
    edited August 2018
    Sounds like some folks here have pre-ordered cards, so hopefully they will post some benchmark numbers. I think 3 or 4 users so far have decided to boldly go where nobody has gone before, so we should have some good info. Based on prices, the 2080ti will have to render in better than half the render time of a 1080ti, which is asking a lot. I think historically the improvements have been more like 25-30% between generations. And with only 11GB, and no NVLink, I think theres good reason for some skepticism
    Post edited by ebergerly on
This discussion has been closed.