Genesis 9 delayed until at least 2020. Will you be buying it?

1235730

Comments

  • TaozTaoz Posts: 9,733
    edited November 2018
    prixat said:

    They're actually going to skip Genesis 9 and go to G10 like Microsoft did with Windows.

    Yes, Victoria 9 will be based on G10.

    I've been working with G10 for some time, I'll post a preview, though I really shouldn't.

    Everything is built on subdivision and G10 has even fewer polygons.

    (Edited to remove texture)

     

    Hm, so constant bugfix updates of G10 that will force DS to restart in the middle of a render?

    Post edited by Taoz on
  • outrider42outrider42 Posts: 3,679

    How about this wacky thought: Genesis 8.5, completely 100% back compatible with G8. You don't even need to autofit, the clothes will just work. And 8.5 has enough small improvements to get people on board. G8 has done very well for Daz, and a lot of that is due to how well most G3 stuff transfers over. Of course they can call it Genesis 9 if they want to.

    The only way I would accept a new Genesis 9 that lacks really strong back compat is if G9 is a true generational leap that we have never seen before. I'm talking proper skin stretching, real muscle flexing, a true body hair system built in, ect.

    Obviously Daz makes it money off people buying content. But rehashing a new model every two years will eventually catch up to them. Think about new users just discovering Daz Studio. There are so many models available, and the naming scheme is super confusing...Hey, doesn't Victoria 4 come after Genesis 3??? Monique 7 for Genesis 3? What about Monique 1-6? Where's Genesis 5-7??? And that is before they open the software and find all these different things inside Daz to use. It can be quite overwhelming. I felt quite overwhelmed when I first started some 2 years ago. I very nearly quit. It is in Daz's own best interest to keep these things as simple as possible.

  • nonesuch00nonesuch00 Posts: 17,929

    Hmmm, Outrider42 has a point really...maybe DAZ 3D should drop the Genesis concept and start a new line with an easy moniker to grasp like 'DAZ 3D 2019' that is inclusive of everything released in 2019 that meets whatever the Genesis 9 specs / DAZ Studio 5.0 specs would require.

  • kyoto kidkyoto kid Posts: 40,581

    ...@ Outrider42.

    I agree.

  • nicsttnicstt Posts: 11,714

    Please, please: NO Studio 5!

  • DAZ pro suite encompassesing Carrara, Hexagon and Bryce all fully integrated and line of figures with boned muscles, skin layers and animated displacement textures for expressions 

  • TaozTaoz Posts: 9,733
    nicstt said:

    Please, please: NO Studio 5!

    So you want them to stop when they hit version 4.99.9.999?

  • Taoz said:
    nicstt said:

    Please, please: NO Studio 5!

    So you want them to stop when they hit version 4.99.9.999?

    Nothing saying they can't go to 4.99999999.999999999.9999999 or more.

  • Kitsumo said:
    Saldaz said:
    PedroC said:

    I think the problem it's not Genesis 9 in 2019, the problem it's Genesis 10 only two years later. Why to want G9 in 2019 if you can have G10 in 2021?.

    But why would go for G10 in 2021 when you can get G11 in 2023?

    :)

    I think I'll wait for Genesis 15 in 2031 myself. wink

    I don't like the system requirements: A quantum processor with 6 cores, 12 cores, or any combination of the two simultaneously. "My render started off ok, but I changed the outcome by observing it."

    No, that only happens if you're rendering a cat inside a box.  ;)

  • Subtropic PixelSubtropic Pixel Posts: 2,378
    edited November 2018
    Taoz said:
    nicstt said:

    Please, please: NO Studio 5!

    So you want them to stop when they hit version 4.99.9.999?

    Nothing saying they can't go to 4.99999999.999999999.9999999 or more.

    That's what they started doing in Star Trek.  They made Warp 10 "Infinite Velocity", and promptly decided not to use it.  So we got silly stuff like "Warp 9.5" and "Warp 9.975". 

    I say bring on DAZ Studio 5 and Genesis 9, although I don't think I have any Gen 8 characters at this time; 'cept maybe the basic thingie, and that's about 85% due to the unexciting 3D printing licensing costs. 

    Gen 9 would help lower prices on G3 and G8 bundles, which, along with my PC+ discount and discounted gift cards, would help make 3D licensing costs/terms somewhat easier to budget for.  That's the thing that might actually encourage me to open my wallet.

    And of course, if G9 has killer features, then I might be willing to bite there too.  But it's gotta be amazing, because I'd rather spend hundreds of future dollars on older models in a few years than thousands of 2018 dollars on those same models now, just because they're new and shiny. 

     

     

    Post edited by Subtropic Pixel on
  • It would have to have a major improvement before I would spend the money on a "9" character.....   

  • To be honest...They can crank out as many characters as they like... If they get so complex that it kills my 12gig GPU card, and I can't render more than one character with any props...What's the point?

    They have to come up with a better rendering method, so that large scenes, with lots of people can be rendered, without you having to be a software engineer. with your special script to remove all the complex textures in distant items.

    I get more done with V4 and M4 for this very reason.

    Newer isn't always better.

    JD

  • Oso3DOso3D Posts: 14,889

    I had my doubts about Genesis 2, then grew to love it.

    I had my doubts about Genesis 3, then grew to love it.

    I had my doubts about Genesis 8, then grew to love it.

    So now I'm shrugging and prepared to love Genesis 9.

     

  • TaozTaoz Posts: 9,733
    Taoz said:
    nicstt said:

    Please, please: NO Studio 5!

    So you want them to stop when they hit version 4.99.9.999?

    Nothing saying they can't go to 4.99999999.999999999.9999999 or more.

    I was only joking. Saying "no studio 5" without explaining why doesn't make a lot of sense.

  • Philippi_ChildPhilippi_Child Posts: 648
    edited November 2018

    Oh my, I've barely gotten started with Genesis 8. Over the past 18+ years I've done precious little actual art. I won't rush into Genesis 9.

    Me as well I'm just now getting used to G8 figures and DS in general.

     

    Post edited by Philippi_Child on
  • TaozTaoz Posts: 9,733
    Kitsumo said:
    Saldaz said:
    PedroC said:

    I think the problem it's not Genesis 9 in 2019, the problem it's Genesis 10 only two years later. Why to want G9 in 2019 if you can have G10 in 2021?.

    But why would go for G10 in 2021 when you can get G11 in 2023?

    :)

    I think I'll wait for Genesis 15 in 2031 myself. wink

    I don't like the system requirements: A quantum processor with 6 cores, 12 cores, or any combination of the two simultaneously. "My render started off ok, but I changed the outcome by observing it."

    No, that only happens if you're rendering a cat inside a box.  ;)

    Are you certain? cheeky

  • outrider42outrider42 Posts: 3,679

    To be honest...They can crank out as many characters as they like... If they get so complex that it kills my 12gig GPU card, and I can't render more than one character with any props...What's the point?

    They have to come up with a better rendering method, so that large scenes, with lots of people can be rendered, without you having to be a software engineer. with your special script to remove all the complex textures in distant items.

    I get more done with V4 and M4 for this very reason.

    Newer isn't always better.

    JD

    That's up to Iray and Nvidia. Octane is doing out of core rendering and eventually Iray will need to keep up.

    Some of the most highly regarded models out now are actually using pretty small textures. I think it was bluejaunte's models, many textures are like 1mb or so, with the exception being the 4 reflection maps. I was very surprised to see how small the actual skin textures are. So overall the impact on VRAM is not as great as some other models. I make this point because more realism does not necessarily mean a massive file size and super dense textures. Its pretty easy to compress textures on your own, and there are a few products in the store that automate optimizations for you.

    Most modern PCs handle high poly counts pretty well, though Daz Studio itself really needs to be updated to use multiple cores. That would make a big difference in Studio's usability. It is something that needs to be done as people make bigger and more ambitious scenes and products. Some items sold in the store will crush most user's PCs.

    Tech always changes, when Iray first came to Daz the biggest gaming card was a 980ti with 6gb. That was only a few short years ago. The 1080ti nearly doubled that. The 2080ti has the same 11gb, but it is capable of pooling its VRAM with Nvlink. (This has been proven, it is no longer speculation.)
  • TaozTaoz Posts: 9,733
     
    Tech always changes, when Iray first came to Daz the biggest gaming card was a 980ti with 6gb. That was only a few short years ago. The 1080ti nearly doubled that. The 2080ti has the same 11gb, but it is capable of pooling its VRAM with Nvlink. (This has been proven, it is no longer speculation.)

    So with 2 x 2080ti you can get 22 GB VRAM? How many can you combine?

  • RAMWolffRAMWolff Posts: 10,146
    Oso3D said:

    I had my doubts about Genesis 2, then grew to love it.

    I had my doubts about Genesis 3, then grew to love it.

    I had my doubts about Genesis 8, then grew to love it.

    So now I'm shrugging and prepared to love Genesis 9.

     

    yes

  • RAMWolffRAMWolff Posts: 10,146

    DAZ pro suite encompassesing Carrara, Hexagon and Bryce all fully integrated and line of figures with boned muscles, skin layers and animated displacement textures for expressions 

    Like this idea allot!  Add in sculpting brushes similar to ZBrush and I would totally be on board with this!  

  • Oso3DOso3D Posts: 14,889

    I also have a related view on Doctor Who...

     

  • Taoz said:
     
    Tech always changes, when Iray first came to Daz the biggest gaming card was a 980ti with 6gb. That was only a few short years ago. The 1080ti nearly doubled that. The 2080ti has the same 11gb, but it is capable of pooling its VRAM with Nvlink. (This has been proven, it is no longer speculation.)

    So with 2 x 2080ti you can get 22 GB VRAM? How many can you combine?

    As far as I'm aware there are only 2 card NVLinks. If you really need more than 21 Gb to render with, there's apparently some VRAM consumed as overhead, you might want to look into Quadros.

  • Every time a new generaton comes out im usually in the forums saying that we didnt need another one because the current generaton worked great. So of course I held off on getting into G8 for as long as I could. Then Daz did what Daz usually does and made me an offer for Vicky 8 at a rediculously low price. I wasnt super impressed to start but when all the great characters started coming out (Daz and PA), I was hooked.

    So now, I can say without a doubt, I am super excited to see Genesis 9! I think daz goes a little crazy with the number of base models they produce these days, but I cant wait to see what the PA's do with G9.

  • I have a lot of content. I'm increasinglly picky. Anything I buy, whatever generation it's from, needs to give me something I don't already have. I see so many characters and outfits I'm sure I've seen before because they're so much like things that are already out there.

  • outrider42outrider42 Posts: 3,679
    Taoz said:
     
    Tech always changes, when Iray first came to Daz the biggest gaming card was a 980ti with 6gb. That was only a few short years ago. The 1080ti nearly doubled that. The 2080ti has the same 11gb, but it is capable of pooling its VRAM with Nvlink. (This has been proven, it is no longer speculation.)

    So with 2 x 2080ti you can get 22 GB VRAM? How many can you combine?

     

    Not the full 22 (then again you don't get a full 11 with a single card either). The reason is some data must be duplicated for performance. The problem is, no app currently reports exactly how much VRAM is pooled, so right now nobody really knows exactly how much you get in Nvlink mode. But vray was able get a scene to run on Nvlink 2080ti's that would not run a single card due to its size. Running in Nvlink is also a bit slower than running the two cards without, but the penalty is not much. My guess (and I stress this is a guess) is that you might get 16-18GB when using Nvlink. You can only link 2 2080ti's or 2080's.

    Here is a link with details on what they did. They also have pics of the scenes they created for testing.

    https://www.chaosgroup.com/blog/profiling-the-nvidia-rtx-cards

  • Subtropic PixelSubtropic Pixel Posts: 2,378
    edited November 2018
    Taoz said:
     
    Tech always changes, when Iray first came to Daz the biggest gaming card was a 980ti with 6gb. That was only a few short years ago. The 1080ti nearly doubled that. The 2080ti has the same 11gb, but it is capable of pooling its VRAM with Nvlink. (This has been proven, it is no longer speculation.)

    So with 2 x 2080ti you can get 22 GB VRAM? How many can you combine?

     

    Not the full 22 (then again you don't get a full 11 with a single card either). The reason is some data must be duplicated for performance. The problem is, no app currently reports exactly how much VRAM is pooled, so right now nobody really knows exactly how much you get in Nvlink mode. But vray was able get a scene to run on Nvlink 2080ti's that would not run a single card due to its size. Running in Nvlink is also a bit slower than running the two cards without, but the penalty is not much. My guess (and I stress this is a guess) is that you might get 16-18GB when using Nvlink. You can only link 2 2080ti's or 2080's.

    Here is a link with details on what they did. They also have pics of the scenes they created for testing.

    https://www.chaosgroup.com/blog/profiling-the-nvidia-rtx-cards

    So this is a cost/benefit evaluation, and nothing more.  I think it would be rare to need more than the 11 GB of one card, but to help with the kitchen table math, I think it would good to have some form of understanding of "how much" would require NVLINK. 

    How much is how big?  11 GB is a lot.  What is the overhead for one card?  Is it a few meg or a few gig?  What is the overhead for two cards running in NVLINK?  Just how big of a scene is that?  Can I leave the connector in place and selectively engage or disengage SLI to address my current project?  The last thing I want to do is to have to crack open the case and attach/detach the cable each time.

    But if I could change it in software, that's doable.  I'd get 2 cards and mostly run them separately, only combining them when I had to render a big scene.

    Post edited by Subtropic Pixel on
  • Llynara said:

    Yep. In it for the long haul. I have every generation available. I use them all, and I enjoy them all. I can't wait to see what 9 looks like. Lovin' 8!

    Me too

  • outrider42outrider42 Posts: 3,679
    Taoz said:
     
    Tech always changes, when Iray first came to Daz the biggest gaming card was a 980ti with 6gb. That was only a few short years ago. The 1080ti nearly doubled that. The 2080ti has the same 11gb, but it is capable of pooling its VRAM with Nvlink. (This has been proven, it is no longer speculation.)

    So with 2 x 2080ti you can get 22 GB VRAM? How many can you combine?

     

    Not the full 22 (then again you don't get a full 11 with a single card either). The reason is some data must be duplicated for performance. The problem is, no app currently reports exactly how much VRAM is pooled, so right now nobody really knows exactly how much you get in Nvlink mode. But vray was able get a scene to run on Nvlink 2080ti's that would not run a single card due to its size. Running in Nvlink is also a bit slower than running the two cards without, but the penalty is not much. My guess (and I stress this is a guess) is that you might get 16-18GB when using Nvlink. You can only link 2 2080ti's or 2080's.

    Here is a link with details on what they did. They also have pics of the scenes they created for testing.

    https://www.chaosgroup.com/blog/profiling-the-nvidia-rtx-cards

    So this is a cost/benefit evaluation, and nothing more.  I think it would be rare to need more than the 11 GB of one card, but to help with the kitchen table math, I think it would good to have some form of understanding of "how much" would require NVLINK. 

    How much is how big?  11 GB is a lot.  What is the overhead for one card?  Is it a few meg or a few gig?  What is the overhead for two cards running in NVLINK?  Just how big of a scene is that?  Can I leave the connector in place and selectively engage or disengage SLI to address my current project?  The last thing I want to do is to have to crack open the case and attach/detach the cable each time.

    But if I could change it in software, that's doable.  I'd get 2 cards and mostly run them separately, only combining them when I had to render a big scene.

    Nvlink can be disabled with a click. 11GB being a lot depends on who you ask. There are plenty of users who want more than 11 because they want to use more than 11. There has been considerable increase in how much data things use in Daz. Many G8 characters can take over half a gig by themselves, and that's not counting clothes. Some environment props will use a ton of data just loading them up. And some people want to render very large images, which greatly increases the amount of VRAM needed.

    The information on Nvlink is still scarce and people are still learning as they go. Nvidia has not provided many details themselves about pooling VRAM, so everything we know is because of people testing it out. Eventually this will come to light. If you have questions, I would direct them to chaosgroup as they are doing a lot of this work.

    But there is always a cost benefit relationship regardless of what you buy. Two 2080ti's with Nvlink are going to be well over $2700. You can buy a Quadro 5000 for $2300 which comes with 16GB on one card. And Quadro can use TCC mode, which means you can use the entire 16GB bank and the card will run faster than normal because this cuts the card off from Windows. Gaming cards cannot do TCC mode, and like I said, there is that small speed penalty running full Nvlink. Even so, I would assume that two 2080ti's will render faster than a Quadro 5000. The 5000 is basically the Quadro version of the non ti 2080. But as you can see, there are trade offs.

  • frankrblowfrankrblow Posts: 2,052

    If you ever use this https://www.daz3d.com/alter-rock-for-daz-studio# you will find that 11GB of VRAM is barely adequate; add in some extra props, lights and characters and 11GB won't be enough.

  • FPFP Posts: 113
    100% compatibility with clothing from G2 up but would like V4 up No triangles no shoe issues. At least Morph transfers from G3 and up + Pose transfer G3 up with no issues. Click compatible to transfer entire G8 characters over and possibly G3. Then I would consider it as being worth a purchase.
Sign In or Register to comment.