I believe RL marketing thinks of DS as a competitor. I suspect the pending IC addition of iRay is an indication of that - even though iRay at present is incapable of realtime animation rendering, and even though still image has never been RL's core strength, or even a minor one. iRay is just another name to drop. It will take more than iRay to move in on DS's still image turf. Marketing is just doing it's job. In animation, DS is a major non-competitor. They have turned themselves into a clearing house for the 3D model cottage industry, and done an excellent job at it too. They seem content to rest on their laurels. If they hadn't picked up iRay (not so long ago) they wouldn't even do a good job with still images. Some competitor they are.

If RL actually does produce a usable and reliable tool to import and use DS clothing, they will kick major dirt on DS, since it is almost all that DS has. And you've said they will. I haven't seen RL say it though. 


"Less clicks good, more clicks bad."

4x would require....4x GPU!
Okay, so I would suggest at minimum trying a dual-GPU setup and rendering with that before you sell it. 
And, you don't even have to wait for iClone Iray, you can see it in action in Daz Iray, now.   
Assuming you install it properly in PCIE slot, and it is able to be powered by your PSU, and it is recognized in Control Panel Device Manager/Nvidia Control Panel, etc, the render settings area of Daz will give you the option to select it as an active GPU to be used for rendering.  You can deselect it too, if you want.  What many in the Octane world do, for example, is use the main display GPU as a non-rendering GPU, and use the other for rendering, while working.
Then, when it comes time to actually let the machine just render, you can also then activate the main one (by clicking it in the menu, as a rendering GPU)
However, I wouldn't use that Iray to benchmark in absolute terms your rendering speed, because it does not have AI denoising....yet

Anyway, this is so cool and amazing talking multi-GPU....with you!!
You are in for a treat....soon you will want 4x GPU.  And it is possible, depending on how many PCIE lanes, not physical connector slots, but 'lanes' (as in peripheral capacity) your mortherboard has.
You can even USB connect GPUs to run externally from your machine with PCIE to USB3 adapters, and then you have a cool-running mini render farm, all on 1 PC.  

This is older, but I used to run 6 Titan Z, and each Titan was Dual-Core, so you had 6x2=12, and then I had Titan X just as the main mobo GPU.  In all, I could use 13 at once in Windows 7.

This is Daz on Left, and the Octane Daz Plugin on right:



Here was a very early test runs I was doing with Indigo, when it first was testing its multi-GPU:



So, it is possible.  On my main PC I use 7 Titan X Pascals now, 6 external, + 1 main display card, and in Octane are almost just as fast as 12 Titan Z cores + 1 Titan X.  Somehow, they are way faster in Iray and Indigo than the 13, I think because there is more GPU resource needed to build a scene, so it doesn't pay to keep adding, like it would in Octane.   
For me, a lower # of higher powered cards is more functional for the majority of rendering environments because it is not too demanding on the PC and power
Oh...You can bitcoin mine too, KT.  You will hash more with 2 GPU than one!  

Anyway, this all begs the question....why not use just 1 GPU on PBR and call it a day?  
Know what, it is a great question, this is what you do in UE4, but then after using Octane and Iray, you can't help but see deficiencies. 
But in the end, it just depends what your rendering output required. 
Path-trace or No-trace.  I say try dual-GPU 1st, and compare it with 2 vs 1 and see if it adds value to you or not.

Right now I'm in that unique position (having the 1080 -- my Alienware PC came with the 1070 which I moved into my old PC, and moved my Titan into the new one).  I've actually just not gotten the time to eBay it (when I first bought the Titan that was my intention, but now months have passed).  I suppose waiting three or (more likely) four months to see what is what with iClone iRay is okay, even if the value of the 1080 will drop dramatically in July (when the new nVidia cards are released).

But (and I realize you and I can only speculate here) if iRay for iClone does come out as I suspect and is nowhere near animation ready then just doubling my speed won't help and I will have wasted both time and money (my new PC can support two GPUs, both in power and slots so nothing more would be needed at least, but I will have wasted time getting it set up that way).  And, let's face it, we don't even know if that iRay plugin WILL support two GPUs, even if it's likely.

Now, I realize I may be WAY off base with my estimates (although I still think it's telling that RL isn't even *trying* to show animation with iRay) and I also realize everyone has a different set of acceptable animation standards, but coming from the "Real Time" environment of iClone I don't think we can say that animation is a possibility until we get typical frame rates, of ALL scenes, of under 10 seconds (that still is pretty brutal -- about three hours for 30 seconds depending on fps).  Reducing THAT in half would be at least acceptable -- you could conceive of doing an animation of 10 minutes in 15 hours of render time.  But anything more?  Just won't be reasonable.

Plus the concerns you raise are very valid -- if iRay doesn't come *close* to transparency when it compares to the render inside of iClone, but rather requires a ton of tweaking of lighting and materials, then it just isn't worth it (and what about Popcorn FX?  Sigh, the more I talk about it, the more I talk myself out of it).

I think at this point I'll wait until I see the next video, supposedly at the end of June.  Maybe that will show me enough to make an intelligent decision (which, at this point, is most likely sell the damn card and move on with my life :>.

---

Alienware Aurora R16, Win 11, i9-149000KF, 3.20GHz CPU, 64GB RAM, RTX 4090 (24GB), Samsung 870 Pro 8TB, Gen3 MVNe M-2 SSD, 4TBx2, 39" Alienware Widescreen Monitor
Mike "ex-genius" Kelley

The young man brought his car to the old engine builder. He said, "I want you to build me an engine that will go really fast!" The old man answered, "Speed costs money, son. How fast do you want to go?" PCs are probably the best hobby I've ever had. I do love the hardware. Sometimes I can forget what my original goal was. When that happens, my wallet reminds me.



"Less clicks good, more clicks bad."

My first "high end" computer cost nearly 6k, and this morning it's sitting on the bottom of my driveway waiting for the garbage collection.

Kid morphs are a pain in CC, but this is a GREAT tutorial that I used to create a standard child base character that I can manipulate however I want now. Pretty happy with the result

https://www.youtube.com/watch?v=TJRrHHyIvCc

The visual thing will be up for debate, that is for sure.  Sometimes PBR does an efficient job to the point where it just won't be needed to have ray-tracing. 
I think the vid shows this, the modeled guy looks like PBR rendering will be just fine, whereas the amazing beautiful vixen, she deserves ray-tracing.  Like, I want to be that light...reflecting on her body, making the sub-surface scatter....
It will be glorious.  But, it will be glorious for 15 seconds, and then I will want to tweak it further.  I think iC7 has a great array of visual tools, very UI friendly.  But even there it is missing a whole slew of PBR channels that Daz has (sheen, dual lobe spec, metallic flakes, etc).  UE4 has a whole bunch of shader modes, specifically for targetted textures.  In iC7, using Super tools is the workaround.  So okay, but fine, you can get a look you want end explore it further with the rest of the tools.   But with rendering, Octane's imager has Brightness, AND Exposure.  Contrast, AND Highlights...Hue, White Point, Bloom, Glare, LUTs, grain, masks, etc....

You are right -The PopcornFX part will either be a very, very pleasant surprise, or a big disappointment.

In the video the colors and lighting seem to match, and I'm like "this looks beautiful...".
Then I see that missing hand shadow, and I'm like "What is that?"....Then I pay more attention to the Tonemapper part, and I'm like, "No...Noooo....NOOOOO....NEED MORE!!!  NEEEEEED MOOOOORE!!!!!! 
Where's me Bloom...WHERE'S ME FRIGGIN' BLOOOOOM???!!!!!" 
Perhaps as this is the CC3 implemented version of Iray, and it has more limited menu options than will the iC7 one.  If so, I can see that being a design choice, fine. 
I just do not want to see what looks to be an extremely brilliant tool get down-throttled in the name of user ease. 
We are your artists.  Don't give us that 8 color Crayola box....Give us that big mega 64, 96, 152 (whatever) color one.     

Dupe post

CC3 Looks amazing - time I start saving up the shackles again

  

Pixar Animation Studios Open Subdivision Technology?

Autodesk dated review of How & What:



Demo subdivision options? Area specific on characters?  The ears, eyes, nose, mouth and hands ran through the Pixar 'PAST' thing which looked impressive. 

These features often get munched up in CC2 if pushed..

I get Quick Renders - it's why I IClone... However.. If CC3 designed characters that render PIXAR quality video in IC with some patience and hardware that would rock.  If it could do it in half the time of DS - Sign me up..

Any info or premonitions appreciated..


Virtual Media® / VTV® / The Virtual Entertainment Network...
http://VTVLive.com