mental ray 3.12 Beta

nVidia is now taking applications to test mental ray in the 3.12 Standalone Beta Program.

This would give you access to a time-limited and fully functional version of mental ray 3.12 with new features including Global Illumination on the GPU.

Take a look on the nVidia ARC website for more details.

About David

I am a VFX artist that specializes in Lighting and Rendering. I spend a fair amount of my time supplying clients with artistic solutions as well as technology solutions. With a background in fine art and technical animation training, I strive to bridge the divide between the artist and technologist.

Posted on December 11, 2013, in Uncategorized. Bookmark the permalink. 10 Comments.

  1. Sooooo, anyone tested it yet? Is it fast? Faster than Redshift3d? 🙂

  2. Hi David, just wanted to ask if there is any news about mental ray on openEXR 2.0 or any deep output. thanks in advance, great blog.

    • We’re not allowed to talk about closed Betas. Deep compositing seems to be the buzzword for EXR 2.0 but I’m not sure anyone knows how much data that actually outputs. Single frames can be in the gigabytes.

      • Thank you for the quick reply. yes i know everyone is hyped up about it and i had to test it with MtoA RC3 but the output data seemed to be reasonable for a freelancer like me working with relatively small scenes.

        The reason i’m looking forward to it is to be able to output maya fluid deep data to nuke because i love the sharpness and density of mental ray fluid renderings.

      • The bigger question will be if Maya supports it rather than mental ray alone. It has to be integrated workflow.

      • sure. i just feel it is gonna be a good year for maya artists with the upcoming updates like opensubdiv, openVDB, exr 2.0, ptex, xgen etc.

  3. So, instead of having a new modern, stable cpu-based GI solution we’ll get the same old FG on the gpu? Wow, it’s really sad seeing Mental Ray ending its days being just a leverage to force users to buy gadgets. It’s dying, if not dead already. Bravo Nvidia. Why not be honest and kill Mental Ray right away instead of disappointing your user base with promises of “something great” that never arrive? It was exciting at first when i heard nvidia to acquaire MI, but of course their goal is always selling hardware, not improving software technnology that make no use of their gpu’s. When all MI products will mandatorily make use of gpu, what will happen with our already purchased non-gpu-updatable render farms which relies all render power on cpus? there a lot of small companies which made an effort buying 5-6 slim-node render farms. I supose we’ll have to eat them.
    Well, speaking about the subject of he post, will it at least be capable of compute brute force FG on the gpu? Is AO on the gpu able to compute its AO thing with displacements? all this w/o buying a $1000 gpu with tons of vram, of course.

    I’m really willing to hear you about my concerns. I’m sorry if this post sounds rude, but this is somewhat the kind of thing i take into account to change to another renderer.

    Thank you for your time in advance.

    • Hi Jorge.

      1. FG on the GPU was an unfortunate marketing mischaracterization. This is NOT Final Gathering on the GPU. It’s called GI on the GPU. It’s a new technique for indirect lighting that can leverage the GPU to render existing or new scenes. You will be seeing more of this explained in the future. The demonstrations at SIGGRAPH were impressive in regards to speed.

      2. GPU compute solutions aren’t really gadgets I don’t think. Both mental ray and Vray are exploring GPUs as a way to enhance rendering (Vray 3.0 will eventually introduce AO on the GPU, a few years behind mental ray)

      3. The AO on the GPU solution has a CPU fallback option that works in a similar way for render farms without a GPU solution. So I can’t say they are excluding non-GPU users. NVIDIA understands wide acceptance is built on many different users.

      4. Is GI on the GPU brute force? Maybe you should sign up for the Beta to begin testing the early implementations and see it evolve. Feedback is useful in early testing so that features aren’t missed and bugs are fixed sooner. But you can imagine sending millions of rays on a GPU is much faster than on a CPU. . .

      5. AO on the GPU does render displacements since mental ray hands the displaced geometry to the GPU. It also supports flags for casting and receiving AO more easily using Maya’s cast and receive shadow flags.

      6. You’d be surprised how little GPU RAM you can get away with. But in any event, you can use a GTX to compute these results instead of a Quadro or Tesla. In the future NVIDIA’s roadmap looks to increase the memory density on their GPUs (Volta series)

      7. Like Apple, NVIDIA is a hardware company. The hybridization of mental ray isn’t a surprise. NVIDIA sells integration technology that makes use of its hardware like CUDA, Optix, and mental ray. But I don’t think this means mental ray is aimed at replacing iray as a GPU-only solution.

      Anything else and you’ll either have to wait for some presentations or join the Beta and take a look at the features. 🙂

      • Ok, i apologize if my post sounded a bit harsh (well, it did, and i’m sorry for that). I really like MR for its simplicity an ease to setup, and of course its powerful materials, and i rely on it all my projects, but i cannot afford buying more dedicated hardware to support its new features. About the beta program, i’d be interested in joining and test this new features provided i have time to spend on it, which i haven’t, and cannot risk my deadlines testing new features on real projects.
        On the other hand, I’m joining a mental ray-based studio in the next weeks as hair r&d/grooming/shading/rendering artist. Maybe i can convince them to join the program and test some features like the new hair shader while we are in r&d stage, mainly because the current solutions are not satisfactory enough regarding time/quality.
        I think nvidia has a very clear roadmap about its renderers and all signs point to a convergence towards gpu only solutions, its their market and they bought MI to exploit that. They tried with gelato and failed, and MR is their second and definitive trial.
        Anyway, thank you for taking your time to answer even if my post was a bit harsh.


      • mental ray will undoubtedly look into hybridizing features into the GPU realm. And as GPUs get better and better (a guarantee) it’s in their best interest to do so. But they know their main customers do not have GPU farms.

        They also have ways to continue to make rendering on the GPU flexible, like the Material Definition Language, which also allows identical rendering on the CPU. 😉

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: