Best of 2016 - GPU software


Let's have a look what happened in GPU industry in 2016. Most of these tools make our jobs just a little easier — or even possible. Others are examples of technology that have migrated to animation and visual effects from other industries.

Let’s have a quick look, in no particular order of importance:

● EXR-IO — Finally, we have the ability to manipulate the AOV channels in CG renders, to quickly and effectively control the look of the render … in Photoshop.



 FurryBall RT 1.4 — Fast GPU rendering and new collaboration between apps - Maya / Max / Cinema. Easy and fast.


● Flowbox Roto — The is all about using RotoSTROKE to snap rotolines to shapes by drawing.


● Tilt Brush — This is painting in 3D space. No, really, in 3D space. With a VR headset and the Tiltbrush, it becomes sculptural painting.


● Oculus Medium — Like ZBrush, but for virtual reality!


● Cryptomatte — This open-source ID matte generator embeds data into the render without explicitly designating object mattes. A free Nuke plugin decodes the data and it’s being used in alShaders (see below).

cryptomatte bunny copy

● AlShaders — Developed by Weta’s Anders Langland (hence “al”), this is an open-source set of production-proven shaders for Arnold. But don’t worry, Vlado at Chaos Group has adopted it for V-ray — with an emphasis on the skin technology.


● Ziva VFX and Ziva Real-Time — The physics-based dynamic muscle system offers tissue volume conservation.

ziva logo

● Universal Scene Description — Pixar’s open-source software provides a consistent format to assemble assets — and then nondestructively edit them in various programs.


● Vulkan — A multiplatform API for making pretty 3D pictures on GPUs as fast as possible.


● PyCharm — Python IDE to help manage, control quality, and version track your Python code.

pycharm copy

● V-Ray for Nuke — The goodness of V-Ray, but accessible from within your comp.

vray nuke

● Wrap 3 — Node-based 3D mesh processing wrapping same asset topology around different shapes. Different characters with the same topography, for instance, which makes texturing, rigging, and such much more efficient.


● Neural Transfer Style — Making Blade Runner look like a Van Gogh!


● Project Voco — An impressive and terrifying technology that can change dialog in audio files by learning the speech patterns of the speaker.


● GPU Accelerated Deep Learning — Having Skynet solve big-data problems like image recognition, natural language processing, engineering efficiency — all pushed to multiple banks of CUDA cores.


● Multiverse — Open-source 3D scene assembly and management to visualize large, complex scenes using Alembic and USD assets.