For several months, Google has taken machine learning to its G Suite to give a corporate level boost to the entertainment industry. In July 2018, the technology giant announced Cloud AutoML Vision in beta. It is a custom translation project and neural networks for the natural language that reached users of Google Cloud.
Since render farms exist, they have been used by the entertainment industry to create visual effects and animations. Performing this process in the cloud implies that the workload is executed without having to update the computation resources. Also, this technology allows companies to scale to achieve better rendering power as needed.
In 2018, Google opened a new Cloud branch in the city of Los Angeles to bring users an enhanced media and entertainment service. Now, Google Cloud and Sony Pictures Imageworks (SPI) have announced a joint project called OpenCue, a new rendering manager for visual effects and animation.
With this project, Google and Sony seek to create a better way to produce effects, where the cloud is involved as a protagonist. According to a statement from Google, it is an open source render administrator that can scale from thousands to millions of images in a ‘hybrid cloud’.
According to the announcement, OpenCue is the evolution of Imageworks Cue 3, the internal tailing system of Sony Pictures. The architecture of this technology supports a large number of machines with multiple concurrences and has labeling systems that give users the possibility to assign specific tasks to various specific machines.
The work done with OpenCue is processed by a central rendering farm, which frees the workstations from the creators of animations and visuals for new tasks. Also, hosts can be divided into many processes, each with its own memory and core specifications.
The open source project is now available through Github, with management guides and examples of future OpenCue projects.
Also published on Medium.