Disney, Google Deepmind And NVIDIA Have Developed Motion Of Physical Processes Of Newton

under the protection of the organization Linux Foundation Transferred The simulation engine of physical processes Newton , jointly developed by NVIDIA, Disney Research and Google Deepmind. It is noted that the transfer of Newton to an independent neutral platform that is not controlled by individual companies will allow us to attract new participants. The Newton code is written in Python and is distributed under the license Apache 2.0.

the engine is created for a quick and scalid simulation of processes and conducting research in the field of robotics. Among other things, Newton allows you to simulate the complex behavior of robots with a large number of interactions, such as walking in snow or manipulation with fragile objects. Among the key advantages of the project, the GPU is noted to accelerate operations and a quick scalable simulation, as well as an expanded architecture that allows you to connect and replace the components.

to accelerate the modeling of processes and spatial calculations using the GPU, the framework is involved href=”https://developer.nvidia.com/ward-python”> nvidia warp , and for the tolerated representation of the hierarchically related data that forms the graphic scene, the Openusd platform.

You can use different baptisms processes (Solver). The Mujoco engine (Multi-Joint Dynamics with Contact) is involved as the main backend, and backens that implement the methods euller , featherstone , implicitmpm, SemiImplicit , style3d , vbd and and xpbd . It is possible to download models in URDF, MJCF and USD formats. href=”https://ru.wikipedia.org/wiki/%D0%93%D1%80%D0%B0%D0%B4%D0%B8%D0%B5%D0%BD%D1%82%D0%BD%D1%8B%D0%B5_%D0%BC%D0%B5%D1%82%D0%B4%D1%8B”> Gradent Methods and calculate the derivatives from the output parameters of the simulation in relation to the input parameters to optimize the simulation parameters or use in machine learning. For example, you can use the engine to optimize processes and teaching robots with accurate manipulations in a simulated environment.

/Reports, release notes, official announcements.