NVIDIA has released Newton: Open-Source Physics Engine for Robotics Simulation
Alisa Davidson
Post: March 21, 2025 11:14 AM Update: March 21, 2025 11:14 AM
Edit and fact confirmation: March 21, 2025 11:14 am
simply
NVIDIA has unveiled the open source and expansion physics engine developed in collaboration with Google Deepmind and Disney Research to improve robot learning and development.
NVIDIA, a technology company, has unveiled the open source and expansion physics engine, which was developed in collaboration with Google Deepmind and Disney Research to aim to strengthen robot learning and development.
Based on NVIDIA WARP, where robots can learn complex tasks with improved accuracy, Newton will work with learning frameworks such as NVIDIA ISAAC LAB, an open source platform for Mujoco Playground and Unified Robot Learning. Designed.
Using the Physical AI model, the robot can autonomously understand, analyze, why and participate in the surrounding environment. The development of robotics depends greatly on accelerated computing and simulations to develop the next -generation robot system.
Physics is essential for robot simulation, and robots are essential because they form a foundation for creating an accurate virtual model that indicates the way robot acts and interacts in a real environment. Through these simulators, researchers and engineers can educate, design, test, and verify control algorithms and prototypes in a safe, efficient and cost -effective way.
Newton is designed to support the entire robot community so that robot engineers can freely use, distribute and contribute to R & D. Based on NVIDIA WARP, a Cuda-X Acceleration Library, this product provides developers with an efficient way to create a kernel-based program accessible to GPUs for simulation, AI, robotics and machine learning (ML). This framework provides a high performance function that executes physical -based simulations using parallel processing capacity of NVIDIA GPUs.
Newton’s noteworthy feature is compatibility with MUJOCO (Multi-Joint Dynamics with Contact), an open source physical engine used for robot research to model complex epidemiological and contact environments. This compatibility allows developers to reuse existing models and code to reduce the time and resources needed to adjust applications for various physical engines.
In addition, Google Deepmind introduced the Mujoco-Warp, an open source robot simulator accelerated by NVIDIA WARP, to achieve 70 times more speed for humanoid simulation and achieved 100 times speed for direct operation. Mujoco-Warp is integrated into NEWTON’s default physical engine, providing developers with improved performance and flexibility of robot engineering applications.
NEWTON’s more major features: differentiated physics, scalability and openusd integration
In addition, the ability to spread the gradient through simulation introduces new opportunities for robotics simulation and learning. Differentiated simulators can produce a forward mode result and calculate the reverse mode gradient of the simulation result, which can be used for reverse to optimize system parameters.
As the robotics field develops, the complexity and various scenarios that need to be simulated are also developed. Newton is designed to be highly adaptable to a robot that supports abundant multi -physical simulations that interact with various materials, including food, cloth and other deformable objects. This flexibility is activated by custom solver, integral and numerical methods. Newton also supports combining various types of solver as demonstrated in the integration of rigid body mechanics and material point method method (MPM) solver to simulate interaction with sand.
NEWTON also provides a variety of data models and configuration engines that consolidate the data needed to explain the robot and the environment using OpenusD frameworks. Custom solver and runtime can be specialized for specific robot functions and environments. In addition, with Disney Research, Google Deepmind, Intrinsic and NVIDIA, Newton helps to define Openusd asset structure for robotics. This structure aims to create an unified data pipeline that provides a common language to all data sources of robotics by adopting a model case in Openusd to standardize the robot workflow.
disclaimer
The trust project guidelines are not intended and should not be interpreted as advice in law, tax, investment, finance or other forms. If you have any doubt, it is important to invest in what you can lose and seek independent financial advice. For more information, please refer to the Terms and Conditions and the Help and Support Pages provided by the publisher or advertiser. Metaversepost is doing its best to accurately and unbiased reports, but market conditions can be changed without notice.
About the author
Alisa, a dedicated reporter for MPOST, specializes in the vast areas of Cryptocurrency, Zero-ehnowedge Proofs, Investments and Web3. She provides a comprehensive coverage that captures a new trend and a keen eye on technology, providing and involving readers in a digital financial environment that constantly evolves.
More
Alisa Davidson
Alisa, a dedicated reporter for MPOST, specializes in the vast areas of Cryptocurrency, Zero-ehnowedge Proofs, Investments and Web3. She provides a comprehensive coverage that captures a new trend and a keen eye on technology, providing and involving readers in a digital financial environment that constantly evolves.
More