time: 11:00 AM – 12:00 PM Pacific Time
Period: 1 hour
High-resolution simulations in science and engineering are widely used in industrial, seismic, weather/climate, and life science applications. However, traditional simulations remain computationally expensive and impractical for real-time applications. It is discretionary, which means it does not easily absorb measured or synthetic data from different sources. Due to the rapid developments in artificial intelligence for science and engineering problems, machine learning has played an important complementary role in addressing critical gaps in traditional methods.
NVIDIA Modulus is a physics-based machine learning platform that contains many modern network and data structures, as well as PDE-led AI techniques to solve real-world science and engineering problems. Modulus has various performance features for single and multiple GPU/node systems, as well as connectivity to many NVIDIA toolkits and technologies. Examples and documentation are provided to ensure smooth learning for students while researchers can customize the framework through various APIs.
This webinar will introduce you to machine learning applications, different areas of science and engineering, as well as a deep dive into code implementation, training, solutions, and visualization aspects of the Physics-ML workflow.
By attending this webinar, you will learn about:
- Machine Learning Applications in Science and Engineering with Physics-ML Framework, NVIDIA Modulus.
- How can you extend / modify the parameter to carry out your own work.
- Lab architecture and functions, and performance improvements for data and physics systems.
- How the Lab framework integrates with other Nvidia toolkits and technologies: PySDF (for engineering), DALI™ (for data loading), Triton™ (for inference), Omniverse™ platform (for visualization).
Join us after the presentation for a live question and answer session with Jianjun Xu, Ph.D. , Sr. Solutions Architect, Amazon Web Services.