Description: Advances in HPC, GPU computing, and AI have transformed the computing landscape of multi-physics simulation. CFD simulations can now be performed at unprecedented levels of fidelity and for long run times with state-of-the-art solvers. This seminar emphasizes how the same advances — specifically related to GPU-optimal algorithm design — can also be used to enable real-time monitoring and prediction of high-fidelity flowfields through the development of lightweight models inspired by the fusion of machine learning algorithms, operational sensor data, and exascale simulation data. As such, the focus here is twofold: (1) to outline how physics can be used within AI to enable GPU-optimal multi-physics flow solvers, and (2) to present a framework for coupling exascale simulations with real-world data to enable real-time flowfield estimation. Emphasis is placed on the role of high-fidelity simulations, hardware scalability, and model interpretability.

Join on your computer or mobile app

Click here to join the meeting

Or call in (audio only)

+1 630-556-7958,,479486641#   United States, Big Rock

Phone Conference ID: 479 486 641#