Event Date/Time: Thursday, September 10 from 10-11 am PT

Overview:

Physics AI has moved quickly from research promise to industrial adoption, but many model architectures still struggle when they leave idealized lab conditions and enter real engineering workflows.

In this webinar, we will discuss Luminary-SMART, Luminary’s new architecture for Large Physics Models, and explain how it is designed to solve one of Physics AI’s hardest problems: building models that learn the underlying physics rather than the simulation mesh.

The session will cover how Luminary-SMART combines mesh independence with scalable model capacity, how it builds on lessons from other model architectures, and what its benchmark performance means for engineering teams deploying Physics AI in production.

What You'll Learn:

• Why architecture matters for Large Physics Models: how model architecture design affects accuracy, generalization, inference speed, and production usefulness
• The mesh-dependence problem: why models can look strong in training but fail when deployed on raw geometry or unfamiliar tessellations
• How Luminary-SMART: uses mesh independence to help the model focus on learning the underlying physics
• How it compares to prior architectures: qualities from other leading model architectures that motivated the development of Luminary-SMART, and how we took that forward
• Benchmark results that matter: performance across SHIFT datasets, quantities of interest, mesh-dependence tests, inference speed, and training time

Who Should Attend:

• AI/ML practitioners working on scientific machine learning
• Engineering and simulation teams evaluating Physics AI for production design, analysis, and validation workflows
• CFD, FEA, structures, loads, thermal, and multidisciplinary simulation engineers interested in applying AI/ML for engineering