Geometric deep learning in Simcenter STAR-CCM+ | From engineer to engineer

Описание: Simcenter STAR-CCM+ 2602 introduces Geometric Deep Learning (GDL) – a breakthrough AI capability that transforms how you perform CFD simulations. Join Rodolfo Piccioli - Principal Engineer, Cédric Tachot - Product Manager and Toby Emm - Application Engineer, as they walk you through one of the most exciting new capabilities in Simcenter.

Design iterations used to be a waiting game: sketch an idea, run CFD, then wait hours—or days—to see if you were right. But creativity doesn’t thrive on waiting.

With Simcenter STAR-CCM+ 2602, GDL capabilities are integrated directly into your CFD workflow. Your existing datasets can train smart, predictive models that can estimate performance in minutes.

Suddenly, exploring bold shapes, subtle surface tweaks, or entirely new concepts becomes fast, fluid, and fun. “What if?” no longer means a weekend of compute time. It means trying it now.

In this video, Rodolfo, Cedric and Toby discuss:
• Build training datasets within Design Manager using existing simulations
• Train AI models with as few as 40 designs using transformer neural networks
• Validate predictions with built-in accuracy metrics
• Run optimization studies exploring thousands of designs in hours
• Leverage GPU acceleration for 10x faster performance

Find out how we continue to provide you with Computational Fluid Dynamics (CFD) capabilities to go faster while modeling the complexity. Learn how to leverage exciting new features to explore engineering possibilities in an integrated environment and turn complexity into a competitive advantage.

#SimcenterCFD #SimcenterAI

📃 Blog article: The art of the automobile: why cars look the way they do (and how we make them better, faster) – Geometric Deep Learning for CFD
https://sie.ag/2LUam2

▶️ About Simcenter:
Engineering departments today must develop smart products that integrate mechanical functions with electronics and controls, utilize new materials and manufacturing methods and deliver new designs within ever shorter design cycles. This requires current engineering practices for product performance verification to evolve into a Digital Twin approach, which enables to follow a more predictive process for systems driven product development.

Simcenter software uniquely combines system simulation, 3D CAE and test to help you predict performance across all critical attributes earlier and throughout the entire product lifecycle. By combining physics-based simulations with insights gained from data analytics, Simcenter helps you optimize design and deliver innovations faster and with greater confidence.