What’s New in In Situ Analysis and Visualization

As high-performance computing systems move past exascale, traditional post-processing workflows have become a limiting factor for simulation-driven applications. In situ analysis with ParaView Catalyst provides a scalable alternative—enabling teams to process and act on data as simulations run.

In this webinar, we’ll explore the latest advancements in Catalyst that support more efficient, scalable, and intelligent workflows, including:

• Asynchronous execution to decouple simulation and analysis for improved performance.
• GPU-resident data pipelines using zero-copy techniques to eliminate costly data movement.
• Real-time simulation feedback using catalyst_results() for dynamic steering.
• Integration with AI/ML pipelines for surrogate modeling and data-driven analysis.

Whether you are already using Catalyst or evaluating in situ approaches to reduce I/O bottlenecks and accelerate time to insight, this session will provide a clear view of how modern Catalyst workflows can be applied in practice.

You’ll also hear directly from the Catalyst development team about the motivation behind these advancements and what they enable for next-generation HPC and AI-integrated simulation environments.

A live Q&A session will follow the presentation.

Key Takeaways

• Understand when and how to use asynchronous execution to improve simulation performance.
• Identify opportunities to eliminate data-movement bottlenecks in GPU-resident workflows.
• Learn how to enable real-time insight and steering within active simulations.
• Evaluate how AI/ML integration can enhance simulation-driven workflows.

Who Should Attend

The capabilities covered in this webinar are designed for developers and teams working with large-scale simulations across domains, including CFD, national labs, and defense research.

You should attend if you are:

• Working with large-scale simulations (CFD, multiphysics, etc.)
• Experiencing I/O bottlenecks or long post-processing times
• Exploring in situ analysis and visualization approaches
• Looking to adopt new ParaView Catalyst performance and AI capabilities