In Situ Analysis with Catalyst: Getting Started

As simulations across scientific disciplines grow in complexity, their demand for computational resources increases dramatically. Traditional post-processing pipelines struggle to keep up and often become inefficient. To address this, there is a growing need to incorporate state-of-the-art analysis and visualization tools directly into simulation platforms.

Catalyst is a flexible and high-performance platform that integrates seamlessly into simulation workflows. It enables in situ analysis and visualization, eliminating the need to store and process massive datasets after simulations complete. Through Catalyst, users gain full access to ParaView’s powerful capabilities directly within their simulations. It also supports data movement technologies like ADIOS, making it adaptable to a variety of HPC environments.

For over a decade, Catalyst has led the shift toward in situ simulation data processing. This evolution has become even more critical as the I/O bottleneck continues to challenge modern HPC workflows. At Kitware, we are expanding what Catalyst can do—moving beyond visualization to support advanced data handling, interactive analysis, and AI-powered coprocessing.

In this tutorial, we will demonstrate how to integrate Catalyst into a simulation code from the ground up. We will showcase how to link Catalyst to your simulation code and explore the fundamentals of describing simulation data in memory for Catalyst.

During this webinar, Kitware’s experts will:

1. Demonstrate how to link Catalyst to a simulation code
2. Show how to describe simulation data in memory with Conduit
3. Explore advanced features for controlling Catalyst at runtime