CERN has a long-standing tradition of developing cutting-edge software tools for various beam physics applications including lattice and optics design, particle tracking, collimation studies, and simulations of collective effects. In recent years however new challenges have emerged which are difficult to tackle with established tools, especially when addressing the design of next-generation colliders. The fragmented software landscape has increasingly hindered users requiring new features and studying the interplay of multiple physical phenomena. At the same time advancements in high-performance computing—particularly GPU-accelerated technology—can enable orders-of-magnitude faster beam dynamics studies, yet retrofitting legacy tools with GPU functionality is often infeasible.

To address these challenges the Xsuite project was launched in 2021. Designed as a Python-native simulation framework, Xsuite supports heterogeneous studies by seamlessly incorporating features such as particle-matter interactions, collective effects, GPU acceleration and multithreading. As part of the Python ecosystem, Xsuite easily interfaces with other Python-based tools, enabling its users to leverage the vast array of scientific libraries available within the Python community. 

By now Xsuite has become an established tool in beam dynamics simulation having gradually replaced several legacy programs. So far, in CERN's operational tasks MAD-X remains the standard for lattice modeling survey and optics. However initial tests of Xsuite in these contexts have shown promising results paving the way for its broader adoption in operational workflows.

This talk explores the rationale behind the transition, outlines the strategy for its implementation, and highlights its benefits by showcasing Xsuite's capabilities and interface. The discussion following the presentation will be an occasion to engage with the user community and collect input on how to move forward.