The production of a sufficiently large number of simulated Monte Carlo events is anticipated to be one of the most significant bottlenecks for many future high-energy physics (HEP) experiments. The simulation of the calorimeter response, in particular, represents a major computational challenge. While substantial efforts have been made by the HEP community to develop machine-learning based fast simulation models, integrating these into realistic experimental setups remains a significant hurdle. Building on the fast simulation tools developed by the ATLAS Collaboration at the LHC, this paper presents recent efforts to create a fully experimentindependent library for fast calorimeter simulation. The library aims to provide a universal interface for both the lateral and longitudinal parameterization of calorimeter showers, as well as for machine-learning based approaches to shower generation.
Joshua Falco Beirer (Tue,) studied this question.