The simulation of physics events in the LHCb experiment uses the majority of the experiment’s distributed computing resources. Most of the computing time is spent on the Geant4-based detailed detector simulation, with approximately 50% of it dedicated to the calorimeter system. Optimizing the calorimeter simulation is therefore crucial for improving the overall computational efficiency. This article presents a solution implemented in the LHCb simulation software to accelerate the calorimeter simulation. During Geant4 transport, the simulation of particles entering the calorimeter is halted and hits are generated using libraries of pre-simulated energy deposits, which undergo a series of transformations to improve simulation accuracy without increasing the library size. This technique reduces the calorimeter computation time to a negligible level without significant loss of accuracy. A comparison between the outputs of the fast and detailed simulations is presented.
Kholodenko et al. (Wed,) studied this question.