GPU accelerated Longitudinal Tomography

Summarise with VIRAT
Authors: Bernardo Abreu Figueiredo, Simon Albright
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Longitudinal phase space tomography is a diagnostic tool deployed in all of the CERN synchrotrons, especially in the accelerators of the PS complex. The first operational version was written in HighPerformance Fortran to minimise the runtime. In 2020, a Python implementation was developed with parts of the code written in C++ to avoid significant loss of performance. Now, to further accelerate the code, a GPU-accelerated version has been developed, which uses CuPy and CUDA to benefit from GPU performance whilst maintaining a user-friendly Python interface. This note describes the latest changes to the Longitudinal Tomography library and gives an overview of the impact on performance

Fields of Study & Subjects

Particle accelerators and beam dynamics Parallel Computing and Optimization Techniques Computational Physics and Python Applications

Publication details

Journal / source
CERN Document Server (European Organization for Nuclear Research)
Publication date
5 March 2026
Language
EN