ACTS is a free and open-source software project for track reconstruction in particle physics experiments. As a modernized version of the particle tracking code used by the ATLAS experiment at the CERN Large Hadron Collider, the project is focused on adoption of modern C++ standards, usability in multi-threaded workflows, and increased use of vectorization.
An essential part of track reconstruction is track fitting. In this step, a hypothetical track is refined and updated using measurements from the detector. A commonly used procedure used for fitting is a Kalman filter. This filter iteratively updates a predicted track. Typically, this consists of linear algebra operations on 1-6 dimensional track parametrizations and corresponding up to 6x6 covariance matrices. After a final smoothing process, the optimal track parameters can be extracted.
Rust is a modern compiled programming language designed for system programming. It is multi-paradigm and focused on sane and safe memory management and robust concurrency. Runtime performance is comparable to established languages such as C++. As such, it seems like a promising candidate for future developments of highly performant tracking software with modern design and principles.
The primary objective of this project is to establish a minimal implementation of the components necessary to assess viability of Rust as a language for future developments. This entails comparing runtime performance, as well as the quality of results to the existing C++ implementations in Acts. Key aspects will be an evaluation of solutions for linear algebra operations in Rust. Additionally, an exploration of possibilities to interface components written in Rust with C++ code can be conducted. Implications and requirements on the C++ event data model with respect to optimal communications between components can be studied.
C++ and or Rust, linear algebra