Improvements to ATLAS Inner Detector Track reconstruction for LHC Run-3

Zachary Michael Schillaci

doi:10.1051/epjconf/202125103048

Abstract

This document summarises the main changes to the ATLAS experiment’s Inner Detector Track reconstruction software chain in preparation of LHC Run 3 (2022-2024). The work was carried out to ensure that the expected high-activity collisions with on average 50 simultaneous proton-proton interactions per bunch crossing (pile-up) can be reconstructed promptly using the available computing resources. Performance figures in terms of CPU consumption for the key components of the reconstruction algorithm chain and their dependence on the pile-up are shown. For the design pile-up value of 60 the updated track reconstruction is a factor of 2 faster than the previous version.

Highlights

The reconstruction of charged particle trajectories in the Inner Detector (ID) is a complex part of the ATLAS [1,2,3] experiment’s event reconstruction chain, making it the most resource intensive component during Run 2 of the LHC (2015-2018)
The clusters need to be combined into short track seeds that are subsequently attempted to be extended through the entire ID to identify the charged particles and precisely reconstruct their trajectories
All the events taken from this run fall under the so-called good-run list (GRL), meaning that they satisfy all data quality requirements to be considered part of the ATLAS physics dataset

Summary

Introduction

The reconstruction of charged particle trajectories (tracking) in the Inner Detector (ID) is a complex part of the ATLAS [1,2,3] experiment’s event reconstruction chain, making it the most resource intensive component during Run 2 of the LHC (2015-2018). The clusters need to be combined into short track seeds that are subsequently attempted to be extended through the entire ID to identify the charged particles (tracks) and precisely reconstruct their trajectories This represents a complex combinatorial problem, which increases in difficulty with pile-up. One main direction not discussed here was the adoption of multi-threading to make more efficient use of the available resources This is not, a priori, expected to improve processing time required by the algorithms. In addition to this infrastructural change, a major effort [5] was carried out to improve the per-thread performance of track reconstruction while maintaining comparable or even superior quality of the reconstructed tracks. The changes made in this effort and their impact are described in this paper

Methods

Results

Conclusion