Line Segment Tracking in the HL-LHC

TL;DR

This paper introduces a novel bottom-up line segment tracking algorithm for the HL-LHC that leverages parallel processing on GPUs, achieving efficient charged particle track reconstruction in high pile-up conditions.

Contribution

It presents a new track reconstruction method that constructs and links track stubs in a parallelizable way, suitable for GPU architectures, for the first time in HL-LHC context.

Findings

Achieves good physics performance in track reconstruction.

Runs efficiently on NVIDIA Tesla V100 GPUs.

Demonstrates potential for future high-luminosity collider experiments.

Abstract

The major challenge posed by the high instantaneous luminosity in the High Luminosity LHC (HL-LHC) motivates efficient and fast reconstruction of charged particle tracks in a high pile-up environment. While there have been efforts to use modern techniques like vectorization to improve the existing classic Kalman Filter based reconstruction algorithms, Line Segment Tracking takes a fundamentally different approach by doing a bottom-up reconstruction of tracks. Small track stubs from adjoining detector regions are constructed, and then these track stubs that are consistent with typical track trajectories are successively linked. Since the production of these track stubs is localized, they can be made in parallel, which lends way into using architectures like GPUs and multi-CPUs to take advantage of the parallelism. The algorithm is implemented in the context of the CMS Phase-2 Tracker and runs on NVIDIA Tesla V100 GPUs. Good physics and timing performance has been obtained, and stepping stones for the future are elaborated.