Physics and Computing Performance of the EggNet Tracking Pipeline

TL;DR

This paper evaluates the EggNet tracking pipeline, a novel GNN-based particle tracking method, focusing on its physics accuracy and computational efficiency on the TrackML dataset, and explores techniques to optimize resource usage.

Contribution

It provides a comprehensive assessment of EggNet’s physics performance and computational scalability, introducing methods to reduce memory and time constraints.

Findings

EggNet achieves improved edge efficiency and purity.

The pipeline demonstrates scalable performance on large datasets.

Optimization techniques effectively reduce computational resource requirements.

Abstract

Particle track reconstruction is traditionally computationally challenging due to the combinatorial nature of the tracking algorithms employed. Recent developments have focused on novel algorithms with graph neural networks (GNNs), which can improve scalability. While most of these GNN-based methods require an input graph to be constructed before performing message passing, a one-shot approach called EggNet that directly takes detector spacepoints as inputs and iteratively apply graph attention networks with an evolving graph structure has been proposed. The graphs are gradually updated to improve the edge efficiency and purity, thus providing a better model performance. In this work, we evaluate the physics and computing performance of the EggNet tracking pipeline on the full TrackML dataset. We also explore different techniques to reduce constraints on computation memory and computing time.