Connecting The Dots 2022

Session

Poster session

31 May 2022, 10:30

Princeton University

25. Improving the performance of dealing with non-ideal tracks within ATLAS reconstruction

Donal Joseph Mc Laughlin (UCL)

31/05/2022, 10:30

Poster

Within high transverse momentum jet cores, the separation between charged-
particles is reduced to the order of the sensor granularity in the ATLAS tracking detectors, resulting in overlapping charged-particle measurements in the detector. This can degrade the efficiency of reconstructing charged-particle trajectories. This presentation identifies the issues within the current reconstruction...

36. ExaTrkX as a Service

Xiangyang Ju (Lawrence Berkeley National Lab. (US))

31/05/2022, 12:30

Poster

Particle tracking plays a pivotal role in almost all physics analyses at the Large Hadron Collider. Yet, it is also one of the most time-consuming parts of the particle reconstruction chain. In recent years, the Exa.TrkX group has developed a promising machine learning-based pipeline that performs the most computationally expensive part of particle tracking, the track finding. As the pipeline...

8. Linearized Track-Fitting on an FPGA

Joachim Zinsser (Ruprecht Karls Universitaet Heidelberg (DE))

31/05/2022, 15:30

Poster

For the ATLAS experiment at the High-Luminosity LHC, a hardware-based track-trigger was originally envisioned, which performs pattern recognition via AM ASICs and track fitting on an FPGA.
A linearized track fitting algorithm is implemented in the Track-Fitter that receives track candidates as well as corresponding fit-constants from a database and performs the $\chi^2$-test of the track as...

30. Optimizing the Exa.TrkX Inference Pipeline for Manycore CPUs

Alina Lazar

31/05/2022, 15:33

Poster

The reconstruction of charged particle trajectories is an essential component of high energy physics experiments. Recently proposed pipelines for track finding, built based on the Graph Neural Networks (GNNs), provide high reconstruction accuracy, but need to be optimized, in terms of speed, especially for online event filtering. Like other deep learning implementations, both the training and...

12. Software Performance of the ATLAS Track Reconstruction for LHC Run 3

Makayla Vessella (University of Massachusetts (US))

31/05/2022, 18:00

Poster

This poster summarizes the main changes to the ATLAS experiment’s Inner Detector track reconstruction software chain in preparation for 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...

15. Deep Learned Algorithmic Combinatorics

Dr Mikael Mieskolainen (Imperial College London (GB))

Poster

We introduce a new algorithmic deep architecture which combines graph neural networks, set transformers and Monte Carlo tree search like random sampling. The algorithm targets large scale combinatorial inverse problems, such as clustering of hypergraphs, encountered in high energy physics and beyond. We demonstrate the method on the tracking problem of high energy collisions.

41. The LHCb Co-Processors Testbed

Andrea Contu (INFN)

Poster

In 2022, at the beginning of the LHC Run 3, the LHCb DAQ and software trigger will reconstruct events at an average bunch crossing rate of 30 MHz. In view of future upgrades steps should be taken towards exploitation of an heterogeneous computing model, in which dedicated co-processors, highly optimised for specific tasks, are used in the same DAQ infrastructure. In order to investigate viable...