2nd NextGenTriggers Hackathon

This hackathon is open to anyone interested in participating. Each participant will have to select one topic that is interested in. The hackathon will be divided in smaller groups per topic. The goal is to brainstorm and share ideas and knowledge, away from emails, meetings, and other distractions.

Topics are the following:

• ML for PUPPI in the Correlator Trigger
• ML Ops
• Particle Flow Taus in L1T Scouting workflow
• Extending the extension: including OT forward disk in the CA algorithm
• Explore job orchestration systems such as Luigi and the CMS Automation Framework to streamline the submission of multi-step workflows for deriving calibrations in the NGT HLT 3.4 task “Optimal Calibrations.” Our goal is to prototype an efficient orchestration strategy within the CMS NGT demonstrator, targeting applications in Run-3 2026 and Phase-2.
• Common graph definition: Many seeding and first stage pattern recognition algorithms deploy a graph model for modelling track candidates or track segments. The target of this hackathon item would be to review existing data model concepts, and identify, find or define efficient graph data structures for CPU and GPU applications.
• detray GPU geometry: Detray is the GPU geometry model of the traccc track reconstruction and is designed to provide efficient navigation and (and propagation) through various detector setups. The target of this hackathon item is to investigate how flexible this library could be expanded describe detectors other than the tested OpenDataDetector, ATLAS ITk and FCC-hh.
• Feasibility Study of the ACTS Framework for CMS Phase-2 Track Reconstruction: This week-long project investigates the feasibility of employing A Common Tracking Software (ACTS) framework as a potential component of the track reconstruction chain for the CMS detector in the HL-LHC Phase-2 upgrade. The study focuses on understanding ACTS’ design principles, data model, and integration requirements, and on exploring how its capabilities align with the performance, flexibility, and scalability needs of CMS. The goal is to provide an informed assessment of whether and how ACTS could be integrated into the CMS software stack, and to identify possible technical challenges or limitations.

For more information, do not hesitate in contacting the people listed under the "contact" section.