Experimental Particle and Astro-Particle Physics Seminar

Europe/Zurich
Stefanos Leontsinis (University of Zurich (CH))
Description

Abstract: The Next-Generation Triggers (NGT) project aims to develop new ideas and technologies for the trigger systems at the High-Luminosity Large Hadron Collider (HL-LHC), fundamentally rethinking how experiments select, reconstruct, and retain collision data in a regime of unprecedented luminosity and pile-up. This contribution highlights recent advances of the NGT programme within the CMS experiment, spanning its two central pillars: the Level-1 Trigger (L1T) and the High-Level Trigger (HLT). 
At the L1T, the Phase-2 upgrade unlocks new capabilities by integrating more advanced FPGA architectures and high-granularity detector information directly into the first decision layer. This enables real-time Machine Learning inference within the available latency, allowing sophisticated event interpretation and anomaly detection. Beyond its traditional filtering role, the L1 scouting paradigm moves toward real-time analysis at 40 MHz, expanding sensitivity to previously inaccessible low-energy and rare signatures. 
At the HLT, the Real-time Reconstruction Revolution (R³) project drives event processing toward a data-scouting paradigm in which all L1T-accepted events are recorded for prompt physics analysis. By combining heterogeneous computing, optimized data formats, and advanced algorithms, the system drastically reduces processing time while preserving physics performance. A real-time calibration loop ensures continuous on-the-fly optimization, while the output format is carefully tuned to retain essential physics information with minimal data volume. 
Together, these innovations transform the trigger from a selective filter into a real-time analysis system, unlocking the full physics potential of every collision.

zoom room link