Shaping the future CMS event selection at the HL-LHC with Next-Generation Triggers
by
1/1-2 - Aula "C. Voci"
Dipartimento di Fisica e Astronomia - Edificio Marzolo
The Next-Generation Triggers (NGT) programme is exploring novel concepts and technologies for the trigger systems of the High-Luminosity Large Hadron Collider (HL-LHC), with the goal of redefining how collision events are selected, reconstructed, and recorded in the challenging environment of extreme luminosity and pile-up. This contribution presents recent developments within the CMS NGT effort, focusing on its two main components: the Level-1 Trigger (L1T) and the High-Level Trigger (HLT).
For the L1T, the CMS Phase-2 upgrade introduces significantly enhanced capabilities through the deployment of modern FPGA technologies together with high-granularity detector information at the earliest stage of the trigger decision. These improvements enable low-latency Machine Learning inference, supporting more sophisticated event classification and anomaly detection. In addition, the evolution of the Level-1 scouting concept extends the trigger beyond conventional event selection toward real-time analysis of the full 40 MHz collision rate, increasing the sensitivity to low-energy processes and other rare physics signatures.
Within the HLT, the Real-time Reconstruction Revolution (R³) project is advancing a new data-scouting strategy in which every event accepted by the L1T can be reconstructed and made available for prompt physics analyses. This is achieved through the combination of heterogeneous computing resources, highly optimized data formats, and efficient reconstruction algorithms that substantially reduce processing time while maintaining excellent physics performance. A continuous real-time calibration framework further enhances reconstruction quality, and the event content is carefully optimized to preserve the most relevant physics information while minimizing data size.
Taken together, these developments represent a major evolution of the CMS trigger system, shifting its role from a pure event-selection mechanism toward a real-time analysis platform capable of maximizing the physics potential of every recorded collision.
The seminar is organized by INFN-PD Gruppo 1.
Mia Tosi, Alberto Zucchetta (INFN)