Next Generation Triggers - Public proposal

The High Energy Physics (HEP) program at CERN has achieved major breakthroughs in particle physics, technology, and algorithms, including the discovery of the Higgs boson in 2012. This allowed the validation of compatibility of the theoretical construction behind the Standard Model (SM) of particle physics with the data, but the existing uncertainties leave room for models beyond the SM. With the experimental collider framework in place, scientific exploration continues to answer questions around
the origin of dark matter, the disproportionately low abundance of antimatter and the nature of the discovered Higgs boson. Hard physics problems aside, much can be gained from improvements to the data acquisition pipeline allowing for capturing a richer set of collision events, furthering scientific understanding.

The Large Hadron Collider (LHC) consists of a 27 km tunnel where superconducting magnets guide bunches of protons, circulating in opposite directions, which are then caused to collide at experimental sites (e.g. ATLAS and CMS) at a rate of 40 million times per second. The collision events emit various particles, which are tracked through a multitude of radiation-hardened detectors and fed into the L1 trigger system, which needs to reject >99% of the events within 10 microseconds due to detector cache
constraints and available network capacity.

This data is further reduced by >99% in the High-Level Trigger (HLT) to conform to the current event analysis and simulation capacity. HEP experimentation is fundamentally stochastic, so without changing other factors, an increase in data collection throughput would allow for higher confidence in current results while increasing the likelihood of detecting novel particles in the current LHC setup. Furthermore, this capacity increase is absolutely needed for future LHC upgrades where each collision will have many more interesting events. The interpretation of the LHC data relies on theoretical simulations of particle interactions in the Standard Model (SM) and in scenarios of new physics beyond the SM (BSM). The full exploitation of the immense HL-LHC datasets, and in perspective of the data from Future Colliders, will require radical improvements in the computing strategies of theory calculations, to increase their accuracy while keeping affordable computing times. A multitude of theoretical tools must be addressed, in a coordinated effort, to preserve their interoperability and harmonize the overall precision. In addition to the several ingredients needed to describe the final states of proton collisions, the infrastructure developed for the triggers, e.g. the GPU cluster, also supports the advancement of software and algorithms for lattice Quantum Field Theory (LQFT) calculations, as a unique approach to control relevant non-perturbative ingredients. The engagement of LQFT experts would also bring to the trigger 1 community complementary expertise and experience in parallel architectures. The progress envisaged with these theoretical tasks complements and augments the benefits of the increased capacity to trigger and record relevant data. The goal of this proposal is to facilitate improvements to LHC data collection and processing beyond current capabilities, while looking forward to future data collection needs, through four work packages.

The R&D work done to optimize the current Run 3 and the following High-Luminosity (HL)-LHC phases will provide critical insight to develop future detectors and data flows for the even more ambitious objectives of the Future Circular Collider (FCC) currently in its Feasibility Study phase. We consider that such an ambitious programme requires co-development partnerships with experts in academia and industry to accelerate the achievement of the objectives.

This work has been funded by the Eric & Wendy Schmidt Fund for Strategic Innovation through the CERN Next Generation Triggers project under grant agreement number SIF-2023-004.