Speaker
Description
The Next Generation Triggers (NGT) R3 (Real-time Reconstruction Revolution) project in CMS aims to rethink the experiment’s data acquisition system, allowing its physics programme to process all collisions accepted by the Level-1 hardware-based trigger system (L1T), in view of the Phase-2 upgrade for the HL-LHC. Its main objective is to expand the High-Level Trigger (HLT) data scouting strategy, reconstructing all the data selected by the L1T and recording them in a physics-analysis-oriented, storage-efficient format. These data, in order to be directly suitable for physics analyses, require optimal calibrations at HLT. In situ calibration must occur during data taking, because scouting data is not amenable to further reprocessing. In order to achieve this, implementation of data buffering of all the incoming HLT input data is foreseen, allowing sufficient time to execute the calibration loop. An NGT Demonstrator system, integrated into the current CMS data acquisition (DAQ) system, has been designed, commissioned, and deployed during LHC Run-3, buffering approximately 1% of the input RAW data. This contribution focuses on the hardware implementation and the processing chain of the NGT demonstrator, which comprises a new data stream going into three dedicated DAQ processing nodes. The data is handled and converted using customised DAQ software, which is packaged in RPMs and deployed via CI/CD. The buffered data are stored for 8 hours, to allow for a run of the NGT calibration loop, and then processed using a variant of the regular HLT reconstruction with the enhanced calibration constants. The process is monitored using the standard CMS data flow and data quality monitoring.