Trigger, Software and Computing

• Published in: ATLAS Vol. 30; pp. 197 - 224
• Format: Book Chapter
• Language: English
• Published: Singapore World Scientific Publishing Company 2019; WORLD SCIENTIFIC
• Series: Advanced Series on Directions in High Energy Physics
• Subjects: Particle & high-energy physics
• ISBN: 9813271795; 9789813271791; 9813271817; 9789813271814; 9813271809; 9789813271807
• ISSN: 1793-1339
• DOI: 10.1142/9789813271807_0007

Developing the trigger selection and offline software is one of the major activities in building a modern high-energy physics experiment like ATLAS. As for any experiment at a hadron collider, the online event selection in the trigger farm requires sophisticated event reconstruction and selection algorithms, in order to reduce the immense amount of collision data produced every second to a rate that the offline computing infrastructure can process and deliver to physicists for analysis. Data selected by the trigger system are subject to careful calibration and a prompt reconstruction at computing facilities at CERN. The data in different formats are distributed worldwide for analysis on the LHC Computing Grid, a federation of more than 130 computing centres across the globe at national laboratories and universities. The real data are complemented by simulated data, with generators modelling collision events of all types of physics processes at the LHC. Those generated events are fed into a highly detailed detector simulation that models the experiment response and produces the same data format as that coming from the real detector. The simulated data are used in physics analysis to understand and correct for detector effects. Figure 7.1 illustrates the ATLAS data processing chain.