An overview of the ATLAS trigger

Summary form only given. This paper gives an overview of the event selection for ATLAS, prior to event building, with the emphasis on the level-2 trigger. (A further stage of selection after event building is described in a separate presentation-ref Event Filter overview.) The level-1 trigger based on custom hardware processes coarse-grain data from the calorimeters and dedicated muon trigger chambers to give a selection of approximately one interaction in 10**5 (one bunch crossing in 10**3). The level-2 trigger processes selected fine-grain data from all detectors to give a further selection of one in 10**2. Studies of the estimated physics rates and detector data volumes have led to a sequential level-2 selection strategy guided by Region of Interest pointers from level-1. The influence of low luminosity (b physics) and high luminosity (high pT) running on the strategy and the architectural options is discussed. The models and testbeds used in developing and testing the strategy are described. Finally the tools, measurements and methodologies to study and compare the various architectural and technical options are summarised.