The ALICE experiment at LHC will implement a High Level Trigger System for online event selection and/or data compression. The largest computing challenge is imposed by the TPC detector, requiring real-time pattern recognition. The main task is to reconstruct the tracks in the TPC, and in a final stage combine the tracking information from all detectors. Based on the physics observables selective readout is done by generation of a software trigger (High Level Trigger), capable of selecting interesting (sub)events from the input data stream. Depending on the physics program various processing options are currently being developed, including region of interest processing, rejecting events based on software trigger and data compression schemes. The system entails a very large processing farm, designed for an anticipated input data stream of 25 GB/s. In this paper we present the architecture of the system and the current state of the tracking methods and data compression applications.

No abstract

No abstract