Optimisation of the ATLAS Track Reconstruction Software for Run-2

The reconstruction of particle trajectories in the tracking detectors of experiments at the Large Hadron Collider (LHC) is one of the most complex parts in analysing the data from beam-beam collisions. To maximise the integrated luminosity during Run-1 of the LHC data taking period, the number of simultaneous proton-proton interactions per beam crossing (pile- up) was steadily increased. The track reconstruction is the most time consuming reconstruction component and scales non-linearly in high luminosity environments. Flat budget projections (at best) for computing resources during the upcoming Run-2 of the LHC together with the demands of reconstructing higher pile-up collision data at rates more than double compared to Run-1 have put pressure on the track reconstruction software to meet the available computing resources. The ATLAS experiment has thus performed a two year long software campaign which led to a reduction of the reconstruction time for Run-2 conditions by a factor of four: a major part of the changes were improvements to the track reconstruction, which was reduced by more than a factor of five without any loss of output information for subsequent physics analysis. We present the methods used for analysing the software, the planning and deployment of updates and new methods implemented to optimise both algorithmic performance and event data.