A Parallel Implementation of the Inspiral Search Algorithm using Triana

Black holes and gravitational waves are among the most fascinating predictions of Einstein’s theory of General Relativity. Today we have indirect evidence for both but have directly observed neither. One of the most promising sources of gravitational waves for detection is the inspiralling compact binary system. This consists of a pair of dense, compact objects (either neutron stars or black holes) with masses of a few to a few tens of Solar masses, orbiting around each other with a period of minutes to hours. Such signals are hoped to be detected using a technique known as matched filtering. We present here an algorithm for performing the matched filtering inspiral search implemented as a workflow of connected Triana units. We use Triana’s ability to distribute taskgraphs over a number of computers to parallelise the algorithm and compare it to a parallel implementation of the same algorithm implemented using MPI. Both implementations run on the same Beowulf cluster, allowing us to compare the trade off between the flexibility of Triana verses the speed of the MPI based code.