Since their first detection in 2015, gravitational waves are rapidly becoming a staple to astronomy and astrophysics, allowing for a new view of the Universe. The Laser Interferometer Space Antenna (LISA) will be the next leap forward for this area of astronomy due to its unique observing frequencies and exceptional sensitivity and size. Discovery and filtering out of the signals coming from galactic binaries will be a primary goal for the LISA data analysis team. Within LISA data there are estimated to be more than 105 galactic binaries visible, requiring the use of complex data analysis tools.
A prototype pipeline for identification of the galactic binaries, called GBMCMC, uses the Bayesian sampling method called Markov Chain Monte Carlo (MCMC). This analysis method allows us to analyze any sized dataset and can identify any number of galactic binaries with calculated confidence levels and uncertainties. However, the highly iterative nature of this method makes the whole process quite computationally expensive. By taking advantage of the parallel tempered methodology used in MCMC, we were able to implement the OpenMP parallelization library within the analysis software. This yielded significant improvements to the analysis time, with time scaling nearly linearly with thread count.