Molecular phylogenetic inference, constructing evolutionary histories from molecular data, is an activity universal to all biological disciplines. Phylogenetic inference is an essential analytical technique used in almost all areas of biology from virology and immunology, to protein evolution, from protein networks to evaluating extinction risks and from organismal studies to the analysis of gene evolution. The use of molecular phylogenies in the literature has grown rapidly over the past 10 years. The volume of molecular data has grown nearly exponentially, due to a roughly 50,000 fold reduction in sequencing costs. During the same time period the accuracy and complexity of analytical models has increased. The volume of data and increase in model complexity far exceed advances in computing power. This has created a widening gap between data sets and methods researchers would like to use and available computing power and scalable software. While there are currently a number of methods to scale phylogenetic inference tools, they tend to only be effective up to hundreds of cores. This project will develop highly scalable algorithms for molecular phylogenetic inference, using a range of techniques, including parallelisation of existing methods, development of highly scalable algorithms and GPU acceleration.