The core of my work is building high performance computing software systems to solve numerical problems. This can be on small embedded board computers at one extreme on supercomputers with hundreds of thousands of nodes. It includes using heterogeneous accelerators on these systems: FPGAs and GPUs. This work involves collaboration with colleagues in several problem domains including: atomic and molecular scattering physics, electromagnetic fields, image analysis and clinical computational physiology.

For the past few years I have worked extensively with colleagues in the QUB Medical School to develop data analytics and machine learning systems that process streams of physiological parameters from the bedside of patient in an intensive care unit.