Modern GPUs (graphics processing units) are highly parallel multicore processors that are capable of non-graphics-related general-purpose computations. For compute-intensive tasks, they have been reported to outperform conventional CPUs by up to a factor of 10.

Mass spectrometry is an analytical technique that identifies the chemical composition of a compound or sample based on the mass-to-charge ratio of charged particles. A sample undergoes chemical fragmentation, thereby forming charged particles (ions). The ratio of charge to mass of the particles is calculated by passing them through electric and magnetic fields in a mass spectrometer. (Source: http://en.wikipedia.org/wiki/Mass_spectrometry)

The goal of this project is to investigate and implement methods to parallelise algorithms to analyse mass spectrometry data such that they can be executed efficiently on modern GPUs.

The School recently acquired an NVIDIA Tesla 1070 http://www.nvidia.com/object/product_tesla_s1070_us.html, which will be used for this project.