Speaker: Dr Horacio González-Vélez, School of Computing, National College of Ireland
Title: "Beyond Fast and Furious: Smart approaches to data-intensive parallel programming for the physical and life sciences"
Abstract: Computationally intensive applications in the physical and life sciences have long posed a challenge to known parallel systems programming techniques, as a result of their inherent heterogeneity and dynamism of their calculations and datasets. This seminar will present some ideas on how to exploit the structure and behaviour of data-intensive parallel applications to efficiently allocate dispersed computational resources taking into account non-functional requirements. By abstracting commonly-used patterns of parallel computation, communication, and interaction, software developers can arguably enable parallel applications with substantial data requirements to adapt to the most suitable computational resources. The seminar will also discuss some lessons learned from the work in cHiPSet, a major European COST Action on High-Performance Modelling and Simulation for Big Data Applications http://chipset-cost.eu/ with specific emphasis on the implications to the physical and life sciences.