Title：MPI on Exascale Systems: Current Trends and Opportunities

Speaker：Dr. Pavan Balaji

Time：10:00-12:00, Wednesday, November 6th, 2013

Venue： Lecture Room, 4rd Floor, Building #5,

Profile：Dr. Pavan Balaji holds appointments as a Computer Scientist at the Argonne National Laboratory, as an Institute Fellow of the Northwestern-Argonne Institute of Science and Engineering at Northwestern University, and as a Research Fellow of the Computation Institute at the University of Chicago. He leads the Programming Models and Runtime Systems group at Argonne. His research interests include parallel programming models and runtime systems for communication and I/O, modern system architecture(multi-core, accelerators, complex memory subsystems, high-speed networks), and cloud computing systems. He has nearly 100 publications in these areas and has delivered nearly 120 talks and tutorials at various conferences and research institutes.

Abstract：Message Passing Interface (MPI) is the most widely used parallel programming model today. However, as we think about Exascale systems that are expected to arrive in the next decade, the fundamental architectural changes that are upcoming need us to rethink how we use parallel computers and in turn parallel programming models. Is MPI still the right programming model for exascale systems? What modifications do we need to make in the MPI standard and/or implementations to be suitable for such architectures? In this talk, I will discuss some of the major expected changes in Exascale systems and the various efforts in my group to address these challenges, both in terms of new capabilities in MPI as well as optimizations to MPI implementations.