Department of Physics, Panjab University, Chandigarh, India

TPSC seminar

Time and date: 4 November 2011 (Friday), 4:00 PM
Venue: Seminar Hall

Title: Doing parallel: Tools, techniques and scope of parallel compuation in scientific research

Speaker: Dr. Jayanti Prasad, Inter-University Centre for Astronomy & Astrophysics, Pune

Abstract: Parallel computation now no longer belongs to users who have access to supercomputers or large clusters, anyone who has a modern computer, including a laptop, can do parallel computation. Parallel computation not only makes it possible to crunch more data in shorter time, or solve many problems in parallel, it can also be used to attack larger problems and make new discoveries. I will begin my talk by discussing why we need parallel computation and introduce some vocabulary which is used in parallel or high performance computation, followed by a discussion on various models of parallel computation i.e., shared and distributed memory computation. I will give examples of some parallel problems and explain how they can be solved using the tools of parallel computation. Without discussing the syntax of any parallel programing technique, I will highlight some of the important elements of various tools like MPI, OpenMP, Pthread, Cuda etc. For my talk I do not expect any exposure of parallel computation or accessibility to any high-end computing system. However, some familiarity with any programming language will be useful. I will end my presentation by giving examples of some parallel problems, I have been working in the area Astrophysics, and take questions.