IMechE-UNITEN Student Chapter Talk Series :Computational Science and Engineering for Discovery

Date :
22 October 2008 (Wednesday)

Time :
4.30 pm to 5.30 pm

Venue:
BW-G-R19 (Lecture Theatre)

Speaker:
Prof. Hasan U. Akay
(Chancellor’s Professor of Mechanical Engineering and Associate Dean
School of Engineering and Technology
Indiana University-Purdue University Indianapolis (IUPUI), USA)

Title of Talk:
Computational Science and Engineering for Discovery

Synopsis:

It is recognized today that the scientists and engineers of this century will be expected to solve problems of unprecedented complexities defining the natural phenomena which can only be solved with computer simulations. While the analytical solutions of these problems will not be possible due to complexity of the equations defining them, it will also be very difficult, if not impossible, to conduct experiments for most of them. These simulations need algorithm complexities and computing powers of unprecedented magnitudes in speed and memory, which will not be overcome by one discipline or one organization alone. As a result, computational science and engineering (CSE) has emerged during the last decade as an indispensable part of scientific and engineering investigations, where the emphasis is on the computation of science rather than the science of computation (i.e., CSE is different than computer science or computer engineering) . Parallel to this, scientific and engineering organizations in many developed countries have been investing heavily on building scientific and engineering computing environments called computing grids for shared usage of resources. These grids are formed by connecting the computer systems at geographically distant locations with fiber optics networks and are administered by various alliances. Mostly targeted for massive data sharing, research, and development, such grid alliances are designed to form community of researchers and educators sharing the hardware and software resources, expertise, and knowledge for scientific discoveries. While the parallel computing algorithms developed by subdividing a large problem into smaller parts and solving each part concurrently on distributed computer systems have long been used, a new computing paradigm called grid computing has emerged during the recent years to harness the enormous power of such computing grids. This has brought many opportunities as well as challenges. One such resource is the TeraGrid supported by the National Science Foundation in USA which houses more than 100 teraflops of computing power distributed at nine sites with facilities capable of data, high-resolution visualization environments, and tools for grid computing. There are already plans to extend the capabilities towards the petaflops level. Indiana and Purdue Universities are two of the sites of this alliance, while IUPUI is the Network Operations Center (NOC) of the facilities. In this talk, in addition to the overview of CSE degree programs and how they can be integrated into traditional engineering and science programs, we will give an overview of the challenges and opportunities brought by the grid computing technology. Experiences gained at the IUPUI Computational Fluid Dynamics Laboratory for solving large-scale computational fluid dynamics problems on the grid will be shared.