20 December 2017 - 30 December 2017
10 Days GiAN Course on Computational Methods and Parallel Processing in Science and Technology
The generic interdisciplinary approach of scientific computing is generally considered to be a main pillar of science and technology, complementary to experiment and theory. To carry out such interdisciplinary approach there is an urgent demand of researchers who are well-trained in scientific computing methods. It is the particular ambition of the emerging research communities to promote the use of computational methods in yet largely unexplored areas such as targeted medicine, cancer therapy, electromagnetics, magnetohydrodynamics, ferrohydrodynamics, naval architecture, aerodynamics, electro-magnetic wave propagation in optical waveguide etc. To meet this demand, the course, computational Methods and parallel processing in science and technology, emphasis the state-of-art in contemporary mathematical methods for the problems related to the fluid dynamics with a wide spectrum of industrial applications.
The increasingly important role of numerical methods in science and engineering will be addressed. While treating traditional and well-developed topics, the emphasis will be placed on the concepts and ideas of importance to the design of accurate and efficient algorithms with the applications to scientific computing. The course combines the deep understanding of the computational methods and algorithms in Applied Mathematics, the implementation of these methods using modern computer science technology and the application to real-world problems.
Topics will include the simulations of complex physical, biological, and engineering systems, optimization and evaluation of simulation models, scientific visualization and parallel computing on GPU (Graphics Processing Unit). The comparison between the CPU programming and the hardware-oriented GPU programming executed in machines with much larger number of processing cores will be dealt. Implementation of modern GPUs in different parallel applications will be taught to reduce a large amount of computing time. Mathematical techniques including finite-difference, finite-volume, finite-element and particle methods with their accuracy, convergence, and stability of numerical methods; turbulence modeling; Parallel computing and multiphysics modeling will be the major contents of this course.
At the end of the course the participants will be able to pursue innovative research work with a strong application-oriented focus anywhere from mathematics, physics and engineering sciences.
Dr. H.P. Rani,
contact no: 91 870 246 2834,