M.Garbey
With the emergence of parallel supercomputers, the art of computational science has leaped to a new era. Problems that were not tractable before are now being solved. The huge amount of computational resources have been used for searching new drugs, building efficient aeroplanes, discovering new composite materials, predicting snow storms, and understanding human genetic structures. Although the physics remains unchanged, revision of existing algorithms for numerical simulations on parallel computers is most likely inevitable. The dynamic computer industry has moved so swiftly that working on parallel computers is no longer limited to researchers at prestigious institutions or national laboratories. Joint Institute for Computational Science (http://www.jics.utk.edu/PCUE/)
Fall 2004.
The goal of this class is to introduce you progressively to the exciting world of High Performance Computing (HPC). You will find that HPC is applicable to many fields such as Mechanics, Physics, Chemistry, Environment studies, Biology, Medicine, etc.
The class will lead you step by step to practical applications.
The plan of the course is as follow :
Brief review of computational science methods Basic concepts of parallel computing Inner Product, Matrix Vector product, Matrix Multiplication and BLAS Fortran 90 and processing of arrays Message Passing Interface (MPI) Iterative solvers for systems of linear equation Direct solver for system of linear equation Heat transfer Potential flow Wave propagation Cellular Automata
We will use extensively the material given on this web server http://www.jics.utk.edu/PCUE/, and many benchmark codes written by the instructor and his collaborators.
The evaluation will be based on
Interaction in the class : 20 % Written exam : 40 % Software project : 40 %
Some Lecture Notes