Parallelizing Sequential Algorithms using MPICH Programming with Case Studies

Parallel computing is getting used to solve problems in multiple domains like scientific, engineering, data mining, transactions processing and so on. Based on the performance requirements of applications, the cost benefits of parallelism are coupled. In this paper, two samples of the diverse applications of parallel computing, namely All-Pairs Shortest-Path problem and Monte Carlo Simulations of Ising Model using MPICH programming are presented, on a cluster and show that this is a viable alternative for high performance computing users in terms of cost and computational time to implement parallel algorithms without compromising on the performance. In this paper foster design methodology [4] is used to demonstrate the parallelization of the algorithms. This design methodology contains 4 steps called, partition, communication, agglomeration, and mapping.