High performance distributed computing

High Performance Distributed Computing has created a new approach to science. High performance can be achieved by distributing the task over different machines, using a better algorithm, using faster computers or by using an MPSoC (Multiple Processor System on Chip). In MPSoC multiple CPUs are connected with each other to accomplish a task as a whole. Each CPU assigned a different task. In this paper we have compared the computation speed of a Linux based cluster and an MPSoC by running the same application and repeated this for different applications, which gives the high performance for the distributed systems. For this we have created a cluster of five Linux based systems by using a complete clustering solution that is OSCAR-5.1b2 which is fully supported for RHEL-5(Red Hat Enterprise Linux-5) and an MPSoC with the help of Embedded System Environment Tool which produce Transaction level model for functional verification and performance estimation.