Friday, December 03, 2004
Wednesday, December 01, 2004
Introduction to this project
In recent years application development support for uni-processors has grown dramatically. A lot of tools are available to assist developers in a wide range of business and scientific domains. In numerical computing and computational tools such as MATLAB, the implementation details of fundamental algorithms used repeatedly in scientific applications.
On massively parallel platforms or distributed environments, application development is still difficult and time consuming due to the scarcity of good supporting tools.
The objective of this research is to analyze the current existing MPI programs, try to test some of them and find out a suitable one for our laboratory use, and if needed try to modify it in order to get higher performance.
On massively parallel platforms or distributed environments, application development is still difficult and time consuming due to the scarcity of good supporting tools.
The objective of this research is to analyze the current existing MPI programs, try to test some of them and find out a suitable one for our laboratory use, and if needed try to modify it in order to get higher performance.
