Publication: Parallel programming with distributed shared memory
Loading...
Date
2006-05-01
Authors
Lee, Pau Hua
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
There is a continual demand for greater computational speed from a computer system than is currently possible because it seems that there will always be applications that require still more computation power. One way of increasing the computation speed is by using multiple computers that linked through network operating together on a single problem that split into parts, and each performed by a separate computer in parallel. Such application can gain advantageous on performance, scalability, and even price. In this reseach, a parallel matrix multiplicaton system will be designed using Parallel Programming with Distributed Shared Memory. With the message-passing method, this system will perform the matrix multiplication computation in parallel by using multiple computers that linked through network. The computation time for the parallel matrix multiplication will be compared with the computation time for serial matrix multiplication and will expecting to have a faster computation speed from the parallel comutation method on the larger matrix scale. In overall, the computation cabability of the conventional computer can be improved by the application of parallel computer system.