The design and implementation of a distributed parallel computing model
| Autor: | Meng-xiang Chen, 陳盟翔 |
|---|---|
| Rok vydání: | 2005 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 93 With the increase of the performance of the PC, users can deal with the data and computation more efficiently. However, it is still very difficult to handle complicated computation by only depending on a single PC. A parallel computer has many CPUs. It can support huge processing and computation. However, it is a big problem for its high cost and meanwhile it is difficult to obtain, too. Connecting many PCs to perform the distributed parallel computation has become more popular due to the high cost of supercomputers. Most distributed parallel computing systems are designed for systems of special purpose. They are not designed for general purpose parallel computing. In this thesis we will design a general-purpose distributed parallel computing model. This model is very easy to implement for general parallel programs or for any general parallel algorithms. We will build the distributed parallel computing model through Java programming language and CORBA. In this thesis, we will generate slave classes by using the concept of Factory Object. The new classes can be created dynamically, and then connects to ORB for the slaveThread of master process. Finally, we will implement the parallel computing model for the quicksort and rank sort programs. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|