A low-cost Java-based parallel computing platform for integrated measurement and data processing
Chun Che Fung, C. Engel, Halit Eren · 2002
This paper reports an investigation into the feasibility of utilizing a low-cost Java-based parallel computing platform for measurement and data processing purposes. The proliferation of high speed, low latency networking has provided a useful and inexpensive alternative to supercomputers and other specialized data processing hardware. The proposal is to use a number of networked PCs for the execution of distributed tasks. In the past, these functions were mainly executed exclusively on high performance computers. Several arbitrary tasks are presented that are executed in parallel on a number of desktop PCs, each of which are networked by a standard IP stack. The programming environment utilized is Java 2.0, from Sun Microsystems. The Java programming language has definite advantages such as independence from platforms and suitability for network executions. The proposed approach has good potential for distributed measurement and data processing applications.