Date of Award
Doctor of Philosophy
Electrical and Computer Engineering
James A. Davis
We have developed a scheme to extend the SASL programming language and its run-time system for fine grain parallel processing. The proposed scheme provides a mechanism that can override the original lazy semantics by augmenting proper "eager" information. This information is first annotated in SASL programs and then translated to the combinator control tags by a new set of optimization rules. The effectiveness of this scheme has been evaluated through the simulation of a set of symbolic-oriented programs on an idealized shared-memory system. The results show that a considerable amount of parallelism can be extracted from a wide variety of application programs.
Digital Repository @ Iowa State University, http://lib.dr.iastate.edu/
Pong Ping Chu
Chu, Pong Ping, "Controlling fine-grain non-numeric parallelism on a combinator-based multiprocessor system " (1989). Retrospective Theses and Dissertations. 9115.