Algorithms for automatic parallelism, optimization, and optimal processor utilization

In this thesis we first investigate the reaching definitions optimization. This compiler optimization collects and stores information about where a variable is defined and how long that definition of the variable stays alive before it is redefined. We compare the old iterative solution to a new algorithm that uses the partialout concept. The partialout solution decreases execution time by eliminating the multiple passes required in the iterative solution. Currently, compilers that find a data dependence between two statements in a loop do not parallelize the loop. Next we investigate automatic parallelism for these loops by breaking the loop into a set of smaller loops, each of which contains no dependencies and thus can be executed in parallel. Finally, we introduce a set of algorithms for optimal processor utilization. The algorithms split the loop into a sequential series of parallel blocks, each block executing in parallel and utilizing the optimal number of processors possible.

Name thesis:degree_name

Master of Science (MS)

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Computer Science

Grantor dc:publisher

University of Nevada, Las Vegas

Year

1993

Author dc:creator

• Macke, Katharine J

Statement dc:rights

• IN COPYRIGHT. For more information about this rights statement, please visit http://rightsstatements.org/vocab/InC/1.0/

Language dc:language

English

Identifier

https://oasis.library.unlv.edu/rtds/315

OAI identifier oai:identifier

oai:oasis.library.unlv.edu:rtds-1314