Using performance analysis to identify candidate operations for GIS software parallelization
Abstract:This paper examines the potential impact of parallel computing on the computational performance of GIS. Performance analyses on existing GIS operations, vector topology creation and interpolation, in commercial GIS products in a serial environment are used to work out the critical computational factors that determine elapsed time performance. Analyses reveal that the extraction of all data necessary to perform a GIS operation at a particular geographical locality dominates the I/O and CPU resource usage. For GIS to benefit from parallel computing, the performance of this task must be addressed.
Document Type: Original Article
Affiliations: 1: Edinburgh Parallel Computing Centre, JCMB, King's Buildings, The University of Edinburgh, Mayfield Road, Edinburgh EH9 3JZ, United Kingdom. 2: Department of Geography, University of Edinburgh, Drummond Street, Edinburgh EH8 9XP, United Kingdom. 3: Department of Geography, University of Portsmouth, Buckingham Building, Lion Terrace, Portsmouth PO1 3HE, United Kingdom.
Publication date: December 1, 1997