Title :
Maintaining dynamic geometric objects on parallel processors
Author_Institution :
Dept. of Inf., Dortmund Univ., Germany
Abstract :
Parallelizing dynamic scientific applications involves solving the dynamic load balancing problem. The balancing should take the communication requirements of the application into account. Especially in the field of computer graphics, many problems deal with objects in k-dimensional space with very special communication patterns. We describe a load balancing strategy, based on a geometric clustering of objects, with provable good performance, that is well suited to applications with interdependencies between neighbouring objects or between objects that have at least one coordinate in common
Keywords :
computer graphics; data visualisation; natural sciences computing; parallel processing; resource allocation; common coordinate; communication requirements; computer graphics; distributed data structures; dynamic geometric object maintenance; dynamic load balancing strategy; dynamic scientific applications; geometric clustering; locality; object interdependencies; orthogonal communication pattern; parallel processors; performance; rendering application; Application software; Chromium; Computer graphics; Gravity; Layout; Load management; Rendering (computer graphics); Scientific computing; Space exploration; Sparse matrices;
Conference_Titel :
Parallel Rendering, 1997. PRS 97. Proceedings. IEEE Symposium on
Conference_Location :
Phoenix, AZ
Print_ISBN :
0-8186-8265-5
DOI :
10.1109/PRS.1997.628293
