A parallel scaled conjugate-gradient algorithm for the solution phase of gathering radiosity on hypercubes
Date
1997Source Title
Visual Computer
Print ISSN
0178-2789
Electronic ISSN
1432-2315
Publisher
Springer
Volume
13
Issue
1
Pages
1 - 19
Language
English
Type
ArticleItem Usage Stats
196
views
views
167
downloads
downloads
Abstract
Gathering radiosity is a popular method for investigating lighting effects in a closed environment. In lighting simulations, with fixed locations of objects and light sources, the intensity and color and/or reflectivity vary. After the form-factor values are computed, the linear system of equations is solved repeatedly to visualize these changes. The scaled conjugate-gradient method is a powerful technique for solving large sparse linear systems of equations with symmetric positive definite matrices. We investigate this method for the solution phase. The nonsymmetric form-factor matrix is transformed into a symmetric matrix. We propose an efficient data redistribution scheme to achieve almost perfect load balance. We also present several parallel algorithms for form-factor computation.
Keywords
Data redistributionGathering radiosity
Hypercube Multicomputer
Parallel Algorithms
Scaled Conjugate-Gradient Method
Computational Methods
Computer Architecture
Image Processing
Matrix Algebra
Form Factor Computation
Gathering Radiosity Method
Linear Equations
Color Computer Graphics