Summary: Image Layer Decomposition for Distributed Rendering on NOWs
Thu D. Nguyen # and John Zahorjan +
tdnguyen@cs.rutgers.edu -- zahorjan@cs.washington.edu
Technical Report UW­CSE­99­10­01
Department of Computer Science & Engineering
University of Washington
(submitted for publication)
October 4, 1999
Abstract
The work in this paper is motivated by our effort to build a distributed rendering system that uses
the multiple hardware graphics accelerators available in a network of workstations/PCs to provide real­
time rendering performance that is one or two generations ahead of what can be achieved using only a
single commodity machine. Specifically, we address one important aspect of building such a system,
how to partition and assign the overall rendering work. We propose a novel approach called Image Layer
Decomposition (ILD). ILD has several advantages over previous partitioning algorithms for our targeted
environment, including its compatibility with the use of hardware graphics accelerators, decoupling of
communication bandwidth requirement from scene complexity, and reduced communication bandwidth
growth as the system size increases. Furthermore, ILD tries to optimize the rendering of a sequence of
frames (of an interactive application) instead of only individual frames.
Using traces taken from a VRML viewer, we simulate a long­running off­line version of ILD to e­

Source: Anderson, Richard - Department of Computer Science and Engineering, University of Washington at Seattle

Collections: Computer Technologies and Information Sciences