Load balancing for multi-projector rendering systems

Publisher Website
Google Scholar
Abstract
Multi-projector systems are increasingly being used to pro vide large-scale and high-resolution displays for next-genera tion inter- active 3D graphics applications, including large-scale da ta visual- ization, immersive virtual environments, and collaborati ve design. These systems must include a very high-performance and scalable 3D rendering subsystem in order to generate high-resolutio n images at real time frame rates. This paper describes a sort-first ba sed par- allel rendering system for a scalable display wall system bu ilt with a network of PCs, graphics accelerators, and portable projec tors. The main challenge is to develop scalable algorithms to partiti on and assign rendering tasks effectively under the performance a nd func- tionality constrains of system area networks, PCs, and commodity 3-D graphics accelerators. We have developed three coarse-grained partitioning algorithms and incorporated them into a worki ng pro- totype system. This paper describes these algorithms and reports our initial experimental results aimed at investigating th e feasibility of constructing a sort-first rendering system using a networ k of PCs and evaluating algorithmic trade-offs and performance bottlenecks within such a system. Our results indicate that the coarse-g rained characteristics of the sort-first architecture are well sui ted for con- structing a parallel rendering system running on a PC cluste r.

This publication has 26 references indexed in Scilit: