Optimal visualization pipeline partitioning technique on the grid environment

Abstract

Remote visualization of large datasets requires the development of distributed visualization pipeline. Additionally, the computational environment should provide easy configuration and automatic selection technique for the pipeline according to the resources available. Grid computing provides the required infrastructure with limited services particularly for remote visualization. The current approaches for grid enabled visualization follow static and manual selection and mapping of the resources. This paper investigates the mechanism for predicting and partitioning the visualization pipeline on the grid environment. We propose a grid computing technique for dynamic partitioning for visualization pipeline with consideration of network bandwidth and total time delay between the visualization pipeline components.