Towards efficient visualization support for single-block and multi-block datasets

Large simulation grids and multi-grid configurations impose many constraints on commercial visualization software. When available RAM is limited and graphics primitives are numbered in millions, alternative techniques for data access and processing are necessary. In this case study, we present our contributions to a visualization environment based on the AVS/Express software. We demonstrate how the efficient visualization of large datasets relies upon several forms of resource sharing, and alternate and efficient data access techniques.