Techniques for Large Data Visualization
Authors: Dan R. Lip¸sa 1, Robert S. Laramee 1, R. Daniel Bergeron 2, and Ted M. Sparr 2
Abstract:
Often scientific datasets are several times larger than the main memory of a computer. The size of datasets, in general, has exceeded that of main memory for several decades and will continue to do so for the foreseeable future. Because of large disk-drive latency, visualization algorithms designed to process data from main memory can rarely be directly applied to data stored on disk without modification. In this paper we review current methods and techniques designed to deal with large data, often larger than the computer’s main memory. Our goal is to provide a student or researcher with understanding of fundamental concepts and knowledge of the most important techniques in the current research literature for visualizing large scientific data. The most important terminology related to
out-of-core visualization is identified and discussed as well as the fundamental challenges faced by this class of techniques. We provide a valuable starting point for readers interested in gaining a concise introduction of techniques for large data visualization.