Parallel Compression of Time-Variant Scientific Data for Collaborative

Research within the physical sciences is becoming increasingly dependent on the ability to create computer simulations of scientific experiments and analyze the results. Many of these visual representations are created by graphing millions of data values to the screen at each discrete sample point for the length of the simulation. Storing and transmitting all of these data points requires powerful communication capabilities, making it hard to share simulations with other collaborators. In order to facilitate the transfer of these visualizations, a number of compression schemes have been developed. This research has focused on the Trajectory Simplification algorithm and the Trajectory Clustering algorithm that have been previously developed. These methods have been successful in reducing the amount of data needed to accurately represent a given simulation. Even so, the amount of time the algorithms take to prepare the data can be extensive. Parallelism has significantly decreased processing times of these algorithms. The parallel Trajectory Simplification algorithm maintained high speed-up factors for up to 64 threads, and the parallel Trajectory Clustering algorithm maintained high speed-up factors for up to 16 threads