Interactive Visualization Of Multivariate Spatiotemporal Data Via GP-GPU
Data visualization is a critical component of all business management processes. To this end, JavaScript-based web dashboard systems are often employed to handle such data visualization and analysis tasks, but generally demonstrate poor visualization capabilities when scaled to realistic, large data sets (millions of records). To overcome this bottleneck, research at the University of Louisiana at Lafayette have designed and prototyped a novel approach that moves the visualization data and processing off of the CPU and onto the graphics hardware (GPU) using WebGL. This approach can handle datasets with more than a million records at prompt interactive rates and demonstrates feasibility as a web browser environment for interactive, real-time information visualization tasks. This invention offers substantial performance advancements over standard JavaScript-based web dashboard systems, and enhances the abilities of visualization and analysis packages for a wide variety of application areas.
