NICS Remote Data Analysis and Visualization Center

This proposal will be awarded using funds made available by the American Recovery and Reinvestment Act of 2009 (Public Law 111-5), and meets the requirements established in Section 2 of the White House Memorandum entitled, Ensuring Responsible Spending of Recovery Act Funds, dated March 20, 2009. I also affirm, as the cognizant Program Officer, that the proposal does not support projects described in Section 1604 of Division A of the Recovery Act.

In recent years, researchers from a growing range of scientific domains have

experienced a widening gap in their abilities to generate data on complex biological and physical systems and translate these data into scientific discovery. Our data analysis and visualization capabilities have failed to keep pace with advances in both the capacity of our computing infrastructure and the resolution and throughput of our data acquisition systems.

Translating raw data generated by detailed simulation or collected by advanced

instruments into scientific discovery requires a coherent suite of powerful hardware and software capabilities that is fully integrated into our national computational infrastructure. The volume of the data and the complexity of the underlying phenomena require an infrastructure that allows scientists to develop and operate flexible, and the same time dependable, end-to-end data analysis and visualization capabilities that span local and national resources. A 2007 report composed by data analysis, management,

and visualization experts concluded that datasets being produced by experiments and simulations are rapidly outstripping our ability to explore and understand them.

One of the most significant challenges to scientific discovery today is the extraction of knowledge and meaning from this vast array of data and the understanding of the correlations, trends, patterns, and interrelationships among disparate elements from an ever-growing array of data sources. Visualization, data analysis, and knowledge discovery are more vital than ever because they generate the data insight and intuition that enable scientific discovery. Scientific simulation has become the third pillar of

science, supporting frameworks and experimental studies in our understanding of

natural phenomena. As simulations become larger, more numerous, more complex, and as the scientific problems we seek to unlock become more challenging, so does the task of understanding the data generated.

This award presents a data visualization and analysis center, based at the University of Tennessee and coupled with the NSF TeraGrid Kraken supercomputer, that will narrow this gap by bringing together a unique team, proven software technologies, and advanced computing and data-handling capabilities. The center will provide the eyes of the TeraGrid XD, our national cyberinfrastructure, as it evolves to the XD era by

empowering scientists to see and understand very large collections of measured or simulated datasets.

The hardware that undergirds the center is a large shared memory SGI UltraViolet system, able to provide 1,024 processors with 4 TB of shared memory for processing large datasets. No other system in the world has this level of shared memory concurrency for analysis and visualization.

Supported by a large (1 PB) filesystem directly connected to the Kraken supercomputer and the TeraGrid network, this system will provide NSF researchers with an unparalleled data understanding resource.

The team of people comprising the center will draw upon experience in visualization,statistical analysis, workflow delivery, portal and dashboard development, remote access, and application development to provide software resources for large scale data understanding. This team will also be supported by dedicated user assistance, system support, education, and application discovery staff.