SEI Team Lends Know-how to National Science Foundation Project
July 24, 2014—In 2011, Carnegie Mellon University (CMU) joined 16 other organizations selected by the National Science Foundation (NSF) to collaborate on the Extreme Science and Engineering Discovery Environment (XSEDE) project, a five-year, $130 million effort. XSEDE builds on the TeraGrid supercomputing network and provides researchers open access to state-of-the-art computational tools and digital resources.
As part of the CMU team, the SEI has played a significant role. "The SEI has been an enormously beneficial partner to the project in helping XSEDE to understand formal software and systems engineering practice and to adapt that practice to the very nontraditional context of XSEDE," said John Towns, XSEDE principal investigator and project director.
To bring formal practice to the NSF community, the SEI developed a twofold approach, first establishing sound engineering practices to enable systematic, measured improvement in products and services. It is also introducing novel engineering practices to address unique challenges arising from XSEDE's status as a highly distributed NSF/Office of Cyberinfrastructure (OCI) sociotechnical ecosystem.
SEI staff members Felix Bachmann, Kurt Wallnau, Linda Northrop, Michael Konrad, Scott Hissam, and Rhonda Brown worked with XSEDE to refine and document software engineering processes to enable effective iterative and incremental development. They further collaborated to define and institutionalize use-case development and active design review—a technique that advances effective communication during software design. The team also conducted an initial study to identify engineering practices for engineering ecosystems.
Bachmann explains, "XSEDE exemplifies a software development ecosystem that many development organizations face or are about to embrace, where social science plays a major role alongside software engineering practices."
Plans are in place to build on XSEDE's success in adopting architecture-centric engineering and other SEI approaches. The goal is to transition them to practices more viable for conducting engineering at ecosystem scale.
For more information about the XSEDE project, please visit www.xsede.org/.
This article originally appeared in the 2013 SEI Year in Review