The National Center for Supercomputing Applications (NCSA) is hosting its monthly colloquium series and invites everyone to participate in the November session. 

Talk Title: Investment in Research-Enabling Cyberinfrastructure Services at Universities Yields a Positive Financial Return to Academia and Society

Abstract: At the University of Illinois Urbana-Champaign, many people understand that NCSA led something called XSEDE from 2011 to 2022 and that XSEDE coordinated and supported the use of large-scale NSF-funded supercomputers, data-storage resources, and cloud systems made available to researchers and students throughout the USA. Also, well remembered is that it was a tremendous success. Not widely known is that the first two years were bumpy—through no fault of the project leadership. In 2013, an NSF program officer said, “I am tired of being asked what is NSF’s return on investment in XSEDE and not having an answer. I would like someone to provide me an answer.” This led to a dozen years of analysis of the quantitative, financial return on investment (ROI) for academic cyberinfrastructure. The first major finding of these studies was that XSEDE, as operated by NCSA and its subcontractors, cost the U.S. government less than it would have cost to purchase similar services from commercial sources. We showed that, in general, any sort of cyberinfrastructure operated locally by an institution of higher education is less expensive than purchasing services from the open market (including commercial cloud providers). The exceptions: when an institution needs cloud-only services or does not have the capacity to operate its own facilities. These are dollar-to-dollar comparisons that do not involve any handwaving about ethereal value of research.

A more recent analysis showed that the net financial value of XSEDE’s services to the country exceeds the costs of that program to the taxpayers of the USA. In plain dollars and cents terms, “Value of XSEDE to the public > Cost of XSEDE to the public.” Current research that will be discussed in this talk includes multiple methodologies useful for analyzing ROI for investment in any kind of research-enabling infrastructure.

Never should we shirk from extolling the value of research in terms of improving the quality of human lives and bettering our understanding of our universe. But we can and should also convey that what we do in academia creates a net positive value to our country in hard-nosed business terms.

Bio: Dr. Craig A. Stewart has for decades been a leading innovator and strategist in advanced computing and “cyberinfrastructure.” Stewart arrived in 1981 at Indiana University Bloomington (IU) to pursue a PhD in Ecology and Evolutionary Biology. Starting in 1996, Stewart was responsible for IU’s supercomputers, which in 2005, expanded to responsibility for all of IU’s computing services and systems used in support of research, scholarship, and artistic creation. Stewart was involved in the TeraGrid from 2003 to 2010 and was part of the leadership team of XSEDE from 2011 to 2020. He was the PI on the IU Data Capacitor and the initial PI on the IU Jetstream Cloud system. Stewart has long been involved in collaborations and partnerships with NCSA, including TeraGrid, XSEDE, and HathiTrust. Stewart retired from full-time work at IU in 2020 and remains active in research and teaching as Professor (Adjunct) of Computer Science, Luddy School of Informatics, Computing, and Engineering at Indiana University.