The University of Kentucky will collaborate on a five-year, $10 million National Science Foundation (NSF) initiative, led by the University of Colorado Boulder (CU Boulder), to reimagine cyberinfrastructure user support services and delivery to keep pace with the evolving needs of academic scientific researchers.
The initiative — Multi-tier Assistance, Training and Computational Help (MATCH) — is part of a larger program called Advanced Cyberinfrastructure Coordination Ecosystem: Services and Support (ACCESS). ACCESS is replacing the Extreme Science and Engineering Discovery Environment (XSEDE), which has been the leading program for NSF-funded U.S. cyberinfrastructure for the past 11 years.
According to the NSF, ACCESS will establish a suite of cyberinfrastructure services “to support a broad and diverse set of requirements, researchers and modes from all areas of science and engineering research and education — set up as five independently managed yet tightly cooperative service tracks supported by a coordination office.”
MATCH — one of the five ACCESS tracks — is spearheaded by CU Boulder’s Research Computing group. MATCH proposes a new model for cyberinfrastructure support services that reflects significant changes in the size and composition of the user group community.
“CU Boulder and our MATCH collaborators will lead this groundbreaking effort nationally by leveraging existing tools, interfaces and community experts to assist researchers using NSF-funded cyberinfrastructure to most effectively conduct their research,” said Shelley Knuth, Ph.D., assistant vice chancellor of research computing at CU Boulder and MATCH principal investigator.
MATCH participants include CU Bolder, University of Kentucky, Ohio Supercomputer Center (OSC), Massachusetts Green High Performance Computing Center and University of Southern California Information Sciences Institute.
The MATCH project will include the design and development of the Pegasus workflow-management system, continued development of the Connect.CI portal currently led by the Northeast Cyberteam and design and development of OSC’s Open OnDemand. MATCH will also develop new documentation and training materials that leverage expertise at UK’s Center for Computational Sciences to assist ACCESS users through the allocation process, assist in the movement of big data, and build on recent advances in natural language processing to support interactive interfaces that quickly direct users to solutions.
“The explosive growth and availability of big data combined with powerful data analysis techniques such as AI and machine learning are transforming all areas of research,” said Jim Griffioen, Ph.D., professor of computer science who directs the UK Center for Computational Sciences and co-investigator on MATCH. “The scale and complexity of today’s national research cyberinfrastructure ecosystems require that we rethink and redesign the ways we support and assist users, helping them to make effective use of these systems and reduce the time-to-discovery.”
The MATCH project has three goals:
- Leverage modern information delivery systems and simplify user interfaces to provide cost-effective scaled support to a broader community.
- Leverage experts from the community to develop training materials and instructions that can reduce the user learning curve for an expanding range of systems, applications and computational techniques.
- Employ a matchmaking service that will maintain a database of specialist mentors and student mentees that can be matched with projects that provide the domain-specific expertise needed to leverage ACCESS resources.
The effort will be discussed at the 2022 Practice and Experience in Advanced Research Computing (PEARC) conference July 10-14 in Boston, Massachusetts.