SC19 Proceedings

The International Conference for High Performance Computing, Networking, Storage, and Analysis

Exchanging Best Practices in Supporting Computational and Data-Intensive Research

Authors: Rudolf Eigenmann (University of Delaware), Robert Sinkovits (San Diego Supercomputer Center), Sandra Gesing (University of Notre Dame), Karen Tomko (Ohio Supercomputer Center), Ian Cosden (Princeton University), Shantenu Jha (Rutgers University, Brookhaven National Laboratory)

Abstract: This BOF aims to bring together:

* Researchers developing and using computational and data-intensive (CDI) applications,

* Those who assist these researchers with expertise in CDI technology,

* Developers of tools supporting the development and use of CDI applications.

The BoF will feature brief presentations of a number of panelists who have been involved in creating and providing application support to domain researchers. The audience is then invited to discuss how best practices can be identified and disseminated with the goal of accelerating computational and data-intensive research in the best possible way.

Long Description: The BOF activity is part of an NSF project that aims to advance science by increasing the productivity of researchers who use computational and data-intensive (CDI) methods for pushing the frontiers of science. CDI-based methods have been recognized as high-impact science enablers and investments in computational experts who help domain researchers increase the productivity of CDI applications have been recognized as having a high return. The project builds on these opportunities by creating synergy among teams of computational experts, which are part of many national projects funded by NSF and other agencies. Furthermore, the effort aims to enhance the toolbox of such experts by bringing them together with those who develop the tool environments of CDI applications. Teams of expert assistants have emerged as essential components of many projects conducting CDI research. By leveraging existing conference and meeting schedules and by adding workshops and online forums where appropriate, the proposed work will enable the exchange of best practices and open problems, and the discussion of supporting tool environments among these teams. The BOF will be an element in a series of online and in-person meetings that bring together expert assistants and tool designers for CDI research. The initiative started in March 2019 with monthly online meetings, involving representatives from such projects as XSEDE, CyVerse, the NSF Software Institutes, ACIRef/CaRCC, as well as from several tool development projects. The first two in-person meetings were held at ICS 2019 (International Conference on Supercomputing) and PEARC19 (Practice and Experience in Advanced Research Computing), respectively. The long-term goals of the BOF and the underlying project are to enhance the national research infrastructure in support of CDI research and thus accelerate science. In doing so, the activities will also yield new advances in principles and methodologies for developing CDI applications. Furthermore, by combining expert teams and developers of supporting tools, new principles and methods for tool design will emerge. In this way, a new knowledge base will be obtained for the efficient development of future CDI applications and their enabling support environments. The BOF will help raise awareness of this research network and solicit community feedback.


Back to Birds of a Feather Archive Listing