Extending High-Performance Computing Beyond its Traditional User Communities

Extending High-Performance Computing Beyond its Traditional User Communities

 

Co-located with the 8th IEEE International Conference on eScience,
Chicago, USA
October 8, 2012

Call for Papers:

Historically, high-performance computing (HPC) has enabled computationally intensive simulations performed in batch mode on a small number of standalone supercomputers, shared among users selected for their computing skills as much as for expertise in their own disciplines. There has been a sustained effort over the past decade to broaden this model by deploying a wider variety of HPC systems tied into emerging national and global cyber-infrastructure (CI), yet only a small fraction of the resources fielded by HPC-based CI programs such as the eXtreme Science and Engineering Discovery Environment (XSEDE) is currently used by people who are not members of communities that have used supercomputing centers since the 1980’s.

Given the digital instruments and methods that are revolutionizing biological, environmental, and physical sciences, as well as the promise of important benefits to social sciences and the arts and humanities, XSEDE is undertaking a proactive effort to work with members of these communities to identify barriers and to develop projects that show how to effectively overcome them.

In this context, the goal of the proposed workshop is to discuss examples of successful projects as well as barriers and practical approaches to overcoming them. After the presentation of selected papers, there will be a discussion among all the participants. The desired outcome is an improved understanding of actions that should be taken by the various stakeholders in order to enable a wide spectrum of practitioners to use HPC resources as part of their work and data flows, and to establish an informal network of people and communities interested in this outcome.

We invite papers that describe projects that have already used HPC systems, or whose requirements analysis indicates a need for HPC systems as part of the infrastructure for their implementation – if specific topics of concern are satisfactorily addressed.

Disciplines of study include, but are not limited to:

  • Genomics and bioinformatics
  • Social, behavioral and economic sciences
  • Digital humanities
  • Public Health
  • Citizen science
  • Computational linguistics
  • Machine learning
  • Digital arts

Topics of concern include, but are not limited to:

  • Campus, Cloud and HPC resources: tradeoffs and interoperation
  • Security and privacy of HPC environments
  • Data management, integration and visualization from Lab to HPC and back
  • Parallelization of compute- or data-intensive tasks
  • Programming paradigms, tools and programming environments
  • Access to and scheduling of HPC environments
  • Community portals and gateways
  • Workflow management and remote collaboration
  • System level support for workflows that include HPC
  • Fault-tolerance of distributed applications
  • Scalability of infrastructures and applications
  • Training and education of current and future practitioners

Organizers

Ralph Roskies, Pittsburgh Supercomputing Center, USA
Nancy Wilkins-Diehr, San Diego Supercomputer Center, USA
Sergiu Sanielevici, Pittsburgh Supercomputing Center, USA
Philip Blood, Pittsburgh Supercomputing Center, USA

Program Committee

Kevin Franklin, University of Illinois, USA
Jane Hunter, University of Queensland, Australia
James Taylor, Emory University, USA
Kenneth Judd, Stanford University, USA
John Grefenstette, University of Pittsburgh, USA
Shaowen Wang, University of Illinois, USA
Robert Sinkovits, San Diego Supercomputer Center, USA
Jeffrey Gardner, University of Washington, USA
Thomas Uram,Argonne National Laboratory, USA

24

No comments