HPDC

It is our great pleasure to welcome you to the first installment of the Science of Cyberinfrastructure: Research, Applications, Experience and Models -- SCREAM-15.

There is a need for comprehensive, balanced and flexible distributed cyberinfrastructure (DCI) in support of science and engineering applications. A fundamental technical challenge is to support a broad range of application usage scenarios and modalities on a range of platforms with varying performance. The current generation of DCI has resulted in important scientific results as well as advances in the state-of-practice of delivering DCI as services to the user community, broadly defined. However, a complete conceptual framework for DCI design principles remains prominent by its absence. This missing framework prevents an objective assessment of important technical as well as policy considerations.

The SCREAM workshop generally aims to address this gap, and specifically aims to understand, through a combination of experience, application requirements, and conceptual models, how to best to create a conceptual framework for the objective design and assessment of distributed cyberinfrastructure. In other words, it aims to build toward the science of cyberinfrastructure upon what has hitherto been a purely empirical approach to cyberinfrastructure design and practice. The SCREAM Workshop is interested in all areas that will further this objective, in particular the interaction of multiple cyberinfrastructure components and systems (distributed computing, broadly defined), including academic and commercial production systems and research testbeds.

Significant effort has been invested in the delivery and practice of DCI with different objectives and varying capabilities, and existing (Open Science Grid, XSEDE, GENI, EGI, PRACE, DAS-n) and previous offerings have yielded valuable information. Enough experience now exists to reflect on what has worked and why, and why some approaches have failed. Thus, we believe the time is appropriate to build upon these lessons towards a next generation of DCI that is designed and architected for well-defined usage modes, performance and capabilities.

Although primarily targeted towards computing scientists, we believe this workshop will have an impact beyond the computing specialist in light of the fact that production cyberinfrastructure impacts the effectiveness of other science & engineering endeavors. This workshop welcomed technical contributions delivered via research-based results, experience papers, and vision papers. Understanding the principles and science of cyberinfrastructure has impact beyond just the computing aspects.

The call for papers attracted 12 submissions with authors from Asia, Canada, Australia, Europe, and the United States. The papers received an average of 4.9 reviews each. As a new workshop seeking to promote discussion, we were generous in our decisions, accepting 8 of the papers, for a 66.7% acceptance rate.

We also encourage attendees to attend the keynote and invited talk presentations, and the closing panel. These valuable and insightful talks can and will guide us to a better understanding of the future:

• Keynote: What can science cyberinfrastructure learn from commercial IT?, Ian Foster, University of Chicago & Argonne National Laboratory

• Invited Talk: Revisiting the Anatomy and Physiology of the Grid, Chris Mattmann, NASA Jet Propulsion Laboratory

• Panel: Designing Distributed Computing Infrastructure for Seamless Multi-site Execution,