A project designed to evolve | National Center for Supercomputing Applications at the University of Illinois

A project designed to evolve

09.06.11 -

What can be done to provide new and expanded infrastructure and services that substantively improve the ability of a broad range of researchers to conduct their work? How can we do this and account for the fact that research no longer typically happens in the context of a single investigator on a single campus? These are just a couple of the questions we asked ourselves while developing the proposal for XSEDE, the Extreme Science and Engineering Discovery Environment.

We recognized that today's investigators collaborate across institutional and geographic boundaries. To be successful, they need dispersed resources—including instruments, data stores, high-performance computers, services, and collaboration tools—at their disposal. The intent with XSEDE is to create the environment in which all of these resources and services are available.

XSEDE is a follow-on project to the TeraGrid, although it is very different in nature in many ways. For more than 10 years, TeraGrid provided access primarily to high-performance computing resources and services. The high-performance computing resources are still critical in general for the success of a lot of the research work, but the focus of the project is now on researcher productivity. For some researchers, computing resources are not as relevant as access to data, access to collaboration tools, and access to instruments. So under the XSEDE project, we'll build on the TeraGrid accomplishments and the lessons that were learned to broaden the array of resources and expand the set of services to encompass more than just HPC. These resources and services will be accessible to researchers around the country and their collaborators internationally. This means shared access to campus-level resources, interoperability with other cyberinfrastructures, access to instrument data, and a whole host of services.

The key element that we are targeting with XSEDE is the productivity of researchers in conducting their day-to-day research. To establish such an environment—a distributed environment—with the resources and services accessible to them, is really the cornerstone of the project.

Why?

As I mentioned in my opening paragraph, research no longer typically happens in the context of a single investigator. Modern researchers are typically collaborating with others at other institutions. They need access not only to resources they have direct control over locally, but also to resources collaborators have, they need access to data sources that are managed by other projects, they need access to instruments, and they need access to computing resources. The intent with XSEDE is to create an environment in which all of these resources and services are available, whether they are formally operated by XSEDE or not.

Those of us involved in the development of the XSEDE project over the last three years have invested a lot of time and a lot of effort in developing a vision for not just providing cycles to researchers, but for establishing a cyberinfrastructure ecosystem in which researchers and educators can be much more productive, and can begin to develop new capabilities.

A key element of our architectural plans is that the project is designed to evolve. We have not designed a system that will be static for the next five or more years. It's specifically designed and has processes put in place to evolve the environment and the system over time as new technologies emerge, as new requirements are understood from the user community, as new communities engage with the project.

What I find most exciting is the potential to include many more disciplinary areas and a lot more researchers who might not have had easy access to the resources in the past. New disciplines, new users, and being able to increase productivity—that's what makes the XSEDE project really exciting.

John Towns
PI and Project Director, XSEDE
Director, Persistent Infrastructure
Directorate, NCSA