Stony Brook Center for Computational Science:
An Interim Operational Plan

2007

James Glimm
Scientific Operations Director

Formation of a Stony Brook Center for Computational Science (SBCCS)
With the arrival of a new and very powerful supercomputer in early 2007, there is an opportunity to raise the level of computational science at Stony Brook. To achieve these goals we have established a Stony Brook Center for Computational Science (SBCCS), in parallel with the Brookhaven Computational Science Center (CSC). The Stony Brook Center will have a mission to develop computational science at Stony Brook. Each center has a secondary mission to cooperate with the other and to support the mission of the other, while they will jointly share outreach missions to a larger New York state community. Thus the two Centers will be closely linked, with many shared members and activities. The exact nature of the linkage of the Centers and their possible merger into a single Center can be addressed later.

The SBCCS will thus be built around core faculty and scientists at Stony Brook It will, in cooperation with CSC, provide a focal point for the community of users of the new supercomputer. It will accelerate the acquisition of supercomputing expertise for scientists at the two institutions interested in taking advantage of the very powerful resource. It will promote a change in the culture of science regionally, to strengthen the excellence in computation and specifically in supercomputing, as supplements to existing strengths in theory and experiment. As a specific task, the Center will compile and post on the www a list of supercomputing related courses offered at Stony Brook and at neighboring institutions, and it will identify gaps if any in this set of offerings, from the perspective of an incoming graduate student. Since we anticipate a cadre of supercomputer users regionally, the mission of the Center will extend to support the development of computational science regionally, and, in cooperation with CSC, to provide a kind of intellectual home for regional computational scientists.

New resources for the SBCCS Center consist of six new Stony Brook faculty lines.

To promote the cross-disciplinary collaborations typical of successful high level scientific computing, the positions have been be filled jointly between departments and/or the two institutions.

All Stony Brook computational scientists will be encouraged to participate in the Center, which will present opportunities to share and trade expertise in supercomputing lore and technologies. In fact, such an informal communication network is essential for both new and advanced supercomputing personnel, as many important facts, tools, tricks of the trade are not readily learned in other ways. For this purpose we request space for Center member offices and meeting rooms.

On consultation with regional institutions having an interest in supercomputing, the NYCCS will coordinate regional activities as well. In view of the (not too long) travel distances involved, we judge that day long topical conferences, with participation open to PhD students and junior faculty/scientists would be popular, and would serve to build a community of shared supercomputing expertise and lore, centered around the users themselves. Tutorials and workshops on supercomputing software and other computing issues will be offered. Some of the user support personnel for the operations will probably consist of Stony Brook graduate students, in the form of staffing a help desk. This help desk and associated student offices should be located within the space assigned to the new Center.

The SBCCS will also be charged with outreach to potential supercomputing users who are not in an academic setting. Here industrial use of supercomputing comes to mind, but in addition, there could well be other users, such as from the financial industry. In addition to the major high tech industries in New York State, we will seek out smaller high tech firms. These firms are often dependent on government research contracts, and we hope that arranging for their access to the advanced supercomputing will be an aid to their overall business position. Here, a proactive strategy will be needed, with a more active and aggressive form of support for users who may be less prepared to invest in the acquisition of new supercomputing technology, as a demonstration of its capabilities in the context of their problems. To carry out the industrial outreach mission, an additional budget will be required. Industrial outreach will be developed in consultation with Stony Brook Vice President for Economic Development Yacov Shamash.

Finally, the Center will be charged with providing a focal point for national and international visibility for the combined Stony Brook-BNL (and regional, if others wish) computational science effort, including presentations as a share of the BNL booth at the annual supercomputing meeting.



Another important goal of the Center will be to establish partnerships with corporate partners throughout the State of NY and the tri-state area.