Skip to main content
Email Print Share
NSF & Congress


Dr. Shirley Malcom
National Science Board

Before the House Science Committee
Subcommittee on Basic Research
April 9, 1997

Chairman Schiff and members of the committee, I am Dr. Shirley Malcom and I am pleased to have the opportunity to testify today. I am here not only as a member of the National Science Board, but also as a member of its Executive Committee. I actively participated in the entire process of the Board's consideration, discussion, and decisions related to the staff recommendations for awards in connection with the Partnerships for Advanced Computational Infrastructure, or PACI program. I am pleased to provide this committee with insights into the processes and workings of the National Science Board in considering the proposals that come to it.

As you are aware, the NSF Act authorizes the Director to make awards and agreements only with the approval of the Board or under authority delegated by the Board. In May 1994, the Board delegated to the Director the authority to approve awards up to $3 million in one year and $15 million over five years. Above that threshold, the NSB reviews and acts directly on the proposals. By statute, Board actions are taken by the full NSB or, if the Board is not able to act, by the Executive Committee.

The NSF staff process includes reviews at higher management levels, including, for packages that come to the NSB, a review by the Director's Review Board, or DRB. Once packages are approved by the DRB, they come to the Board and are assigned to one of our committees for in-depth consideration, then presented to the Board for action.

In the case of the PACI awards, the extraordinary number of real and perceived conflicts of interest existed in connection with these proposals as a result of the call for extended partnerships. Consequently, the Board was left with only eight confirmed members out of sixteen and three consultants out of a possible nine able to participate in the deliberations. As a result, the staff recommendations were presented directly to the Board to insure that the packages were subjected to a thorough review and discussion by as large a number of Board members as possible.

NSF staff provided a presentation to the Board on the proposal packages, after which a lead reviewer and a secondary reviewer provided detailed reviews, commented on issues for which more information was needed, and made comments as to the fairness of the procedures and the appropriateness of the recommendations from staff. The lead reviewers and Board members all participated in the discussions and asked hard questions of the staff. The staff provided additional information and answers that satisfied the NSB members and consultants participating in the discussion.

Board discussion considered issues such as assurance that a diverse set of computer architectures were used by the partnerships, the management of large, far flung partnerships, the effects of budget reductions on the overall coherence of the proposed projects, and the transition process to the new program and its impact on the user community. Members of the Board were made aware in advance of Cornell University's concerns about the composition of the review panels and the uniformity of information provided to competitors, as well as the position of the Pittsburgh Supercomputing Center. We scrutinized the proposals very carefully, reflected on what we heard, and concluded that the staff recommendations should be supported fully.

I can assure you that we asked hard, direct questions and demanded satisfactory answers. We reviewed reports from the merit review process and assured ourselves that the review process was thorough, fair, and consistent with NSF's high standards. The discussions among our members and consultants overwhelmingly affirmed the staff's recommendation. This strong sense of support provided a firm basis for the Executive Committee to take action on the PACI proposals, and we did so by a vote of four to zero to approve the staff recommendations.

Of course, it is more desirable that the full Board be able to act on projects of this scope, magnitude and importance. However, the Senate has not yet acted to confirm the nine NSB nominees, leaving the Board with only 16 confirmed members. At this meeting, we were unable to achieve participation of thirteen voting members to constitute a quorum.

The Executive Committee acted on behalf of the Board, exercising its judgment within the statutory authority and with the full and firm support of our colleagues. If the NSB had not acted last week, the result would have been a delay in beginning negotiations with the PACI awardees as well delays in making plans for phasing out existing centers. In addition, the research community would have been subjected to needless uncertainty. Such a delay would have been inconsistent with the effective exercise of our responsibility -- to the institutions involved and to the community of users of supercomputing facilities -- for supporting the U.S. scientific and industrial leadership in supercomputing. Thank you for the opportunity to provide additional information to the Committee on this important decision. I would be pleased to respond to any questions that you might have.

See also: Hearing Summary.