|
|
|
| News |
| Advanced Scientific Computing Advisory Committee |
| February 2007 ASCAC Meeting Highlights |
| The primary business at the DOE Office of Science (SC) Advanced Scientific Computing Advisory Committee (ASCAC) 27-28 FEB 2007 meeting was to review and accept the Petascale Metrics Panel Report as an ASCAC Report. This Report, presented to the ASCAC by Petascale Metrics Panel Chair, ASCAC member Gordon Bell, provides the ASCAC response to the March 2006 Charge from Dr. Raymond Orbach, Under Secretary of Energy for Science, “... to weigh and review the approach to performance measurement and assessment at [the Department of Energy Office of Science High Performance Computing Centers and related projects], the appropriateness and comprehensiveness of the measures, the [computational science component] of the science accomplishments and their effects on the Office of Science’s science programs, ... [and] the role and computational needs over the next 3–5 years” (“Highlights of the ASCAC March 2006 Meeting,” SciDAC Review, Fall 2006, p4). |
| The Petascale Metrics Panel Report describes four aspects of science accomplishment that the Panel recommends that the ASCR High Performance Computing Centers should report and track “to assist in the measurement of scientific output from its projects: Publications, Code & Datasets, People, and Technology Transfer…; Project Milestones versus the Proposal’s Project Plan …; Exploiting Parallelism and/or Improved Efficiency: aka Code Improvement …; and Break-throughs, an immeasurable goal. The Panel could not identify metrics or any method that could be used to anticipate discoveries that occur on the leading edge of fundamental science.” |
| The ASCAC unanimously endorsed the Petascale Metrics Panel Report, noting that the difficult problem of measuring scientific understanding and accomplishment needs additional study. As observed by ASCAC Chair, Dr. Jill Dahlburg, in a paper titled “The S&T Innovation Conundrum,” there are at least two distinct phases in the history of a major scientific accomplishment. “There is an early, searching phase that is evocative of prospecting. It is characterized by a few discrete but high-impact events. There is little functional capability produced during this phase and the individuals contributing to the discrete events, while generally sure that they are involved in profoundly exciting research and technology, often have no idea what the ultimate functional capability will be. Continuing with the analogy, this early phase is followed by a later, more predictable phase that is much like mining. This latter phase seems to be dominated by continuous improvement in functional capability as characterized by a larger number of lower-impact innovations than occurred in the early phase. During the mining phase, the capability produced can usually be related to the funding applied and to the inherent potential of the technology being exploited, resulting in desirable features such as measurable and predictable return on investment.” Metrics, while in some sense essential for the mining phase, are generally not applicable to the quantification of success in the prospecting phase of a major scientific breakthrough. |
| Dr. Ellen Stechel, ASCAC member and Chair of the ASCAC Networking Subcommitee, provided an update about that status of the report to address Dr. Orbach’s 10 MAR 2006 Charge to examine the role and efficiency of networking and networking research within the DOE SC. In the report, which is due in November 2007, the Subcommittee hopes to produce useful guidance on networking research and the priority of networking needs.
|
On February 28, two new Charges from Dr. Orbach were provided to the ASCAC. The first is a Committee of Visitors (COV) Review of SciDAC, which will be led by ASCAC member Dr. Robert Voigt. The second is a Joint Committee with the Biological and Environmental Research Advisory Committee (BERAC) to examine the issue of computational models for Genomes to Life (GTL). ASCAC member Dr. Rick Stevens has agreed to serve as ASCAC Co-Chair of this Joint Committee, the report for which is due at the August 2007 ASCAC meeting.
|
Contributor:
Written by Dr. Jill Dahlburg of the Naval Research Laboratory, these highlights were largely summarized from the ASCAC Petascale Metrics Report, The S&T Innovation Conundrum report, and the February 2007 ASCAC Meeting Minutes (Fred O’Hara, recording secretary)
|
