Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
APPENDIX A STATEMENT OF TASK The task of the committee, as originally conceived, is described in the following excerpt from the Notice of Financial Assistance Award from the U.S. Department of Energy and the Research Grant from the Office of Naval Research to the National Academy of Sciences: The committee will carry out Phase I of a study to define criteria for the development and performance measurement of large-scale computers for scientific and engineering applications. Phase I will synthesize a collective judgment on the best study plan for the substantive aspects of the supercomputer criteria question. (Phase II will carry out the substantive study, provided that the plan developed in Phase I appears promising.) The committee will perform the following tasks: 1. Select important, computationally intensive problems related to large-scale scientific and engineering applications to constitute the subject matter of the study. 2. Examine the computational characteristics of these problems together with the capabilities of existing and prospective large-scale computers so as to develop a study plan that will assess the prospects of improved criteria for large-scale computer development and performance measurement, the nature of such improvements, and the anticipated consequences of these improvements. 3. Document the results of Phase T The committee will consist of approximately ten recognized experts in the following areas: numerical processing, symbolic processing, algorithm design, computer architecture, computational physics, computational mechanics, and computer performance measurement. 53
54 During the course of the study, a different emphasis of the Phase I task was recommended by the committee and accepted by the representatives of the sponsors. In particular, Phase I undertook to describe the problem of performance measurement in some depth and to point broadly to desirable directions of inquiry. However it was believed that the detailed definition of a research program to resolve the problem could be done more productively by the supercomputer community itself than by this, or any other, committee. Consequently no study plan is proposed for Phase II. Instead a different need is seen--a forum to track, assess, and disseminate progress, so that research results get promptly translated into practice.