Computational Science and the Future of Computing Research

• Authors: John R. Rice
• Venue: IEEE Computational Science & Engineering
• Year: 1995

This is the first of IEEE CS&E's articles in a new category, "Ideas, Issues, and Trends," represented by an icon showing the ripple effect of commentary and observations that may from time to time be thrown into the pond of the professional community of computational scientists and engineers. In this issue John Rice, an old hand in computer science and numerical analysis and former chair of the Computing Research Association, offers his perspective on a topic of vital interest to researchers.Though policy details shift with politics, it seems that public pressure to use tax money in an efficient--perhaps frugal--way is likely to affect research funding for some time to come, if not indefinitely. This change is seen in such manifestations as the emphasis on National Challenges that has captured some of the fanfare and research dollars that the Grand Challenges once held. The catchword for the new policy trend is "strategic" research. Strategic, that is, in terms of a focus on the funding of work that directly furthers a set of goals defined with the overall benefit of society in mind. Some strategic research is basic, some applied.Without stating a value judgment on whether the "strategic" emphasis is good, Rice makes the point that some form of it is most likely inevitable. Regardless, he argues that computer scientists, at least, are in an enviable position: with a little creativity, almost all computing research can be legitimately tied to the defined strategic goals. One of the best vehicles for doing this is computational science and engineering. Rice adduces four examples to make his point. Automatic control of blast furnaces, interactive computational environments for the design and optimization of mechanical parts, virtual reality, and robots all qualify as strategic, and are all part of or directly connected to CSE. The advances required to make these techologies viable can keep researchers in just about all subfields of computer science busy for a long time, if they make the connections clear. The writing is on the wall, Rice says. Some may not like it, but woe to the researcher who ignores the message.