Date: Tue, 14 Jan 1997 22:05:08 GMT Server: Apache/1.0.3 Content-type: text/html Content-length: 4389 Last-modified: Sun, 05 May 1996 15:43:48 GMT
Scientific computing involves all aspects of tackling problems, and typically includes the formulation of a description of the physics of a problem, the development of a mathematical model of the physical problem, the selection or creation of algorithms to solve the equations in the mathematical model, the implementation on a computer, the analysis and validation of the computer results, and visualization of the results. Carrying out research in this area is necessarily multidisciplinary, calling on advanced skills in areas that span many classical divisions of academia. Modern advances in computational science typically involve teams of researchers with a variety of skills and interests.
Computer scientists are a key component in the mix needed for a good scientific computing team. Although dusting off a numerical analysis program and labeling it ``scientific computing'' can help draw more attention and grant money, scientific computing is not numerical analysis, it is not the analysis of algorithms, it is not high performance computing, and it is not computer graphics, but it instead consists of the combination of all those fields (and others) to craft solution strategies for applied problems.
A good example that can illustrate aspects of computer science and its role in computational science is computational fluid dynamics (CFD). The Scientific Computing Group is closely allied with the Extreme Computing Group and has work connected with the Center for Innovative Computing Applications.