Recognizing the need to educate science and engineering students in the methods of high performance computing, a growing number of universities are introducing computational science and engineering courses and programs. This talk will describe several representative computational science programs at the graduate and undergraduate levels and discuss trends in courses and curricula.
While computer science departments have been well established in our universities for many years, with a relatively well defined curriculum, computational science remains an interdisciplinary subject with an emerging curriculum, requiring the development of new courses not offered by traditional departments. Courses in four defined areas of computational science (applications, mathematics, computer science, and visualization) will be discussed.
The twenty-two graduate computational science programs included in this survey include eight that offer degrees in computational science, ten that offer specialty degrees (usually in the form of a minor or certificate) and four that offer graduate courses in computational science without any degree. Thus the trend in graduate computational science education is the specialty degree in which the student is part of a traditional science or engineering department (the "home" department) and meets additional requirements in computational science. These can include computational science and engineering courses and a computationally based thesis related to the home department's discipline. In addition, the student's examination committee may be required to include computational scientists. Upon graduation, the student receives an annotated degree, e.g., "PhD in Chemical Engineering and Computational Science."
The fifteen undergraduate computational science programs included in this survey include two that offer specialty degrees (minor or certificate) and thirteen that offer undergraduate courses in computational science without any degree. There are no undergraduate computational science departments or stand-alone degree programs. The undergraduate trend is toward interdisciplinary computational science courses supplementing existing course offerings.
Most of the current graduate and undergraduate computational science programs are at large research universities that own supercomputers or that have close associations with state or national centers. An issue for the near future is to extend computational science education into four-year liberal arts colleges and small universities. Network access to supercomputers and the availability of high-performance workstations and relatively low-cost departmental supercomputers provide opportunities for these schools to introduce computational science into their curricula.
Chuck Swanson email: cds@cray.com University Marketing phone: (612) 683-3425 Cray Research, Inc. fax: (612) 683-3599