Computational Science

Last updated: February, 2009

A computational scientist is a person skilled in scientific computing. This person is usually a scientist, an engineer, or an applied mathematician who applies the IT infrastructure (computers, storage, networks and software) in different ways to advance the state-of-the-art in their respective applied disciplines such as physics, chemistry or engineering. Scientific computing has increasingly also impacted other areas including economics, and medicine. A computational scientist works on the boundaries of science and information technology. In the early days of computational science this meant working mainly with super computers. Nowadays complex scientific problems do not only need huge amounts of computing power, they also need to be able to store, transfer and manage, huge amounts (Peta to Exabytes) of generated data. Example of (big) scientific experiments are the LHC experiments at CERN. Such experiments require massive amount of computing resources (more than just one super computer) as well as storage and network resources. Ideally a computational scientist is a domain expert, skilled software engineer and experienced information technologist all in one!

Computational science is the field of study using IT infrastructure to analyze and solve scientific, social scientific and engineering problems. In practical use, it is typically the application of (large scale) computer simulation and other forms of computation to problems in various scientific disciplines. Traditionally computational science focused on application development on supercomputers, however during the last decades distributed off the shelf and data intensive computing meant that storage, curation and transfer of extremely large data sets have become important. The field is distinct from computer science (the mathematical study of computation, computers and information processing). It is also different from theory and experiment which are the traditional forms of science and engineering. Within science, computational science is sometimes called the third pillar next to theory and experiment.

The interdisciplinary field of computational science also penetrated the (business) domain of finance and banking. Financial institutions nowadays rely in part on computational scientists to design, optimize and implement models as well has deploying and executing the applications (simulations) on large scale IT infrastructures. As (quantitative) financial institutions came to rely more on sophisticated financial models to develop and price their products and raw computing power became more affordable, a new type of department emerged between the back and the front office : 'the middle office' . The middle office, the third pillar between front and back office is typically the place within an organization where computational scientists work.

Computational scientists work on the boundaries of information technology and science (or business) and therefore need to have an extraordinary skill set, which typically includes: working in a multidisciplinary environment, ability to translate requirements and models to software, able to communicate with domain experts as well as with IT personnel, optimize models and design/run applications on large IT infrastructures, design and develop software (ad hoc or within larger projects).

Governments and companies increasingly use models to predict, correlate and gain insight into questions of climate science, the origins of our universe, medicine and the financial markets. It is therefore important for countries and corporations to invest in computational science for reasons of (economic) security, prosperity and competitiveness. In 2005 the Information Technology Adivsory Committee produced the report: "Computational Science: Ensuring America's Competitiveness", while in 2004 the Office of Science produced the report: "Data Management Challenge". Both reports highlight the importance of computational science, as well as a shift away from pure raw computing power to more data driven applications.