Scientific Computing

Einstein and Pythagoras fighting over C2

What is science? The English word science is derived from the Latin word scientia for knowledge. A simple definition of science is the activity concerned with the systematic acquisition of knowledge about the universe. See the definition of science in the Cambridge and Oxford dictionaries.

How do we acquire knowledge? There are three ways for acquiring scientific knowledge, known as the three pillars (or legs) of the scientific method: 1) observations and experimental measurements often obtained by instruments; 2) theory and inductive hypotheses often formulated in the language of mathematics; and 3) computation using computational algorithms and computer simulations.

Scientific computing (or computational science) constitutes the third pillar of the scientific method. It uses computational algorithms to represent mathematical models of physical phenomena and solve them on computers. The computer outputs will then be used to predict the future or to reconstruct the past and hence to aquire knowledge. Computational science has been successful in a wide range of science and engineering fields, from medicine to climate prediction.

According to Prof. J. Tinsley Oden: “the emergence of computational science is one of the most important developments in human history, a development that will revolutionize virtually every aspect of science in the future.”

Beskow at KTH

High performance scientific computing refers to the systematic practice of aggregating computational algorithms and parallel computing to efficiently and reliably solve large problems in science and engineering. It delivers a much higher performance than one could get out of a typical desktop computer or workstation, especially when applied to systems that function above a teraflop or \(10^{12}\) floating-point operations per second.

Predictive computational science is an emerging, interdisciplinary field concerned with assessing the predictability of mathematical and computational tools, particularly in the presence of inevitable uncertainty and limited information. It is based on a systematic incorporation of all three pillars of the scientific method to acquire knowledge. Predictive computational science involves methods and principles of a variety of disciplines, such as mathematics, statistics, computer science, natural sciences, engineering, medicine, uncertainty quantification, and parallel computing.