Validating the Need to Validate Code

Abstract

Douglass Post and Lawrence Votta’s article admirably stresses the need for validation of computation. If I understand their main point correctly, it’s that computation is science, not mathematics.The article motivates me to contrast two fields I work between that use very different paradigms to describe the electrodiffusion of charge. In computational electronics, 1,2 1. C. Jacoboni, P. Lugli, The Monte Carlo Method for Semiconductor Device Simulation, Springer-Verlag, New York (1989) https://doi.org/10.1007/978-3-7091-6963-6.2. S. Selberherr, Analysis and Simulation of Semiconductor Devices, Springer-Verlag, New York (1984) https://doi.org/10.1007/978-3-7091-8752-4. the electric field is traditionally calculated by solving Poisson’s equation with far-field boundary conditions but at relatively low resolution. Poisson is solved anew whenever charges move. Computational chemistry, starting more or less with computer simulation of fluids 3 3. M. P. Allen, D. J. Tildesley, Computer Simulation of Liquids, Oxford U. Press, New York (1987). and computational biology, 4 4. T. Schlick, Molecular Modeling and Simulation, Springer-Verlag, New York (2002). computes the electric field at high resolution and does not deal clearly with far-field boundary conditions. Electrodiffusion, which has been at the center of electro- and physical chemistry since Michael Faraday’s time, is also at the center of electronics, where it describes the movement of charge in semiconductors and most of our digital devices. Electrodiffusion is no less important in biophysics, where it is responsible for the electrical properties of cells and tissues, and controls many biological functions.It seems to me that these different treatments of similar physics are distinct and unlikely to be equally precise under all conditions. I hope the article helps motivate workers in each tradition to discuss other treatments beyond their own and try to understand the differences. I hope such workers can determine the conditions under which each treatment is accurate. That way we may learn to use each computational tradition of electrodiffusion only in appropriate situations.REFERENCESSection:ChooseTop of pageREFERENCES <<CITING ARTICLES1. C. Jacoboni, P. Lugli, The Monte Carlo Method for Semiconductor Device Simulation, Springer-Verlag, New York (1989) https://doi.org/10.1007/978-3-7091-6963-6. Google ScholarCrossref2. S. Selberherr, Analysis and Simulation of Semiconductor Devices, Springer-Verlag, New York (1984) https://doi.org/10.1007/978-3-7091-8752-4. Google ScholarCrossref3. M. P. Allen, D. J. Tildesley, Computer Simulation of Liquids, Oxford U. Press, New York (1987). Google Scholar4. T. Schlick, Molecular Modeling and Simulation, Springer-Verlag, New York (2002). Google ScholarCrossref© 2005 American Institute of Physics.