CURRENT ISSUES IN COMPUTATIONAL METHODS-ROUNDTABLE: MANAGING MODELING AND SIMULATION RESEARCH AND INNOVATION IN AN APPLICATIONS-DRIVEN ENVIRONMENT

摘要

As computational power has grown, the capabilities in modeling and simulation have expanded accordingly. The maturation of modeling and simulation as a field has naturally placed applications as the driver for projects and funding. Accordingly, increasing emphasis is placed on end-user requirements such as input, output, and features leaving little time left for numerical methods development. Furthermore, high-impact research will necessarily entail investigation in multi-dimensional problems that are difficult to protoype. These conditions combine to create a set of challenges for the future evolution of technologies in modeling and simulation: (a) what is the best mechanism for prototyping innovative methods in multidimensional, multiscale, parallel computational physics application codes within a given project scope, (b) how do we convert prototyped methods into production code, and (c) what impacts does this environment pose for the teaching and mentoring of the next generation of computational scientists and engineers.