Generic programming for high-performance scientific applications

We present case studies that apply generic programming to the development of high-performance parallel code for solving two archetypal partial differential equations (PDEs). We examine the overall structure of the example scientific codes and consider their generic implementation. With a generic approach it is a straightforward matter to reuse software components from different sources; implementations with components from the Iterative Template Library (ITL), the Matrix Template Library (MTL), Blitz++, A++/P++, and Fortran BLAS are presented. Our newly developed Generic Message Passing library is used for communication. We compare the generic implementations with equivalent implementations developed with alternative libraries and languages and discuss performance as well as software engineering issues. Copyright (c) 2005 John Wiley & Sons, Ltd.