Program Development for Computational Grids Using Skeletons and Peformance Prediction

We address the challenging problem of algorithm and program design for the Computational Grid by providing the application user with a set of high-level, parameterized components called skeletons. We describe a Java-based Grid programming system in which algorithms are composed of skeletons and the computational resources for executing individual skeletons are chosen using performance prediction. The advantage of our approach is that skeletons are reusable for dierent applications and that skeletons' implementations can be tuned to particular machines. The focus of this paper is on predicting performance for Grid applications constructed using skeletons