The past few years have been witness to a growing interest in grid computing as a means of realizing the need for scientific computing capabilities. The grid environment has one severe setback i.e. the requirement of a very steep learning curve. In this work we propose a model, the SMCG (Simple Model for Computational Grid), which is much simpler in comparison to the conventional grid model, for a robust environment that caters to the needs of scientific computation. The major goal of this model is the ease of using this environment for creating applications as well as for programming. This work was motivated by the latest trend in integrating grid computing and cluster computing.