Computational Smart Grid and DAM Queue

Abstract

The layout of the Smart Computational Grid is an integral part of our information services infrastructure. In this paper, an anticipation of the architectural model of Smart Computational Grid is derived, which provides petabytes storage, tera-byte processing speed, and Gigabyte data transmission, where many supercomputers forms node with parallel data transmission. Impulsive demands posed by unexpected flows of data from users for complex computations, however, is exceedingly challenging. The paper specifies the design and implementation of a Dynamic Allocation of Memory (DAM) algorithm. DAM algorithm identifies the design of memory at supercomputing node, for ergodic functioning and for ensuring that no data overflow, using the methodology of queuing theory. The overall performance of the DAM algorithm is shown using MM1 and GG1 queue models and results are compared accordingly