International Journal of Engineering Inventions e-ISSN: 2278-7461, p-ISSN: 2319-6491 Volume 5, Issue 9 [Oct. 2016] PP: 25-41 www.ijeijournal.com Page | 25 COMPUTATIONAL GRIDS: CHALLENGES AND NEW TRENDS Napoleão Póvoa Ribeiro Filho 1,2 , Fernando Jorge Ebrahim Lima e Silva 1,2 , Leandro Guimarães Garcia 1,3 , Cleórbete Santos 1 1 Universidade Federal do Tocantins, Mestrado em Modelagem Computacional de Sistemas, Av. NS 15, 109 N, Palmas, TO 77001-090, Brasil 2 Instituto Federal do Tocantins, Endereço: AE 310 Sul, Av. LO 5, s/nº - Centro, Palmas, TO 77021-090, Brasil 3 Universidade Federal do Tocantins, Faculdade de Medicina, Av. NS 15, 109 N, Palmas, TO 77001-090, Brasil ABSTRACT: This paper presents a literature review on computational grids, analyzing their use by various organizations and companies. It aims to identify the main challenges found in their usage, its trends and application. This computing model, although not new, has a considerable amount of users that operate in several areas. As a model that allows the use of heterogeneous computing environments, computational grids have many issues to consider for their use. This decentralized processing also allows their use to be adapted and improved by newer technologies. The findings of this study may provide insights for researchers and users on architecture improvements and computational grids usage possibilities. I.INTRODUCTION The Internet growth, powerful computers availability, high-speed computer networks just as the low cost of components raw materials, is changing the way that scientists and engineers do computing and also the way society in general manages information and services [23]. In 90’s, inspi red by the availability of long-range high-speed computers and challenged by the computational new applications networks requirements, researchers began to imagine a computing infrastructure that could provide access to on-demand computing and enable resource sharing, as in processing and storage cycles, in a secure and adaptable way, between dynamic individuals groups and institutions [17]. Considering the available technology and resource needs required by new applications, Ian Foster and Karl Kesselman conceptualized a computational model for sharing resources and coordinated problem solving in dynamic virtual organizations and multi-institutional [1, 2]. In this new model, it is possible to access computational resources, according to the work need to be performed by interconnected computers by a network likes it as can be seen in [1], [2], [6], [9] [10], [11], [12], [15], [16], [17], [18], [20], [23], [24], [25], [26] and [ 27]. What makes this archetype even more attractive is the possibility of common computers with different operating systems and resources, be able to donate processing cycles or storage space when they are idle [1], [2], [6], [11], [12], [15], [24], [25]. This paradigm is known as Grid Computing. The following figure illustrate this idea: