Abstract: Scientific applications often require large computational power and availability massive numbers of processing data intensive applications. Traditionally these needs were addressed using grids, clusters and supercomputers for processing and managing large scale applications, but cloud computing offers users with a new model for utilizing computing infrastructure. Scientific applications include mathematical model and numerical solutions techniques to solve engineering and scientific problems. That model often required large number of computing resource and to cut down such computing complexities a dedicated high performance infrastructure such as the grid, clusters were used. The resources for execution of scientific application may be seen in some parts of the year which can lead to long waiting times for utilization of these resources ,or the available resources for one applications Mayer be insufficient to complete the applications before its deadline. In these cases, scientific resources may be complemented by cloud resources. Leasing cloud computing services on pay per use basic, even minor institutions can easily access a large number of resources, which are utilized and paid for only for the time they are actually utilized.

Keywords: Paas, IaaS, Service Level Agreement.