Scientific and commercial computing is undergoing an immense change with increasing demands for processing of large-scale datasets for a variety of needs such as simulation, modeling, and calculations of multivariate equations. Computation has largely been used as a method for achieving the results and is now shifting to a data intensive computing model to accomplish some of the most demanding scientific challenges in existence. Current data and storage architectures are not sufficient to provide for Petascale and Exascale data processing and analysis which will require new ways of data access at multi-terabyte per second speeds. An architectural framework is proposed for data intensive cloud computing called Datalanx.