Since the inception of von-Neumann architecture for computer design, there has been no new paradigms or revolutions in computer architectures. Computer applications have been increasing at an exponential rate, however, the basic computer architectures remained the same. The conventional computer architectures, which are based on primitive building blocks including arithmetic logic units, floating point processor units, logical shift units, and register file units created tremendous semantic-gap and inefficiencies in information system processing. It is about time to revisit the standard von-Neumann computation model and argue about its efficiencies, as we are entering into a new era of information processing where applications don't have any boundaries in computation, communication, and information storage.

In this paper, we propose a revolutionary computer architecture which avoids the semantic-gap and inefficiencies, and is based on an object-oriented paradigm to provide the benefits of abstraction, inheritance, hierarchy, modularity, extensibility, and polymorphism. We will describe the fundamental building blocks for this architecture and propose a possible approach for implementing these new generation of computers which will not make software and hardware obsolete before coming to existence. We will present the design issues related to such architectures and research directions needed to study the feasibility of these architectures.