Overcoming the limitations of von Neumann architecture in big data systems

Current methods of surmounting the limitations of von Neumann architectures in building Big Data systems do not address the core architectural problems, thereby incurring resource utilization inefficiencies. This paper proposes the replacement of von Neumann architectures with a non-volatile computer architecture that is specifically designed to scale up to Big Data. This new architecture is assessed against prevalent Big Data models, in the context of performance and power consumption, the results of which show that the prerequisite technology is already available for its implementation.