EAGER: Exploration of topological self-organizing non-linear dynamical systems with memory as efficient scalable computing fabric

University Of California-San Diego, La Jolla CA

Investigators

Abstract

Unconventional computing models that depart substantially from how our traditional computers work are emerging as an important research direction in both fundamental science and technology. Quantum computing is the most celebrated example, where a phenomenon known as entanglement and unique to quantum mechanical systems, is expected to help in the solution of some difficult problems, such as prime factorization. However, there is a wide range of physical phenomena one can exploit other than entanglement, and that do not require extreme technological conditions to operate. This means that there is a vast area of research yet to explore between our traditional computers and quantum computing. This project will investigate a radically novel computing fabric that falls within this uncharted territory, and is able to solve difficult problems efficiently by means of non-quantum dynamical systems. Such a computing fabric is called memcomputing because it exploits the innate time non-locality of any physical system, namely memory of their past dynamics. If successful, this project is expected to usher in a paradigm shift in computing that will reverberate across practically all aspects of technology. The proposed activities will contribute to the professional development of highly skilled personnel with expertise that cuts across different disciplines involved in this project. To prove the fundamental value of such a non-von-Neumann computing fabric, this award will support the development of a simulator for digital memcomputing machines based on the numerical integration of the differential equations of appropriate non-quantum, non-linear dynamical systems with memory. To facilitate an independent verification, the simulator will then be employed to tackle challenging combinatorial problems, such as prime factorization, whose solution can be easily checked. This award will thus serve as the stepping stone towards the ultimate goal of realizing such digital memcomputing machines in hardware. This project will also advance our knowledge of the physical requirements to compute efficiently by investigating the deep relation between computing, topology, and collective behavior of physical systems. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.