| Abstract: Over the last decade, it has become clear that conventional VLSI is reaching key scaling limits. Moreover, the energy efficiency of human-engineered electronic devices is many orders of magnitude lower as compared to biological computational structures. Such inefficiencies, for example, severely limit our ability to move from 2D to 3D architectures in materials systems as is needed to achieve high performance and span the complexity required for applications using big data or artificial intelligence. We therefore need a new class of materials that can enable computing, but which are not bound by the rules of conventional VLSI, and we are inspired by the fact that the brain is an existence proof for such low-energy high-computing materials that do not rely on a von Neumann architecture. We have pursued the use of polymer-networked nanoparticles as a possible alternative. We will summarize our progress on characterizing these materials and report the primitive structures that we have created computationally thus far. |
