Molecular Architectures for QCA-Inspired Boolean Networks
R. Rinaldi, G. Maruccio, V. Arima, G.P. Spada, P. Samori, G. Cuniberti, J. Boland, and A.P. Bramanti
International Journal of Unconventional Computing 8, 301 (2012)
Most of the current research in nanoelectronics fall into two mainstreams. The first stream attempts to transpose the principles of traditional electronics into new technological scenarios. Basically, new ways to build and interconnect transistors are sought for. The second stream is that of research of more innovative solutions tailored on the nanoscale properties of matter, rather than adaptation and reuse old concepts developed at the microscale. In our view, both philosophies fall short of their goal. As to the first, since the transistor has been the very successful elementary brick of electronics for decades, trying to lengthen its life beyond the present constraints of technology obeys a somewhat reasonable economy principle of design. However, molecular-scale transistors generally perform worse than their silicon-based ancestors. Moreover, technologies for interconnecting them into complex, patterned configurations seem far off. The second approach has the merit of looking for a closer linking between the physics of devices and the computational paradigms implemented. The most outstanding and representative topic here is the Quantum-dot Cellular Automata (QCA) paradigm, from which, in fact, this project moves. QCA exploit quantum charge confinement and tunneling, as well as electrostatic interaction, to implement digital computation. Yet, in their original formulation, QCA do not fully account for real world systems, for instance, the cells confining the charges are implicitly supposed to be highly symmetric, as real world molecules are not. And again, no convincing solutions are outlined to pattern computing networks. Overall, realistic post-Moore technologies call for a globally new design style, accounting for physics, technology and computing architecture all at once, so that all the levels (abstract to physical) of the new nanoscale machines are conceived to fit each other. This is just the vision underlying the MolArNet project, which aims to take the fundamental steps toward realistic, out-of-the-lab implementation of molecular QCA.