Uma metodologia de síntese automática com o foco na tecnologia Quantum-Cellular-Automata

Abstract

The complementary metal-oxide-semiconductor (CMOS) has been used as the main technology to design integrated circuits in the last decades. However, this technology is near to its physical limits. This way, several new devices have been proposed for the replacement of the transistors, such as the technology Field-Coupled Nanocomputing (FCN). Many FCN paradigms are being studied, including quantumdot cellular automata (QCA) and nanomagnetic logic (NML). This work proposes a new methodology for QCA synthesis, which attaches different logic and physical synthesis tools, filling gaps, and proposing new methods and design elements for integration. A new version of the state-of-the-art QCA automatic synthesis method has been proposed, which has been called Migortho. In addition, a new data structure was adopted to be used, in this case, the Majority Inverter Graph (MIG). In the same way, new design elements were proposed, such as MajX, Double Wire, and Double Wire Lent. Furthermore, a series of commands were implemented that invoke MIG optimization methods through the EPFL Logic Libraries library called Mockturtle. In short, it was possible to evaluate the impact of the new elements proposed in this work, as well as the new data structure for logic representation. Likewise, it was possible to identify which command implemented could impact the greatest reduction in input circuits, reaching an average optimization rate of 15; 8%. Besides, several circuits were evaluated through synthesis by the Migortho method, which generally presented gains concerning the result of the version of the Ortho method. In addition, two versions of the used benchmarks were proposed for further evaluations of the QCA synthesis. Finally, evaluations were made with the benchmark specific state-of-the-art for the MIG structure, which made it possible to obtain both the estimate of the final QCA layout and to apply optimizations through the Mockturtle commands, that it optimized the circuits described by the state-of-the-art method.