Recently, the research team headed by Professor Xie Guangjun at the Integrated Circuit Design Research Center, School of Microelectronics, Hefei University of Technology, has made significant strides in the automated design of Molecular Field-Coupled Nanoscale Circuits (MolFCN). The team has successfully developed and released the open-source, full-process EDA platform, iFCN. This innovative platform facilitates a seamless, automated design workflow, spanning from RTL circuit input all the way to device-level layout output.

To achieve this, the platform incorporates several advanced features: a Morton code-driven quadtree data structure for efficient data organization, topological embedding capabilities sourced from Graph Convolutional Networks (GCN) to enhance design flexibility, an optimized A* routing algorithm for improved pathfinding, and a lightweight adaptive clock phase allocation mechanism for precise timing control. These innovations collectively boost layout compactness, placement quality, and routing success rates.

Test results indicate that iFCN outperforms traditional methods, offering a roughly 30% increase in runtime efficiency and a 10% reduction in layout area. Moreover, it adeptly manages circuit crossovers in designs with over 5,000 nodes. When benchmarked against international advanced tools, iFCN holds its own, providing even richer functionality. Notably, it surpasses 24nm and 12nm CMOS process designs in terms of both area utilization and power consumption.

At present, iFCN is fully accessible on GitHub, complete with its source code, test data, and sample circuits, inviting the global research community to explore and contribute.