Post-layout simulation automation based on Reinforcement Learning using Schematic to Layout sync module

Due to complexity and difficulty of circuit design, attempts to automate circuit design increase recently. In particular, reinforcement learning based circuit optimization increases. Traditional research focused on pre-layout simulation because of difficulty of automation layout. In this paper, we proposed post-layout simulation automation method using 3 kinds of modules. Modules are consisted of pre-processing data module, schematic to layout sync module, and parasitic extraction tool automation module. These modules contribute to optimize circuit accurately. As a result, modifying layout DB time decreases nearly 75.9%, and the variation of the skew and the duty cycle error decrease 53.4%, 12.0% respectively.