AMSDAT: Integrated Analog and Mixed-Signal Design Optimization Framework for SoC Applications

Abstract

Traditionally, analog and mixed-signal circuit design involves a long, iterative process based on a thorough understanding of circuit behavior and the intuition of the designer, and yet resulting in sub-optimal designs with high probability. In the proposed work, a fully automated, efficient, and robust design optimization framework to find the optimal solution, over a large design search space, for the multi-objective design of analog ICs is presented. The methodology integrates the metaheuristic algorithms such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) with SPICE for analog and mixed-signal circuit design automation by optimizing its performance metrics to meet the design specifications. A user-friendly graphical interface is implemented in MATLAB, and the design verification is seamlessly simulated in SPICE for accuracy. The tool is demonstrated to achieve target design and even overperformance of the design, for 2-stage op-amp and analog-to-digital converter applications. The framework tool exemplifies a robust, scalable, and accurate methodology for multi-objective optimization of analog and mixed-signal circuits of varying scale and design complexity. Further, the framework is applicable to any process technology defined in the netlist of the circuit under design consideration and hence design migration across process nodes is a simple process.