Design and Optimization of Low-Dropout Voltage Regulator Using Relational Graph Neural Network and Reinforcement Learning in Open-Source SKY130 Process
Authors:
- 28 October 2023
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 01-09
Abstract
Design automation and optimization for analog integrated circuits (ICs) are challenging, especially for transistor sizing. Given certain design specifications and circuit topology, circuit designers need to size various components to achieve the desired performance, possibly involving many optimization iterations. Recently, reinforcement learning (RL) has been applied to optimize analog circuits. The trained RL agents can achieve very high sample efficiency over evolutionary-based algorithms. By using the ability of transfer learning, the trained agent can be applied to optimize the same circuit across different technology nodes and even the circuits with different topologies. However, a significant bottleneck in applying machine learning (ML) techniques to analog IC design is the non-disclosure agreement (NDA) of the process development kit (PDK), which makes reproducibility of the prior art a big challenge. This work presents an RL framework that leverages the open-source SKY130 PDK to address the limitation above. We apply a novel heterogeneous graph neural network (GNN) called relational graph convolutional network (RGCN) as the function approximator of RL to capture more topological information about a circuit. As a proof-of-concept, low-dropout voltage regulators (LDO) are optimized by our proposed RL circuit optimizer framework to show its feasibility, achieving promising results.Keywords
Funding Information
- University of Toronto
This publication has 19 references indexed in Scilit:
- BagNet: Berkeley Analog Generator with Layout Optimizer Boosted with Deep Neural Networks2019
- An Efficient Bayesian Optimization Approach for Automated Optimization of Analog CircuitsIEEE Transactions on Circuits and Systems I: Regular Papers, 2017
- Parasitic-aware GP-based many-objective sizing methodology for analog and RF integrated circuits2017
- Optimization of LDO voltage regulators by NSGA-II2016
- Analog Circuit Optimization Based on Hybrid Particle Swarm Optimization2015
- Optimizing an LDO voltage regulator by evolutionary algorithms considering tolerances of the circuit elements2015
- An enhanced MOEA/D-DE and its application to multiobjective analog cell sizing2010
- High PSR Low Drop-Out Regulator With Feed-Forward Ripple Cancellation TechniqueIEEE Journal of Solid-State Circuits, 2010
- An automated design approach for CMOS LDO regulators2009
- Full On-Chip CMOS Low-Dropout Voltage RegulatorIEEE Transactions on Circuits and Systems I: Regular Papers, 2007