Electric Analog Circuit Design with Hypernetworks and a Differential Simulator

TL;DR

This paper introduces a deep learning-based method for fully automatic analog circuit design, utilizing hypernetworks and a differential simulator to generate and refine circuit configurations and parameters.

Contribution

It presents a novel two-stage neural network approach with hypernetworks for automatic component selection and parameter prediction, advancing beyond existing methods that only fit parameters.

Findings

The proposed model efficiently designs circuits with superior performance.

Hypernetwork conditioning improves component placement accuracy.

Differential simulation refines component parameters effectively.

Abstract

The manual design of analog circuits is a tedious task of parameter tuning that requires hours of work by human experts. In this work, we make a significant step towards a fully automatic design method that is based on deep learning. The method selects the components and their configuration, as well as their numerical parameters. By contrast, the current literature methods are limited to the parameter fitting part only. A two-stage network is used, which first generates a chain of circuit components and then predicts their parameters. A hypernetwork scheme is used in which a weight generating network, which is conditioned on the circuit's power spectrum, produces the parameters of a primal RNN network that places the components. A differential simulator is used for refining the numerical values of the components. We show that our model provides an efficient design solution, and is superior to alternative solutions.