Parameter optimization method and system of analog integrated circuit

Abstract

The invention discloses a parameter optimization method and system for an analog integrated circuit. The method comprises the following steps: acquiring a circuit structure, performance index requirements, design parameters and a value range thereof, and an iteration stop condition of the analog integrated circuit to be subjected to parameter optimization; based on the obtained design parameters and the value range thereof, sampling to obtain a training sample set; respectively inputting each sample in the training sample set into a circuit simulator for simulation to obtain a corresponding training response set; and performing iteration based on the training sample set and the training response set to realize parameter optimization of the analog integrated circuit. The invention provides a step-by-step optimization strategy based on mutual information analysis in order to alleviate the problem of curse of dimensionality and enable an algorithm to have the capability of processing a high-dimensional optimization problem. In each iteration, only part of design variables are selected for optimization, and the curse of dimensionality can be relieved.

Description

technical field [0001] The invention belongs to the technical field of analog integrated circuit design, relates to the field of Bayesian algorithm optimization, in particular to a parameter optimization method and system of an analog integrated circuit. Background technique [0002] Analog integrated circuits usually occupy a small part of the chip area of ​​the mixed-signal system-on-chip. Due to the lack of mature automatic design solutions, the parameters of analog integrated circuits need to be manually selected and adjusted. The process is time-consuming and tedious and highly dependent on the designer's intuition and According to experience, the design of analog integrated circuits has become a key bottleneck in the time to market of a SoC. With the development of machine learning, parameter optimization methods for analog integrated circuits based on machine learning have received extensive attention in recent years. [0003] The current optimization method of analo...

AI Technical Summary

Problems solved by technology

When the above-mentioned existing methods optimize the more commonly used analog integrated circuits (usually the dimension is between 20 and 40), it is difficult to converge to the optimal solution under limited computing resources, resulting in loss of use value

[0006] In addition, from the perspective of computational complexity of the algorithm, since the computational complexity of Bayesian optimization based on the Gaussian process is O(n 3 ) (n is the number of iterations), under limited computing resources, the number of iterations of the Bayesian optimization algorithm is limited, which further increases the difficulty of the Bayesian optimization algorithm to deal with high-dimensional problems

Images