Circuit parameter optimization method and system based on reinforcement learning

Abstract

The invention discloses a circuit parameter optimization method and system based on reinforcement learning, and the method comprises the steps: obtaining an optimization parameter and observation information of a simulation circuit, and carrying out the initialization of the optimization parameter; inputting the observation information into a pre-trained neural network model to output an update amount of the optimization parameters; and updating the optimization parameters according to the update amount to achieve an optimization target. According to the method, the optimal optimization parameters under the given design parameters and design objectives can be rapidly obtained, the circuit design efficiency is improved, and the listing time of circuit products is shortened.

Description

technical field [0001] The invention relates to the technical field of integrated circuit design, in particular to a circuit parameter optimization method and system based on reinforcement learning. Background technique [0002] In circuit design work, parameter optimization is an inevitable problem, especially in analog circuit design, which requires a high level of experience for designers. With the continuous improvement of the scale and performance of integrated circuits, the demand for circuit design optimization is increasing. The traditional optimization algorithm needs a large number of iterations, because the emulator needs to be called to calculate the objective function in the iteration, and the cost of calling the emulator is very high, so it takes a lot of time to optimize the circuit parameters. This greatly restricts the efficiency of circuit design, not only increases the labor cost of circuit design, but also prolongs the time to market of circuit products,...

AI Technical Summary

Problems solved by technology

In the case of a large number of parameters, it is difficult for novice designers to quickly obtain better parameters to meet the current circuit design goals.

In addition, the problem of process size migration is often encountered in circuit design. After migrating from a large-scale process to a small-scale process, although the circuit topology remains unchanged, the circuit parameters need to be redesigned by the designer to meet the design requirements.

Images