Fast Machine-Learning-Driven Supply Noise-Aware Macromodeling for High-Speed Nonlinear Drivers
Songyu Sun,Xiao Dong,Qi Sun,Xunzhao Yin,Quan Chen,Cheng Zhuo
DOI: https://doi.org/10.1109/tcad.2024.3455242
IF: 2.9
2024-01-01
IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
Abstract:Emerging domains such as artificial intelligence, 5G mobile and automotive are increasingly reliant on high-speed circuits for efficient processing, in which achieving high operating frequencies and data rates is crucial to enable productive data exchange and rapid responses. High-speed data as well as low noise margin in the high-speed serial links call for efficient models of drivers. In this paper, we propose a fast machine-learning-driven macromodel for high-speed drivers, which can efficiently capture the nonlinear characteristics of drivers considering dynamic supply noise with low model complexity. A decoupling-superposition strategy is employed to effectively calculate the impact of power supply noise. Additionally, we introduce a piecewise-segmented method for macromodel solving to further enhance the speed of model utilization. Experimental results demonstrate that compared to HSPICE, the proposed macromodel achieves up to 50-1200× speedup while maintaining sufficient accuracy, even for signals with GHz data rate.