Second-Order Arnoldi Reduction Using Weighted Gaussian Kernel

Modeling and design of on-chip interconnect continue to be a fundamental roadblock for high-speed electronics. The continuous scaling devices interconnects generates self mutual inductances, resulting in generating second-order dynamical systems. model order reduction is an essential part any modern computer-aided tool prefabrication verification the components interconnects. existing methods use expensive matrix inversion generate orthogonal projection matrices often do not preserve stability passivity original system. In this work, Arnoldi method proposed, which selectively picks interpolation points weighted with Gaussian kernel given range frequencies interest matrix. proposed ensures reduced-order over desired frequency range. simulation results show that combination multi-shift selective dynamically optimal better accuracy numerical compared techniques.