Directly performance-constrained template-based layout retargeting and optimization for analog integrated circuits

Due to intrinsic intricacy, layout parasitics exhibit a significant impact on the performance of analog integrated circuits. In this paper a directly performance-constrained parasitic-aware automatic layout retargeting and optimization algorithm is presented. Unlike the conventional sensitivity analysis, a general central-difference based scheme using any simulator for sensitivity computation is deployed. We propose a piecewise sensitivity model to enforce more accurate sensitivity computation during parasitic optimization. Moreover, mixed-integer performance constraints due to parasitics are included in the formulated mixed integer nonlinear programming problem rather than through either indirect parasitic-bound constraints or inaccurate worst-case sensitivities. A graph technique and mixed-integer nonlinear programming are effectively combined to solve the formulated parasitic optimization problem. The automatically generated target layouts can satisfy performance constraints to ensure the desired specifications. The experimental results show that the proposed algorithm can achieve effective retargeting of analog circuits with less layout area and significant reduction in execution time.