Advanced analog layout design automation in compliance with density uniformity

To fabricate a reliable integrated circuit chip, foundries follow specific design rules and layout processing techniques. One of the parameters, which affect circuit performance and final electronic product quality, is the variation of thickness for each semiconductor layer within the fabricated chips. The thickness is closely dependent on the density of geometric features on that layer. Therefore, to ensure consistent thickness, foundries normally have to seriously control distribution of the feature density on each layer by using post-processing operations. In this research, the methods of controlling feature density distribution on different layers of an analog layout during the process of layout migration from an old technology to a new one or updated design specifications in the same technology have been investigated. We aim to achieve density-uniformity-aware layout retargeting for facilitating manufacturing process in the advanced technologies. This can offer an advantage right to the design stage for the designers to evaluate the effects of applying density uniformity to their drafted layouts, which are otherwise usually done by the foundries at the final manufacturing stage without considering circuit performance. Layout modification for density uniformity includes component position change and size modification, which may induce crosstalk noise caused by extra parasitic capacitance. To effectively control this effect, we have also investigated and proposed a simple yet accurate analytic method to model the parasitic capacitance on multi-layer VLSI chips. Supported by this capacitance modeling research, a unique methodology to deal with density-uniformity-aware analog layout retargeting with the capability of parasitic capacitance control has been presented. The proposed operations include layout geometry position rearrangement, interconnect size modification, and extra dummy fill insertion for enhancing layout density uniformity. All of these operations are holistically coordinated by a linear programming optimization scheme. The experimental results demonstrate the efficacy of the proposed methodology compared to the popular digital solutions in terms of minimum density variation and acute parasitic capacitance control