一個最佳化光罩設計演算法在系統晶片積體電路佈局應用之研究

Abstract

針對系統晶片積體電路佈局之需要，本論文首度提出一個整合型之最佳化光罩設計演算法;此方法成功的整合了基因演算法、光學鄰近效應修正 規則、光學鄰近效應修正數值模擬，以及積體電路實作上之經驗法則，同時輔以叢集平行計算技術。光學鄰近效應修正之基本概念為藉由加入適當 的輔助性圖樣或是移動原有電路佈局圖樣之部分邊界，以補償在光學微影過程中因光學鄰近效應在晶圓曝光時的影像失真。

In this work, various rule-based and model-based OPC methods are investigated so that a trade off between speed and correctness is found. Accordingly, we propose a hybrid intelligent OPC system for obtaining both the benefits of rule-based and model-based methods, which can be applied to optimize the mask design for system-on-a-chip layout automation. In the proposed OPC system, the genetic algorithm, lithography numerical simulation and the empirical experiences are introduced systematically. A Linux based PC cluster system is also constructed that some parallel computing methods are employed to perfect the efficiency.