A study of multigon mask data preparation performance and its dependence on multigon data volume

The increasing adoption of complex curvilinear masks in leading-edge semiconductor manufacturing has introduced new challenges related to data volume and runtime. Multigons - a native curved geometry format, provides a more compact and efficient representation of curvilinear data. However, simply converting shapes to Multigons does not automatically improve performance. The benefits of Multigons depend on how efficiently they represent the actual mask shape, especially the number of control points used in Multigon representation during post-tape-out steps such as OPC and MPC. High control point count increase runtime complexity and data volume which affects the performance of downstream processes like fracture. In this study, we investigate how optimizing Multigon control points in MDP steps before fracture improves runtime and output file size. Results show that fewer control points lead to improvements in both runtime and output data volume, reinforcing the importance of upstream data volume optimization.