DEC 01, 2025

GA-Optimized Overlay Sampling

10-2025-0187557 · KP25137

Inventors

Taekwon Jee, Jihoon Jung, Jeng-Hun Suh

Applicant

SemiAI, Co.,Ltd.

Official Title

METHOD FOR OPTIMIZING SEMICONDUCTOR OVERLAY SAMPLING BASED ON A GENETIC ALGORITHM USING RULE-BASED SEEDING AND OVERLAY SAMPLING DEVICE USING THE SAME

Summary

This invention relates to a genetic algorithm-based method and apparatus for optimizing semiconductor overlay sampling using rule-based seeding to improve both efficiency and accuracy in photolithography processes, addressing limitations of conventional rule-based and heuristic sampling methods that fail to capture nonlinear distortions and process variations. An initial population is generated by combining rule-based sampling results with random sampling, and optimal sampling positions are iteratively derived through genetic algorithm operations including fitness evaluation, selection, crossover, and mutation, where performance is assessed based on spatial uniformity, prediction error, and uncertainty. This approach enables high measurement accuracy with fewer sampling points, improves convergence speed, mitigates local optima issues, and provides a scalable and adaptive solution for dynamic semiconductor manufacturing environments.