Optical Proximity Correction Using Segmented Patch Averaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing complexity and miniaturization of semiconductor processes lead to distortion issues during photolithography due to light interference and diffraction, resulting in abnormal circuit operation, necessitating improved optical proximity correction to ensure accurate layout printing.
Innovation Solution
An optical proximity correction system and method that divides layouts into segments, calculates hash values for each segment, and applies biasing to reduce errors, using a master-slave device architecture to optimize turnaround time and accuracy by minimizing the number of patches and maximizing population for average value calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If the layout is divided into many small patches for optical proximity correction, then the correction accuracy is improved, but the turnaround time increases
Solution Approach 1:
The patent divides the layout into patches and further segments each patch into multiple segments for parallel processing. This segmentation allows the system to maintain high correction accuracy by processing each segment individually while reducing overall turnaround time through parallel computation across multiple slave devices.
Solution Approach 2:
The master device performs preliminary actions by generating hash values for each segment and determining which segments should be processed by which slave devices before the actual optical proximity correction computation begins. This preliminary organization optimizes the subsequent parallel processing efficiency.
Solution Approach 3:
The patent creates multiple copies of patch data distributed across different slave devices for parallel processing. Each slave device receives and processes copies of relevant patch segments independently, then returns results to the master device, thereby achieving speedup through parallel computation while maintaining accuracy.
2Productivity
If the number of patches is reduced to decrease turnaround time, then processing speed is improved, but the population for average value calculation decreases, affecting accuracy
Solution Approach 1:
The patent resolves this contradiction by moving from a two-dimensional patch structure to a three-dimensional segment structure within patches. By dividing each patch into multiple segments that can be independently processed and aggregated, the system maintains sufficient population for accurate average value calculation while enabling parallel processing across segments and slave devices, thus improving processing speed without sacrificing accuracy.
Solution Approach 2:
The master device merges results from multiple segment calculations to compute accurate average values. By combining data from multiple segments across different patches, the system maintains sufficient population for statistically significant average value calculation while keeping individual patch sizes manageable for parallel processing.
Data Source
AI summary
An optical proximity correction system and an operating method are provided. Provided is an optical proximity correction system comprising, a plurality of patch blocks which include a plurality of patches including a segment information table, a plurality of slave devices which receive the segment information table from the plurality of patch blocks to generate a minimum patch table, and a master device which receives the minimum patch table from the plurality of slave devices, generates a segment average calculation table, and performs an optical proximity correction on the patches recorded in the segment average calculation table.