DRC Violation Correction via ML Recipe Selection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current Design Rule Check (DRC) violation correction techniques rely heavily on manual designer experience and are inefficient, as they lack the ability to automatically identify suitable recipes for fixing DRC violations and analyzing recipe inconsistencies, leading to increased time and resource consumption.

Innovation Solution

A DRC verification system that uses machine learning and artificial intelligence algorithms for outlier detection, spectral clustering, and recipe selection, allowing for automatic identification of suitable recipes and reducing manual review efforts by classifying layout patterns and selecting recipes based on previously successful corrections.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual designer experience is used for DRC violation correction, then accuracy in selecting suitable recipes can be maintained, but productivity and efficiency deteriorate due to increased time and resource consumption

Engineering Contradiction:

Improveaccuracy in selecting suitable recipesVSAvoidspeed of DRC violation correction

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The system performs self-service by automatically analyzing DRC violations, classifying layout patterns, and selecting appropriate correction recipes without requiring manual designer intervention. The machine learning model independently processes violations and generates correction recommendations, enabling the system to serve itself rather than relying on external human expertise for each violation case.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces the mechanical system of manual designer review and selection with an automated machine learning-based system. The ML model processes layout patterns, identifies violations, and selects recipes through computational algorithms, substituting human cognitive and manual processes with automated electronic processing that operates faster and at scale.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If automated DRC violation correction systems are implemented, then productivity and speed are improved, but reliability deteriorates due to inability to accurately identify suitable recipes and analyze recipe inconsistencies

Engineering Contradiction:

Improvespeed of DRC violation correctionVSAvoidaccuracy in identifying suitable recipes

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system performs preliminary action by pre-classifying layout patterns and pre-processing violation data before actual correction is needed. The machine learning model is trained in advance on historical DRC violation data, learning from previously successful corrections. This preliminary training and classification enable the system to quickly and accurately identify suitable recipes when actual violations occur, without compromising reliability for speed.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements feedback mechanisms by analyzing the results of applied correction recipes and using this information to improve future selections. The ML model learns from the outcomes of previous corrections, adjusting its predictions to avoid inconsistent recipe applications. This feedback loop ensures that automated selection maintains high reliability by continuously improving based on actual performance data.

Inventive Principle:

Principle #23Feedback

3Reliability

If manual review processes are used for DRC violations, then reliability in recipe selection is maintained, but loss of time increases significantly

Engineering Contradiction:

Improveaccuracy in recipe selectionVSAvoidtime for DRC violation analysis

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent replaces the time-consuming manual review mechanical system with automated machine learning processing. The ML model rapidly analyzes layout patterns, identifies violations, and selects recipes through computational operations that execute in seconds rather than the minutes or hours required for manual designer review, dramatically reducing time loss while maintaining reliability through algorithmic consistency.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system uses copying by replicating successful correction patterns from historical data. Instead of manually analyzing each new violation from scratch, the ML model copies and adapts proven correction recipes from similar previously analyzed cases. This copying approach maintains reliability by relying on validated solutions while reducing time through pattern recognition rather than repeated manual analysis.

Inventive Principle:

Principle #26Copying

Data Source

PatentUS20240330563A1System and method for diagnosing design rule check violations

Publication Date: 2024.10.03 TAIWAN SEMICONDUCTOR MANUFACTURING CO LTD

AI summary

A system and method for fixing DRC violations includes receiving a layout pattern having a design rule check (DRC) violation therein, determining that the layout pattern is an inlier based upon a comparison of the layout pattern with a plurality of previously analyzed layout patterns. The comparison may be performed by an anomaly detection algorithm. The system and method may also include selecting a recipe from a pool of recipes previously applied to the plurality of previously analyzed layout patterns for fixing the DRC violation in the layout clip upon determining that the layout pattern is an inlier.