Video Overlay for Semiconductor Failure Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current failure analysis techniques for integrated circuit chips require tedious manual alignment and image processing of numerous time-resolved images generated by the LADA technique, hindering efficient identification of failing devices in dense chip layouts.

Innovation Solution

A video overlay system and process that allows users to load and align multiple images generated by the LADA technique, enabling playback and navigation like a video, with hotspot marking and analysis to identify critical areas and potential killer nets.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual alignment and image processing of numerous time-resolved images is performed, then identification accuracy of failing devices can be achieved, but analysis time and operational complexity increase significantly

Engineering Contradiction:
Improveidentification accuracyVSAvoidanalysis time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary alignment by applying a single alignment matrix to all images in the sequence based on alignment data from one reference image. This preliminary automated alignment eliminates the need for manual alignment of each individual image, significantly reducing analysis time while maintaining identification accuracy through subsequent hotspot analysis.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system creates a video overlay by copying and stacking multiple time-resolved images in sequence, allowing operators to view the complete set of images as a continuous video. This copying approach enables rapid visualization of device behavior across time intervals without requiring manual processing of each individual image.

Inventive Principle:
Principle #26Copying

2Manufacturing precision

If manual alignment of each image is performed, then precise positioning can be achieved, but operational efficiency decreases

Engineering Contradiction:
Improvepositioning precisionVSAvoidoperational efficiency
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The system applies a universal alignment matrix to all images in the time-resolved sequence, making the alignment process multi-functional across the entire image set. This single alignment operation positions all images precisely relative to each other, maintaining positioning precision while dramatically improving operational efficiency by eliminating repetitive manual alignment tasks.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system enables self-service alignment by automatically applying the alignment matrix to all images based on data from one reference image. This self-service approach maintains precise positioning without requiring continuous manual intervention, thereby improving operational efficiency while preserving positioning accuracy.

Inventive Principle:
Principle #25Self-service

3Loss of information

If 100+ images are viewed at various time intervals, then comprehensive analysis can be performed, but user workload and complexity increase

Engineering Contradiction:
Improvecomprehensive analysisVSAvoiduser workload
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The system merges 100+ time-resolved images into a single video overlay display, combining all images into one unified visual interface. This merging approach enables comprehensive analysis of device behavior across all time intervals while reducing user workload by eliminating the need to manually navigate and switch between individual images.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The system adds the time dimension to the spatial display by stacking images in temporal sequence within the video overlay. This dimensional transformation allows comprehensive time-based analysis of failing devices while simplifying user interaction, as operators can observe temporal evolution directly in the overlay without complex navigation through multiple separate images.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Data Source

PatentUS10650509B2Video overlay
Publication Date: 2020.05.12 SYNOPSYS INC
  • US10650509B2 patent drawing
  • US10650509B2 patent drawing
  • US10650509B2 patent drawing

AI summary

A method includes automatically aligning a laser-based timing analysis image of a semiconductor device with an image of a layout of the device. The method further includes controlling a speed at which a multitude of images subsequently obtained by the laser-based timing analysis are compared to the layout of the device to create a video overlay. The method further includes analyzing a multitude of potential failures of the semiconductor device by detecting movements of a multitude of hotspots on the layout as shown by the video overlay.