Case study · Semiconductors & electronics

Intel automates wafer failure-pattern detection for yield analysis

Manual end-of-line yield analysis could not cover every wafer in every lot, accuracy depended on engineer experience, and sharing knowledge across 15 fabs was slow.

Manufacturer
Intel
Use cases
1
Evidence
Manufacturer-reported
Year
2021

Manufacturer-reported  This result comes from the manufacturer's own publications, such as newsroom posts, earnings materials or engineering blogs. It has not been independently verified.

The challenge

Manual end-of-line yield analysis could not cover every wafer in every lot, accuracy depended on engineer experience, and sharing knowledge across 15 fabs was slow.

The approach

Intel IT built an AI workflow using machine learning, deep learning and image processing to detect, cluster and classify wafer failure patterns automatically and push results to yield engineers.

1 use case in this implementation

  1. Root-cause analysis for quality & yield

    AI detects, clusters and classifies end-of-line wafer failure patterns (including multiple per wafer) to speed root-cause analysis and yield ramp.

    • 100 percent coverage of the wafers and lots
    • >90 percent accuracy in detecting baseline patterns
    • helps us maintain the high yield we had in previous process generations while keeping the same level of yield analysis staff

Outcomes as reported by the source. We do not adjust, combine or extrapolate reported figures.

Related case studies

Put AI to work

Find the AI opportunities hiding in your plant.

Genvo, our implementation partner, works with manufacturers to pick the right first project, connect the systems you already run and put AI agents and tools into daily use, with people approving what goes out.

Assessments and consultations are provided by Genvo. How our partnership works.