2025 L5: Defects, Contaminants, Yields (2)/Mathematics of Failure-and-Reliability(1)

2025 L5: Defects, Contaminants, Yields (2)/Mathematics of Failure-and-Reliability(1)

by Tian-Li Wu (吳添立)

Applied Sciences & Engineering Engineering & Technology TGPProduction and industrial engineeringTGPRReliability engineering
🎙 Tian-Li Wu (吳添立) 👥 11K 📅 October 16, 2025 ⏱ 159 min 👁 442 📄 lecture 🧭 2026-08-16
Available in: English (current) Français

Keywords

yielddefect densitybinomial distributionPoisson approximationMurphy's integralcritical areareliability

Summary

This lecture, part of a course on semiconductor device reliability and failure physics, focuses on the statistical foundations of yield modeling and the distinction between yield and reliability. It begins by reviewing the binomial distribution, using coin toss examples to illustrate the probability of outcomes. The Poisson approximation is then introduced as a simplification for rare events, leading to the basic yield equation Y = exp(-D*A), where D is defect density and A is die area. The lecture emphasizes that yield decreases exponentially with increasing area and defect density. To account for non-uniform defect distributions, Murphy’s integral is presented, integrating the yield over a defect density distribution. Several distribution models (constant, triangular, exponential, Gaussian, gamma) are discussed, with graphical representations for estimation. The lecture then contrasts yield and reliability: yield defects cause complete opens or shorts, while reliability defects cause performance degradation (e.g., increased resistance) without complete failure. The concept of critical areas (yield critical area vs. reliability critical area) is introduced, along with a relationship R = Y^K. The lecture concludes by summarizing key takeaways: defects are unavoidable, yield depends on defect density and area, and defect clustering in reality deviates from the random distribution assumption.

197 words

Critical Evaluation

Value of the Information & Strength of the Argument

The lecture provides a solid foundation in yield modeling, deriving the exponential yield equation from statistical principles and clearly explaining the assumptions and limitations. The argumentation is logical and progressive, building from basic probability to practical yield models. The distinction between yield and reliability critical areas is particularly valuable, as it clarifies a common source of confusion. The use of graphical examples and practical illustrations (e.g., die size increase) enhances understanding. However, the lecture could benefit from more concrete examples of defect distributions and their impact on yield in real manufacturing scenarios.

Scientific Rigor, Source Quality, Title Accuracy

The lecture is scientifically rigorous, presenting well-established statistical models and their application to semiconductor yield. The content aligns with standard textbooks on semiconductor manufacturing and reliability. However, the video does not cite specific external sources, relying instead on the instructor’s expertise and course materials. The title accurately reflects the content, covering defects, yields, and the mathematics of failure and reliability. The lecture is part of a structured course, which adds to its credibility.

180 words

Title / Content Match

The title accurately reflects the content: it covers defects, contaminants, yields, and introduces the mathematics of failure and reliability.

Quality & Reliability

8/10

The lecture is part of a university course on semiconductor device reliability and failure physics. It presents established statistical models (binomial, Poisson, Murphy's integral) and distinguishes yield vs. reliability critical areas. The content is consistent with standard textbooks and academic literature, though no external sources are cited in the video itself.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Cited Sources

Concurring Sources

  • Semiconductor Manufacturing Technology — Standard textbook covering yield modeling and defect statistics.

Contribution & Novelties

The lecture provides a clear and structured introduction to yield modeling, emphasizing the statistical basis and the distinction between yield and reliability. It effectively bridges theoretical probability distributions with practical semiconductor manufacturing concerns. The discussion of Murphy’s integral and various defect distribution models is particularly useful for understanding real-world yield variations.

Pour aller plus loin :

  • Poisson distribution — The Poisson approximation is central to the yield equation; this page provides a thorough mathematical background.
  • Binomial distribution — The foundational distribution used to derive the yield model; this page explains its properties and applications.
  • Semiconductor device fabrication — Provides context on the manufacturing processes where yield and reliability are critical.
  • Murphy’s yield model — A specific yield model mentioned in the lecture; this page details its formulation and assumptions.
  • Integrated circuit — Background on the product whose yield is being modeled.

141 words

Radar Profile

The radar profile shows high scores in quantity and quality of information, with a slightly lower technical level, indicating a lecture that is comprehensive and reliable but accessible. The overall high scores reflect the solid educational value and scientific rigor.

Reliability 8/10