Adaptive Outlier Detection for Power MOSFETs Based on Gaussian Process Regression

@article{Shimozato2022AdaptiveOD,
  title={Adaptive Outlier Detection for Power MOSFETs Based on Gaussian Process Regression},
  author={Kyohei Shimozato and Michihiro Shintani and Takashi Sato},
  journal={2022 IEEE Applied Power Electronics Conference and Exposition (APEC)},
  year={2022},
  pages={1709-1714}
}
Outlier detection of semiconductor devices is important since manufacturing variation is inherently inevitable. In order to properly detect outliers, it is necessary to consider the discrepancy from underlying trend. Conventional methods are insufficient as they cannot track spatial changes of the trend. This study proposes an adaptive outlier detection using Gaussian process regression (GPR) with Student-t likelihood, which captures a gradual spatial change of characteristic variation… 

Figures and Tables from this paper

References

SHOWING 1-10 OF 13 REFERENCES
Comparison of wafer-level spatial I/sub DDQ/ estimation methods: NNR versus NCR
  • S. SabadeD. Walker
  • Computer Science
    Proceedings. 2004 IEEE International Workshop on Current and Defect Based Testing (IEEE Cat. No.04EX1004)
  • 2004
TLDR
Two methods that use deterministic variation in I/ sub DDQ/ at the wafer level for estimating fault-free I/subDDQ/ of a chip are proposed and evaluated: nearest neighbor residual (NNR) and neighbor current ratio (NCR).
Analysis and decomposition of spatial variation in integrated circuit processes and devices
Variation is a key concern in semiconductor manufacturing and is manifest in several forms. Spatial variation across each wafer results from equipment or process limitations, and variation within
Gaussian process regression with Student-t likelihood
TLDR
This work discusses the properties of a Gaussian process regression model with the Student-t likelihood and utilizes the Laplace approximation for approximate inference and compares the approach to a variational approximation and a Markov chain Monte Carlo scheme, which utilize the commonly used scale mixture representation of the student-t distribution.
Bulk and epitaxial growth of silicon carbide
Evolution of lattice distortions in 4H-SiC wafers with varying doping
TLDR
From analysis of peak widths, it was determined that threading dislocations in the top 6 microns of the wafer increase after epitaxial layer growth, thereby reducing device reliability.
Spatial variation of lattice plane bending of 4H-SiC substrates
Basal plane bending of on- and off-axis 4H-SiC substrates was measured by high-resolution X-ray diffractometry (HRXRD). In order to evaluate basal plane bending properties in detail, line scans of
Variance reduction using wafer patterns in IDDQ data
  • Proceedings of International Test Conference (ITS), Oct. 2000, pp. 189–198.
  • 2000
Which is harder SOA test for SiC MOSFET to do Unclamped Inductive Switching (UIS) or Unloaded Short Circuit mode Switching (USCS)? Does UIS play a role of USCS?
It is tough for SiC MOSFET to maintain the same time width (tw) of Unloaded Short Circuit mode Switching (USCS), which we define as the index of the Short Circuit Safety Operation Area (SCSOA)
Statistical post-processing at wafersort-an alternative to burn-in and a manufacturable solution to test limit setting for sub-micron technologies
TLDR
Burn-in data and customer Defects per Million units (DPM) data show a 30-60% decrease in failure rate with SPP implementation with very acceptable yield loss, and challenges of production implementation have been overcome.
Gaussian Processes for Machine Learning, ser. Adaptive Computation and Machine Learning series
    ...
    ...