Wafer-View Defect-Pattern-Prominent GDBN Method Using MetaFormer Variant

Good-Die-in-Bad-Neighborhood (GDBN) is a technique employed to identify chips that pass initial tests but may have defects. Previous research used neural networks and expanded observation windows but ignored the impact of isolated dice. This paper improves wafer pattern information through denoising...

Full description

Saved in:
Bibliographic Details
Published inProceedings - International Test Conference pp. 76 - 80
Main Authors Li, Shu-Wen, Yen, Chia-Heng, Chang, Shuo-Wen, Chu, Ying-Hua, Wu, Kai-Chiang, Chao, Mango Chia-Tso
Format Conference Proceeding
LanguageEnglish
Published IEEE 03.11.2024
Subjects
Authorization
defective parts per million (dppm)
geographical part average testing (gpat)
good-die-in-bad-neighborhood (gdbn)
IEEE merchandise
latent defect
neural network
Neural networks
Noise reduction
Semiconductor device modeling
Training
Online AccessGet full text

Cover

More Information
Summary:Good-Die-in-Bad-Neighborhood (GDBN) is a technique employed to identify chips that pass initial tests but may have defects. Previous research used neural networks and expanded observation windows but ignored the impact of isolated dice. This paper improves wafer pattern information through denoising and creates a lightweight model. It also reduces training time by annotating multiple dice simultaneously. Experiments on real-world datasets show the model effectively captures more Test Escapes, reducing Defective Parts Per Million (DPPM) and improving return merchandise authorization gains.
ISSN:2378-2250
DOI:10.1109/ITC51657.2024.00023