Volume diagnosis data mining

Wu-Tung Cheng; Yue Tian; Sudhakar M Reddy

doi:10.1109/ETS.2017.7968238

Back

Conference proceeding

Volume diagnosis data mining

Wu-Tung Cheng, Yue Tian and Sudhakar M Reddy

2017 22nd IEEE European Test Symposium (ETS), pp.1-10

05/2017

DOI: 10.1109/ETS.2017.7968238

View Online

Abstract

With decreasing feature sizes and increasing complexity of fabrication processes for manufacturing VLSI semiconductor devices, more systematic defects occur at the advanced technology nodes. Product yield ramp up is mostly determined by how fast systematic defects are identified and fixed. Given the long times and expense of physical failure analysis (PFA), use PFA on a large number of failing devices to find systematic defects is becoming infeasible. For this reason, volume diagnosis data mining for root cause identification based on statistical methods is used to reduce turnaround time and cost to speed up the process of systematic defect identification. The identified root cause information not only can be used to improve yield analysis but also can reduce PFA cost by focusing on failing devices with systematic defects.

Bridges

Data Mining

Games

Systematics

Bayes methods

Maximum likelihood estimation

Probability distribution

scan diagnosis

Solid modeling

volume diagnosis

yield analysis

Details

Title: Subtitle: Volume diagnosis data mining
Creators: Wu-Tung Cheng - Mentor Graphics
Yue Tian - University of Iowa
Sudhakar M Reddy - University of Iowa
Resource Type: Conference proceeding
Publication Details: 2017 22nd IEEE European Test Symposium (ETS), pp.1-10
DOI: 10.1109/ETS.2017.7968238
ISSN: 1530-1877
eISSN: 1558-1780
Publisher: IEEE
Language: English
Date published: 05/2017
Academic Unit: Electrical and Computer Engineering
Record Identifier: 9984197275602771

Metrics

14 Record Views