Comprehensive Analysis of Random Telegraph Signal Noise (RTN) Using a Physics-Informed Machine Learning Framework on Advanced 3-demensional NAND Memory Devices

Project description

This project aims at measuring and analysing random telegraph signal noise (RTN) on advanced 3-dimensional memory devices by using a recently developed physics-informed machine learning framework. 

RTN, caused by discrete charge trapping and de-trapping in defects inside the material, is one of the measured reliability issues on scaled devices since the amplitude of the RTN increases with decreasing device areas.  However, RTN generated by multiple defects creates very complex data, resulting in difficulties of systematic and comprehensive analysis. 

Recently, imec developed a physics-informed machine learning framework based on a Bayesian-based algorithm together with affinity propagation clustering, which can efficiently and accurately analyse complex RTN data with multiple traps. In addition to the novel analysis framework, imec has developed advanced measurement methodologies and statistical analysis techniques to extract, interpret and understand the reliability physics reflecting in RTN phenomena.

This project provides opportunities and experiences to learn the physics-informed machine learning framework, advanced electrical measurements methodologies, and statistical analysis techniques, as well as the reliability physics of novel devices.

Project tasks

• The student will measure and analyse random telegraph signal noise (RTN) on advanced 3-dimensional NAND test devices
• The student will propose possible improvements of the measurement and analysis methodologies 
• The student will extract time constants of individual charging defects, relate these properties to the materials in the NAND stack, and investigate the trends with geometry of the devices.

Type of Project: Combination of internship and thesis; Internship; Thesis 

Master's degree: Master of Engineering Technology; Master of Engineering Science; Master of Science 

Master program: Nanoscience & Nanotechnology; Materials Engineering; Physics; Electrotechnics/Electrical Engineering; Electromechanical engineering; Chemistry/Chemical Engineering 

Duration: >6 months  

Supervisor: Jan Van Houdt (EE, Nano) 

For more information or application, please contact the supervising scientist Yusuke Higashi (yusuke.higashi@imec.be).

Only for self-supporting students.