Deep learning-based yield prediction for the die bonding semiconductor manufacturing process

Ali Akbar, Muhammad and Jazlan, Ahmad and Mohd Ibrahim, Azhar and Ahmad, Arfah (2024) Deep learning-based yield prediction for the die bonding semiconductor manufacturing process. In: Proceedings of the 7th International Conference on Electrical, Control and Computer Engineering, Kuala Lumpur, Malaysia.

Text Deep Learning-Based Yield Prediction for the Die Bonding Semiconductor Manufacturing Process.pdf Restricted to Registered users only Download (257kB)

Abstract

In the semiconductor manufacturing industry, consistently achieving a high yield is the primary target to meet customer demands and ensure continuous profitability. The ability to predict the yield of a particular manufacturing process at either the Front of Line or End of Line facilities is therefore essential in order to analyze Return of Investments (ROI), predictive maintenance and condition monitoring. However, achieving high quality predictions with good accuracy is challenging due to the various uncertainties in the manufacturing process such as unexpected machine downtime and stoppage for maintenance. In this paper we propose a method using Deep Learning Long Short-Term Memory Recurrent Neural Networks (LSTM-RNN) to perform day ahead forecasting of the yield from the Die Bond process at a particular semiconductor manufacturing facility. The method was implemented using MATLAB software, and the results demonstrate that the proposed approach achieves accurate yield forecasts with less than 8% error. Further improvements can be made by utilizing hourly data instead of daily data.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	Die bonding, Long short-term memory recurrent neural networks, Semiconductor manufacturing process
Divisions:	Faculty Of Electrical Technology And Engineering
Depositing User:	NUR FARISAH JAFRIN
Date Deposited:	23 Jul 2026 00:26
Last Modified:	23 Jul 2026 00:26
URI:	http://eprints.utem.edu.my/id/eprint/29876
Statistic Details:	View Download Statistic

Actions (login required)

View Item