Semiconductor device manufacture procedure prediction system and method

Abstract

The present invention provides a semiconductor device manufacture procedure prediction system and method, and relates to the technical field of semiconductors. The system comprises a data acquisition module, a data conversion module and a result prediction module with a neural network model. The method comprises the steps of collecting prediction-related original data of to-be-predicted information, converting the original data into computable standardized data, and calculating a to-be-predicted information prediction result by using a neural network prediction model. With adoption of the system and the method, the to-be-predicted information prediction result can be timely calculated by using the neural network model according to on-line data, thereby preventing major reliability and / or yield problems occurred in a semiconductor device manufacture procedure.

Description

technical field [0001] The invention relates to the technical field of semiconductors, in particular to a semiconductor device manufacturing process prediction system and method. Background technique [0002] In the field of semiconductor technology, yield and reliability are two important factors affecting the development and profitability of the semiconductor manufacturing industry. Generally, the reliability of a semiconductor device is calculated and obtained according to test results after the manufacturing process of the semiconductor device is completed. Similarly, the yield rate of semiconductor devices can only be obtained from the wafer test results after the process is completed. [0003] Since yield and reliability are very important factors to profitability, how to accurately predict reliability risk and yield is a necessary factor for semiconductor device manufacturers to ensure profitability. [0004] Currently, assessments of reliability risk and yield can ...

AI Technical Summary

Problems solved by technology

This can lead to high-risk wafers being eventually scrapped due to risk, which in turn increases the cost of the semiconductor device manufacturing process

[0005] It can be seen that there is no solution in the prior art to prevent major reliability or yield problems in the manufacturing process of semiconductor devices through real-time prediction of inline data

Images