US 11,791,184 B2
Semiconductor fabrication process and method of optimizing the same
Jiho Kim, Suwon-si (KR); Minhyeok Kwon, Suwon-si (KR); Shigenobu Maeda, Seongnam-si (KR); Jooyeok Seo, Suwon-si (KR); and Minuk Lee, Suwon-si (KR)
Assigned to SAMSUNG ELECTRONICS CO., LTD., Yongin-si (KR)
Filed by SAMSUNG ELECTRONICS CO., LTD., Suwon-si (KR)
Filed on Apr. 13, 2022, as Appl. No. 17/719,722.
Claims priority of application No. 10-2021-0097892 (KR), filed on Jul. 26, 2021.
Prior Publication US 2023/0023762 A1, Jan. 26, 2023
Int. Cl. H01L 21/67 (2006.01); G05B 19/418 (2006.01)
CPC H01L 21/67276 (2013.01) [G05B 19/41865 (2013.01); G05B 2219/32291 (2013.01); G05B 2219/33034 (2013.01); G05B 2219/45031 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A non-transitory computer-readable medium storing program code for optimizing a semiconductor fabricating process, the program code including a framework and a machine learning model, the program code, when executed by a processor, causing the processor to:
input fabrication data including a plurality of parameters associated with the semiconductor fabricating process to the framework to generate a first class for analyzing the fabrication data, wherein the parameters are different from one another;
extract a first parameter targeted for analysis a second parameter associated with the first parameter from the plurality of parameters to generate a second class for analyzing the first parameter as a sub class of the first class;
modify the first parameter and the second parameter into a data structure having a format appropriate for storage in the second class;
perform data analysis on the data structure to calculate a correlation between the first parameter and the second parameter;
transform the first parameter and the second parameter into a tensor when the correlation exceeds a threshold and otherwise transform the first parameter and a third parameter of the parameters associated with the first parameter into the tensor;
input the tensor to the machine learning model to perform a machine learning algorithm for predicting a characteristic of the semiconductor fabricating process; and
change the semiconductor fabricating process based on the predicted characteristic.