US-20260126772-A1 - SYSTEM, DEVICE, AND METHOD FOR A MANUFACTURING PROCESS

US20260126772A1US 20260126772 A1US20260126772 A1US 20260126772A1US-20260126772-A1

Abstract

Provided is a system, device, and method. The method includes estimating, using a first model, respective first yields of first wafers based on first data on a plurality of factors about a semiconductor manufacturing process regarding the plurality of first wafers, the first data obtained before the semiconductor manufacturing process of the plurality of first wafers has been completed, generating, using a second model generated based on the first model, respective yield contribution values of the plurality of factors based on the first data, data on the estimated respective first yields of the plurality of first wafers, second data on the plurality of factors regarding a plurality of second wafers of which the semiconductor manufacturing process has completed, and determining one or more factors, among the plurality of factors, that contribute to a yield reduction based on the generated respective yield contribution values of the plurality of factors.

Inventors

Jaewon Lee
Namyeong Kwon

Assignees

SAMSUNG ELECTRONICS CO., LTD.

Dates

Publication Date: 20260507
Application Date: 20250509
Priority Date: 20241104

Claims (20)

1 . A processor-implemented method with respect to a semiconductor manufacturing process, the method comprising: estimating, using a first model, respective first yields of a plurality of first wafers based on first data on a plurality of factors about the semiconductor manufacturing process regarding the plurality of first wafers, the first data obtained before the semiconductor manufacturing process of the plurality of first wafers has been completed; generating, using a second model generated based on the first model, respective yield contribution values of the plurality of factors based on the first data, data on the estimated respective first yields of the plurality of first wafers, second data on the plurality of factors regarding a plurality of second wafers of which the semiconductor manufacturing process has completed, and data on respective second yields of the plurality of second wafers; and determining one or more factors, among the plurality of factors, that contribute to a yield reduction based on the generated respective yield contribution values of the plurality of factors.
2 . The method of claim 1 , wherein the plurality of factors include factors regarding a plurality of equipment pieces used in the semiconductor manufacturing process and factors regarding a plurality of measurement items that are measured in the semiconductor manufacturing process, wherein data, among the first data, regarding one wafer of the plurality of first wafers includes category data on one or more pieces of first equipment used among the plurality of equipment pieces in a first process of the semiconductor manufacturing process performed with respect to the one wafer, category data on one or more pieces of second equipment predetermined among the plurality of equipment pieces to be used in a second process of the semiconductor manufacturing process with respect to the one wafer, measurement data on one or more first measurement items, among the plurality of measurement items, that have been measured with respect to the one wafer in the semiconductor manufacturing process, and measurement data on one or more second measurement items, among the plurality of measurement items, predetermined to be measured with respect to the first wafer in the semiconductor manufacturing process, and wherein another data, among the second data, regarding another one wafer of the plurality of second wafers includes category data on the plurality of equipment pieces with respect to the other one wafer and measurement data on the plurality of measurement items with respect to the other one wafer.
3 . The method of claim 2 , wherein the category data on the one or more pieces of second equipment includes corresponding data obtained based on process history of the one or more pieces of second equipment, and wherein the measurement data on the one or more second measurement items includes other corresponding data obtained based on process history of the one or more second measurement items.
4 . The method of claim 1 , further comprising training the first model based on the second data and the data on the respective second yields of the plurality of second wafers.
5 . The method of claim 1 , wherein the first model comprises a neural network including at least a plurality of layers, and wherein the method further comprises generating the second model based on weights of the plurality of layers, where the second model comprises an explainable artificial intelligence (XAI) model.
6 . The method of claim 1 , wherein the determining of the one or more factors that contribute to the yield reduction comprises: identifying date-dependent changes in respective yield reduction contribution rankings of the plurality of factors, based on the generated respective yield contribution values of the plurality of factors; and identifying, as the determined one or more factors that contribute to the yield reduction, corresponding one or more factors, among the plurality of factors, of which the respective yield reduction contribution rankings rise based on the date-dependent changes in the respective yield reduction contribution rankings of the plurality of factors.
7 . The method of claim 6 , wherein the identifying of the date-dependent changes in the respective yield reduction contribution rankings of the plurality of factors comprises: generating a plurality of wafer sets by grouping wafers for which an identical process in the semiconductor manufacturing process was performed on an identical date among the plurality of first wafers and the plurality of second wafers; identifying the generated respective yield contribution values of the plurality of factors for each of the plurality of wafer sets; and determining the respective yield reduction contribution rankings of the plurality of factors for each of the plurality of wafer sets based on the identified generated respective yield contribution values.
8 . The method of claim 7 , wherein a yield reduction contribution ranking of a first factor for a first wafer set among the respective yield reduction contribution rankings of the plurality of factors for each of the plurality of wafer sets includes one of: a yield reduction contribution ranking corresponding to a date on which a first process of the semiconductor manufacturing process corresponding to the first factor is executed for the first wafer set; a yield reduction contribution ranking corresponding to a date on which a second process of the semiconductor manufacturing process corresponding to the first factor is predetermined to be executed for the first wafer set; a yield reduction contribution ranking corresponding to a completion date of the first wafer set with respect to the semiconductor manufacturing process; and a yield reduction contribution ranking corresponding to an expected completion date of the first wafer set with respect to the semiconductor manufacturing process.
9 . The method of claim 7 , wherein the determined respective yield reduction contribution rankings of the plurality of factors for a first wafer set, among the determined respective yield reduction contribution rankings of the plurality of factors for each of the plurality of wafer sets, are determined to be higher as the identified respective yield contribution values of the plurality of factors for the first wafer set are lower.
10 . The method of claim 7 , wherein the identified generated respective yield contribution value of a first factor for a first wafer set, among the identified generated respective yield contribution values of the plurality of factors for each of the plurality of wafer sets, is determined as an average value of the generated respective yield contribution values of the first factor corresponding to wafers included in the first wafer set.
11 . The method of claim 1 , wherein the plurality of first wafers include one wafer for which the semiconductor manufacturing process is set to be completed within a first period of time with reference to a set date, and wherein the plurality of second wafers include another one wafer for which the semiconductor manufacturing process was completed within a second period of time with reference to the set date.
12 . The method of claim 1 , further comprising providing a user terminal with information on the determined one or more factors, wherein the information on the determined one or more factors comprises at least one of information indicating date-dependent change in yield reduction contribution rankings of the determined one or more factors, information on equipment corresponding to the determined one or more factors, or information corresponding to a measurement item corresponding to the determined one or more factors.
13 . A non-transitory computer-readable recording medium storing code, which when executed by one or more processors, configures the one or more processors to execute the method of claim 1 .
14 . An electronic device comprising: a memory storing code; and one or more processors configured to execute the code, wherein, execution of the code by the one or more processors, configures the one or more processors to: estimate, using a first model, respective first yields of a plurality of first wafers based on first data on a plurality of factors about the semiconductor manufacturing process regarding the plurality of first wafers, the first data obtained before the semiconductor manufacturing process of the plurality of first wafers has been completed; generate, using a second model generated based on the first model, respective yield contribution values of the plurality of factors based on the first data, data on the estimated respective first yields of the plurality of first wafers, second data on the plurality of factors regarding a plurality of second wafers of which the semiconductor manufacturing process has completed, and data on respective second yields of the plurality of second wafers; and determine one or more factors, among the plurality of factors, that contribute to a yield reduction based on the generated respective yield contribution values of the plurality of factors.
15 . The electronic device of claim 14 , wherein the plurality of factors include factors regarding a plurality of equipment pieces used in the semiconductor manufacturing process and factors regarding a plurality of measurement items that are measured in the semiconductor manufacturing process, wherein data, among the first data, regarding one wafer of the plurality of first wafers includes category data on one or more pieces of first equipment used among the plurality of equipment pieces in a first process of the semiconductor manufacturing process performed with respect to the one wafer, category data on one or more pieces of second equipment predetermined among the plurality of equipment pieces to be used in a second process of the semiconductor manufacturing process with respect to the one wafer, measurement data on one or more first measurement items, among the plurality of measurement items, that have been measured with respect to the one wafer in the semiconductor manufacturing process, and measurement data on one or more second measurement items, among the plurality of measurement items, predetermined to be measured with respect to the first wafer in the semiconductor manufacturing process, and wherein another data, among the second data, regarding another one wafer of the plurality of second wafers includes category data on the plurality of equipment pieces with respect to the other one wafer and measurement data on the plurality of measurement items with respect to the other one wafer.
16 . The electronic device of claim 15 , wherein the category data on the one or more pieces of second equipment includes corresponding data obtained based on process history of the one or more pieces of second equipment, and wherein the measurement data on the one or more second measurement items includes other corresponding data obtained based on process history of the one or more second measurement items.
17 . The electronic device of claim 14 , wherein the execution of the code further configures the one or more processors to train the first model based on the second data and the data on the respective second yields of the plurality of second wafers.
18 . The electronic device of claim 14 , wherein the first model comprises a neural network including at least a plurality of layers, and wherein the execution of the code further configures the one or more processors to generate the second model based on weights of the plurality of layers, where the second model comprises an explainable artificial intelligence (XAI) model.
19 . The electronic device of claim 14 , wherein, for the determining of the one or more factors that contribute to the yield reduction, the execution of the code configures the one or more processors to: identify date-dependent changes in respective yield reduction contribution rankings of the plurality of factors, based on the generated respective yield contribution values of the plurality of factors; and identify, as the determined one or more factors that contribute to the yield reduction, corresponding one or more factors, among the plurality of factors, of which the respective yield reduction contribution rankings rise based on the date-dependent changes in the respective yield reduction contribution rankings of the plurality of factors.
20 . The electronic device of claim 19 , wherein, for the identifying of the date-dependent changes in the respective yield reduction contribution rankings of the plurality of factors, the execution of the code configures the one or more processors to: generate a plurality of wafer sets by grouping wafers for which an identical process in the semiconductor manufacturing process was performed on an identical date among the plurality of first wafers and the plurality of second wafers; identify the generated respective yield contribution values of the plurality of factors for each of the plurality of wafer sets; and determine the respective yield reduction contribution rankings of the plurality of factors for each of the plurality of wafer sets based on the identified generated respective yield contribution values.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit under 35 USC § 119(a) of Korean Patent Application No. 10-2024-0154167, filed on Nov. 4, 2024, in the Korean Intellectual Property Office, the disclosure of which is incorporated herein by reference in its entirety. BACKGROUND 1. Field One or more embodiments relate to a system, device, and method for a manufacturing process. 2. Description of Related Art The semiconductor industry has diversified and become more sophisticated in demand, from personal computers in the past to mobile devices, home appliances, and automobiles, along with the development of the industry. Further, there is a higher integration of semiconductor devices in such devices. Semiconductor circuitries are being formed with ultra-fine line widths. SUMMARY This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter. In one general aspect, a processor-implemented method with respect to a semiconductor manufacturing process includes estimating, using a first model, respective first yields of a plurality of first wafers based on first data on a plurality of factors about a semiconductor manufacturing process regarding the plurality of first wafers obtained before the semiconductor manufacturing process of the plurality of first wafers has been completed, generating, using a second model generated based on the first model, respective yield contribution values of the plurality of factors based on the first data, data on the estimated respective first yields of the plurality of first wafers, second data on the plurality of factors regarding a plurality of second wafers of which the semiconductor manufacturing process has completed, and data on respective second yields of the plurality of second wafers, determining one or more factors, among the plurality of factors, that contribute to a yield reduction based on the generated respective yield contribution values of the plurality of factors. The plurality of factors may include factors regarding a plurality of equipment pieces used in the semiconductor manufacturing process and factors regarding a plurality of measurement items that may be measured in the semiconductor manufacturing process, where data, among the first data, regarding one wafer of the plurality of first wafers may include category data on one or more pieces of first equipment used among the plurality of equipment pieces in a first process of the semiconductor manufacturing process performed with respect to the one wafer, category data on one or more pieces of second equipment predetermined among the plurality of equipment pieces to be used in a second process of the semiconductor manufacturing process with respect to the one wafer, measurement data on one or more first measurement items, among the plurality of measurement items, that have been measured with respect to the one wafer in the semiconductor manufacturing process, and measurement data on one or more second measurement items, among the plurality of measurement items, predetermined to be measured with respect to the first wafer in the semiconductor manufacturing process, and where another data, among the second data, regarding another one wafer of the plurality of second wafers may include category data on the plurality of equipment pieces with respect to the other one wafer and measurement data on the plurality of measurement items with respect to the other one wafer. The category data on the one or more pieces of second equipment may include corresponding data obtained based on process history of the one or more pieces of second equipment, and the measurement data on the one or more second measurement items may include other corresponding data obtained based on process history of the one or more second measurement items. The method may further include training the first model based on the second data and the data on the respective second yields of the plurality of second wafers. The first model may include a neural network including at least a plurality of layers, and the method further include generating the second model based on weights of the plurality of layers, where the second model may include an explainable artificial intelligence (XAI) model. The determining of the one or more factors that contribute to the yield reduction may include identifying date-dependent changes in respective yield reduction contribution rankings of the plurality of factors, based on the generated respective yield contribution values of the plurality of factors, identifying, as the determined one or more factors that contribute to the yield reduction, corresponding one or more factors, among the plurality of factors, of which the