C3 IoT Case Studies Optimizing Semiconductor Manufacturing Yield with IoT

	Optimizing Semiconductor Manufacturing Yield with IoT C3 IoT

Optimizing Semiconductor Manufacturing Yield with IoT

C3 IoT

Technology Category	Analytics & Modeling - Machine Learning Functional Applications - Manufacturing Execution Systems (MES)
Applicable Industries	Equipment & Machinery Semiconductors
Applicable Functions	Product Research & Development Quality Assurance
Use Cases	Additive Manufacturing Manufacturing Process Simulation
Services	Data Science Services Testing & Certification
Challenge	A large U.S.-based manufacturer of high-performance semiconductors was facing a significant challenge in optimizing the manufacturing process of its wireless products. The company, which designs and delivers a broad set of cutting-edge products including radio frequency filters, amplifiers, modulators, attenuators, and more, was experiencing lower than expected overall yield in some of its most complex products. This was affecting the company's productivity and profitability, and there was a need for a solution that could predict low-yield wafers early in the process and identify process improvements to increase overall yield. Read More
About Customer	The customer is a large U.S.-based manufacturer of high-performance semiconductors. The company has a broad portfolio of cutting-edge products, including radio frequency filters, amplifiers, modulators, attenuators, and more. The company has a significant global presence with 35 facilities worldwide and employs 6,000 people. The company's products are widely adopted in leading mobile devices and IoT ecosystems. The company generates $3.6 billion in revenue annually. Read More
Customer Name	Not disclosed Read More
Solution	The manufacturer implemented C3 AI Process Optimization to address the challenge. This solution enabled the company to identify bad wafers, quantify time and costs saved, and tune design and manufacturing processes to optimize yields. The implementation of the project took 10 weeks from kickoff to pre-production application. During this period, a 3 terabyte unified data image of 830,000 files was created, along with more than 1,500 relevant features. The company also built 20 machine learning algorithms to predict die quality and 12 machine learning algorithms to predict the bottom 10% and 20% of low-yield wafers. The C3 AI Process Optimization user interface was configured to deliver insights on manufacturing optimization based on analysis. Read More Log in to view content
Contents

Technology Category

Analytics & Modeling - Machine Learning

Functional Applications - Manufacturing Execution Systems (MES)

Applicable Industries

Equipment & Machinery

Semiconductors

Applicable Functions

Product Research & Development

Quality Assurance

Use Cases

Additive Manufacturing

Manufacturing Process Simulation

Services

Data Science Services

Testing & Certification

Challenge

A large U.S.-based manufacturer of high-performance semiconductors was facing a significant challenge in optimizing the manufacturing process of its wireless products. The company, which designs and delivers a broad set of cutting-edge products including radio frequency filters, amplifiers, modulators, attenuators, and more, was experiencing lower than expected overall yield in some of its most complex products. This was affecting the company's productivity and profitability, and there was a need for a solution that could predict low-yield wafers early in the process and identify process improvements to increase overall yield.

About Customer

The customer is a large U.S.-based manufacturer of high-performance semiconductors. The company has a broad portfolio of cutting-edge products, including radio frequency filters, amplifiers, modulators, attenuators, and more. The company has a significant global presence with 35 facilities worldwide and employs 6,000 people. The company's products are widely adopted in leading mobile devices and IoT ecosystems. The company generates $3.6 billion in revenue annually.

Customer Name

Not disclosed

Solution

The manufacturer implemented C3 AI Process Optimization to address the challenge. This solution enabled the company to identify bad wafers, quantify time and costs saved, and tune design and manufacturing processes to optimize yields. The implementation of the project took 10 weeks from kickoff to pre-production application. During this period, a 3 terabyte unified data image of 830,000 files was created, along with more than 1,500 relevant features. The company also built 20 machine learning algorithms to predict die quality and 12 machine learning algorithms to predict the bottom 10% and 20% of low-yield wafers. The C3 AI Process Optimization user interface was configured to deliver insights on manufacturing optimization based on analysis.

Impact #1	The implementation of the C3 AI Process Optimization solution has had a significant impact on the manufacturer's operations. The ability to predict low-yield wafers early in the process and identify process improvements has led to an increase in overall yield. The solution has also enabled the company to identify bad wafers, quantify time and costs saved, and tune design and manufacturing processes to optimize yields. The insights delivered through the C3 AI Process Optimization user interface have been instrumental in driving manufacturing optimization. The solution has also demonstrated the potential of AI in manufacturing, with the company now able to build machine learning algorithms to predict die quality and low-yield wafers.

Impact #1

The implementation of the C3 AI Process Optimization solution has had a significant impact on the manufacturer's operations. The ability to predict low-yield wafers early in the process and identify process improvements has led to an increase in overall yield. The solution has also enabled the company to identify bad wafers, quantify time and costs saved, and tune design and manufacturing processes to optimize yields. The insights delivered through the C3 AI Process Optimization user interface have been instrumental in driving manufacturing optimization. The solution has also demonstrated the potential of AI in manufacturing, with the company now able to build machine learning algorithms to predict die quality and low-yield wafers.

Benefit #1	$39 million estimated annual economic impact
Benefit #2	Created 3 terabyte unified data image of 830,000 files
Benefit #3	Created more than 1,500 relevant features

Benefit #1

$39 million estimated annual economic impact

Benefit #2

Created 3 terabyte unified data image of 830,000 files

Benefit #3

Created more than 1,500 relevant features

Download PDF Version

Overview

Optimizing Semiconductor Manufacturing Yield with IoT

Operational Impact

Quantitative Benefit