DocMemory
 
Home
News
Products
Shop
Memory
Corporate
Contact
 

News
Industry News
Publications
CST News
Help/Support
Member Area
Tester Brochure
Demo Library
Software
Tester FAQs

biology medicine news product technology definition

Tuesday, December 12, 2017
Memory Industry News
Email ArticlePrinter Format PreviousNext

Micron use AI and Big Data to improve yield


Tuesday, October 31, 2017

Emerging technologies, such as cloud, mobile, and social media, have driven exponential growths in global data, propelled information system reforms, and supported business transformation. To promptly respond to market developments, an increasing number of manufacturers are optimizing processes, capabilities to increase product values through new technologies. "To swiftly respond to market needs," said Micron Technology global front-end operations VP Buddy Nicoson, "many manufacturers are improving quality, yield rates, productivity, and cost through big data analytics. Micron is no exception."

Besides big data analytics project, Micron has recruited a group of global data scientists to work closely with partners to introduce appropriate AI and other data analytics solutions and implement advanced processes.

New technologies increase the speed to deconstruct and reconstruct industry chains. Companies are required to rebuild partnerships and ecosystems amid innovations and transformations, including Micron. At the initial stage in big data analytics project, Nicoson continued, Micron coordinates with partners to collect various data in wafer foundries. After validating benefits and returns on investment with smaller trials, company executives will align with big data analytics projects, and expand application coverage based on business needs.

Ramping up productivity and yield rates

Through AI and big data analytics solutions, Micron is able to optimize performances in quality, yield rate, output, production cycle and operating cost.

For quality, Micron's Remote Operations Center deploys mechanisms, such as Sensor Based Fault Detection, Predictive Maintenance, Real Time Process Control, and Predictive Analytics, in foundries to enhance operational efficiency by 35%.

For yield rate, Micron's system automatically identifies and categorizes defects on wafers through deep learning, and evaluates root causes as process or production tools. Analysis results and suggested solutions are later sent to team members to correct issues and improve yield rates.

For output, Micron collects and analyzes engineering and operational data, such as facility operation, error test results, and process control, to monitor real-time production status in foundries. This has helped improve performances and quality management by adjusting and optimizing all production lines.

For production cycle, Micron optimizes schedules by analyzing wafer production processes and demand projection.

For operating cost, Micron collects and analyzes unstructured data, such as wafer production wastes, to predict needs. Micron also progressively reduces operating costs via component management, waste reduction, and lower power consumption.

These five aspects all have contributed to higher yield rates and output. Micron is also encouraging foundries to learn from one another. Foundries in Japan, Singapore, Taiwan and the US can access to big data analytics performances in each site for reference, in order to improve at the same time. Through AI and big data analytics, Micron can manage and optimize each foundry via its remote operations center, Nicoson indicated.

"We don't have to send people every time when issues occur," said Nicoson. "Our staff in remote operations center can access real-time information via dashboard, arrange predictive maintenance, and ensure agile adjustments in response to customer needs."

Maximizing big data analytics performance

"Besides introducing new technologies such as AI and big data analytics, it's also important to assist employee transformation and restructure partner ecosystem," Nicoson explained. Big data analytics project helps Micron shift wafer management task from manual work to automation, and centralize foundry management to remote operations center. It also integrates distributed systems into an integrated solution, and starts to make decisions and overcome challenges through data analytics. To ensure successful transformations, Micron has launched a Job Transformation program. With data analytics courses, employees around the world can collaborate to learn more about big data and other new technologies faster. They can brainstorm together to propose better business operation methods, and implement with new technologies.

For example, after data analytics courses, a process operator, who has been working in Micron for 24 years, now knows about algorithms and uses predictive analytics results from remote operations center for daily tasks. By knowing to request maintenance before equipment encounters problems, this process operator has successfully transformed to a process technician for more sophisticated tasks, and served as a mentor to new employees.

Besides investing funds and resources for internal developments, Micron is actively working with external partners to aggregate information and operational technology data, such as advance process control, big data analytics, facility operation, and error test, for advanced manufacturing. It ensures Micron to allocate resources in time to respond to fast-changing market.

When it comes to AI and big data analytics plans in the future, Nicoson stated that "AI and big data analytics have factually enhanced our market responsiveness. Micron will continue to invest in our team members' abilities for data analyzation, and gradually optimize the performance of each foundry."

By: DocMemory
Copyright 2017 CST, Inc. All Rights Reserved

Email ArticlePrinter Format PreviousNext
Latest Industry News
Samsung spending $26 billion to dominate the memory market12/11/2017
Nvidia offers graphic card for AI on desktop12/11/2017
Toshiba reached agreement to settle dispute with WD12/11/2017
MediaTek and Qualcomm to embrace entry level Android Go 12/11/2017
Samsung to hire 1000 "cream-of-the-crop" engineers in India12/8/2017
TSMC estimate 3nm setup to cost $20 billion12/8/2017
Phison to ship PCIe 3.0 NVMe controllers12/8/2017
Broadcom's strong profit to help acquisition fight12/8/2017
Chip manufacturers are seeking high-voltage MOSFET production supports 12/7/2017
Volkswagen senior manager to get jail term12/7/2017

CST Inc. Memory Tester DDR Tester
Copyright © 1994 - 2017 CST, Inc. All Rights Reserved