Altis Semiconductor Improves Product Quality with SAS®

Two former IBM managers specializing in the semi-microelectronics industry formed Altis Semiconductor in Corbeil-Essonnes, France in July of 1999. The company has created a powerful and reliable quality control program using SAS solutions to improve the electrical performance of its products.

Altis Semiconductor is now one of the most important copper production high-tech companies in the world (0.18 µm and higher). Altis controls all of the building blocks required for the development of S.O.C. (the single-chip or "System on a Chip" concept) including analog and digital transistors, dense SRAM (standard random access memory), on-board DRAM (dynamic random access memory), and flash memories.

The organization is a standout beyond its classic role as a "caster" with its offering of high-level engineering services: production and accelerated prototype testing, detailed fault analysis, and performance improvement. SAS offered Altis Semiconductor assistance in order to achieve its objectives in those areas.

Quality Improvement
The mission of the Altis Engineering Department is to ensure the introduction of the latest technology and production of new products as well as the optimization of fabrication procedures, while always taking into account intense variations in the marketplace.

In the initial phase the Engineering Department determined, as early as 1992, a need for automatic reporting systems. Their objectives were to:

• Provide access to an automatic and interactive reporting system.
• Install a query system for users.
• Provide functions created in relation to user need, allowing analysis of data in the engineering environment.

As explained by Pascal Chesneau, data analysis manager of the project: "We selected SAS solutions from the very beginning because of their capability in managing the automatic and interactive creation of menus based on automatic collection systems of all production data."

Deploying the Project with SAS
The second phase, started in 1997, included the implementation of new applications accomplished with SAS solutions. A team was created in order to develop a computation engine dedicated to test results and to migrate the existing MVS application to UNIX. SAS assisted with the implementation of the computer for the final test, followed by the installation of an interactive platform for all Engineering Department engineers.

For this phase, the objective was to create a system that allows users access to all technical and logistical databases while making available dedicated professional tools for quality control and statistical analysis.

SAS solutions are present throughout the Altis Semiconductor production process, on four main levels:

• Data warehouse: This platform, accessible to all users, covers a series of solutions using SAS technology, mapped over dedicated interfaces. These functions allow access, for example, to on-site test data and final test data, but also to all logistic events. The SAS solution allows the establishment of access parameters to the database while making the query and analysis system relatively transparent to the user. Reports are created with specific SAS tables allowing the users to perform weekly or monthly tracking calculations. Hundreds of reports are generated for all production lines. Collection, analysis, and reproduction are all performed in the SAS Data Warehouse environment.
  
  
• Data storage and structuring: Altis Semiconductor has a considerable number of logistic and technical tables in DB2 and did not think it was necessary to create new specific tables for production; native access links to the DB2 bases and the SAS tables were implemented to create the SAS Data Warehouse.
  
  
• Data Mining: Another objective, in terms of the engineering mission, was the improvement of performance and procedures as well as the investigation of factors faulted with lowering performance.
  
  "We wanted to determine which technique or model had to be developed in order to find negative factors in performance, without performing a pre-analysis by engineers. We achieved that with the help of the SAS Data Mining Solution," explains Chesneau.
  
  
• Users, with transit offered on the Java Web interface: Analysts can extract data by product and technology using the SAS query system and identify faulty operations or equipment. "From now on, the deployment and automation phase offers users an interface that allows them to set the entry parameters directly. A project is underway with SAS to create a Java Internet interface," according to Chesneau.

An Excellent Return on Investment
The project with SAS, an example of SAS Supply Chain Intelligence, has allowed Altis Semiconductor to create a technical platform that is recognized by its users. For example, the data mining solution allowed the validation of a model resulting in a 1.7 percent increase in performance, which represents considerable sums on high volumes.

Chesneau concludes: "SAS responded to our industrial problems by providing scalable, solid solutions that were accepted by all. We were able to join users and developers through a series of techniques and analysis tools. And all this contributed to the goal of reaching our objectives for quality of service, performance, and profitability."

The Assets of Altis Semiconductor:

• Almost 40 years of know-how in semiconductors, specifically logic.
• A strong concentration of expertise onsite in Corbeil-Essonnes.
• Three partners who are world leaders in their markets, who invest and install their technological skills centers for Europe on the Corbeil site (Dupont Photomasks, Air Liquide and BOC Edwards).
• Proven experience in the mass production of semiconductors (DRAM).
• Strategic geographic position for access to the European market.

Copyright © SAS Institute Inc. All Rights Reserved.