Integration of all manufacturing data and wafer-testing

The integration of manufacturing data and wafer testing is essential for a number of reasons related to yield improvement and the optimization of the manufacturing process.

In a top-down diagnostics approach, the detection of production or wafer-test related anomalies starts with the final test results. Due to the close relationship between all the steps in a process, there needs to be tight integration of the essential data with automated diagnostics for the important relationships. This provides engineers with a fast overview of the associations between data and allows them to use this knowledge for model development or modification.

CenterPoint’s products support a wide range of equipment integration standards and wafer final test standards, including SEMI EDA, SECSII, KLARF, STDF V4, SEMI G85. The seamless integration of these standards allows solutions for data collection to be uniformly set up throughout the manufacturing process.

CenterPoint’s products provide various off-the-shelf data-mining techniques integrated in an automated learning concept that supports the merging of online and/or offline data from:

• Process Tools
• Metrology
• Wafer Electrical Tests
• Wafer Final Tests

Interactive tools provide filtering capabilities and investigate the strength of associations between data from different sources. An integrated expert system framework enables automated decision making.

Multi-user access to the engineering knowledge provided by the CenterPoint solution leads to collaborative use and further development of the corporate engineering knowledge management.

The exchange of stored engineering knowledge between different installations and sites enables higher flexibility in process transfer and reuse of knowledge.

The integration of manufacturing and wafer test data provides a number of opportunities for the further optimization of the overall manufacturing process.

Reduction of Test Time
The association between process data, metrology data and defect measurements leads to a final testing scenario where the performing of single tests or groups of tests is dependent on the appropriate manufacturing results. Under certain conditions, this allows test times to be reduced by up to 20%.

Faster Root-Cause Analysis
CenterPoint’s solutions provide off-the-shelf merging algorithms for the correlation of process tool data, metrology data, wafer electrical test results, defect measurement results, wafer final tests and binning results.

Automated research into the associations between the multiple input parameters is an integral feature based on the correlation results. This provides engineers with an immediate overview of the interdependence and strength of interdependence between single parameters and groups of parameters.

Sets and groupings of parameters can be customized using intuitive editors and powerful scripting support (Python Script or CenterPoint Script Language).

Within the semiconductor industry measurement data is generated for a vast number of different sources. Due to the complexity of the semiconductor manufacturing process, these measurements are often closely related to each other and the relationships are becoming increasingly difficult to model based on analytical approaches.

CenterPoint’s APC\Platform therefore provides alternative algorithms based on data-mining techniques to enhance and accelerate the process of gaining knowledge about data associations and root causes.

Furthermore, this knowledge held electronically can be transformed into error classification models and decision making rules that are applied automatically to factory-wide online data for the purpose of real-time health monitoring.