Semiconductor Data Monitoring

Data monitoring in the semiconductor industry is the collection and analysis of all chip manufacturing data, including test data, wafer defect inspection data, probe tests, WAT, final inspection tests and manufacturing data from the hundreds of processes that each chip undergoes. Easy access to these massive amounts of extremely complex data saves engineering time and costs.

We often get questions about what kinds of semiconductor data can be analyzed. In fact, all numerical data can be monitored, in any kind of format (STDF, ATDF, CSV. RITdb, …) and from any source: Wafer Sort, Final Test, Parameter Test (E-Test, PCM), Inline Measurement Data and Defect Data.

Our yield analysis system YieldWatchDog collects and monitors data, checking for anomalies, irregularities or early warnings of a problem. Yield excursions occur when a tool or manufacturing process begins to operate outside its predetermined specifications. If a potential problem is detected, an alert is immediately sent to an engineer, enabling a quick reaction, thereby avoiding yield excursions and waste, significantly reducing costs and optimizing yield.

Analysis of all the chip manufacturing data allows engineers to monitor the manufacturing process from start to finish and use heat maps and other enhanced visualisation tools to identify specific processes or pieces of equipment that lead to yield loss.

These integrated visualization and calculation tools, such as wafer maps, trend charts, correlation charts, histograms, line charts, Pareto analyses, fail flip maps, fail trends, fail category maps, gallery views and others, enable engineers to get a thorough understanding of manufacturing data. Product and test engineers can display the analysis data exactly as they need it to get actionable insights into the data, enabling them to tackle the issues that have the biggest impact on yield.

YieldWatchDog-XI goes even further with machine learning capabilities, uncovering hidden correlations, possibly predicting issues in manufacturing before they occur and performing root cause analysis. Yields and profits can be increased with advanced data analytics based on intelligent algorithms.

End-to-end monitoring of manufacturing data reduces the engineering resources needed, not only for product and process engineers, while it also allows for cross-functional involvement in the manufacturing processes, such as from business management and customer requirements. Data monitoring with YieldWatchDog yield analysis software provides enormous potential for efficiency improvement, workload reduction and maximizing profitability.

Data monitoring software saves manufacturers time and money, optimizing yield, while also helping to discover where to implement quality initiatives thanks to the constant easy access to data and analytics.