image

For some semiconductor manufacturers it is important to be able to track and trace the status of dice as they travel through the test supply chain. In the event of quality issues, the ability to quickly and accurately identify a defective die, analyze the root cause and contain impacted modules and systems is critical. For these yieldHUB’s customers, each die is traced through the test process and is readable later should there be an issue in the field. Each die’s history is then available for analysis in the yieldHUB database well after the die was released into the field. This traceability data is kept in the yieldHUB central database for that customer.

yieldHUB uses the traceability data to cleanse the data in wafer sort and final test. It’s easy to detect when a die is retested for example in post-packaging final test so that only one version of the test results of a die is recorded for population and drift analysis.

You can also search the database for a unique ID. The unique ID can then be analysed across all tests and compared with the general population of dice.

Traceability information can also help create wafer representations in Final Test, even allowing the direct comparisons of wafer sort or CP with final test results. “Bin flip” analysis is important for yield improvement and can be done also in final test using this method.

Such traceability can also combat counterfeiting. When the test results for each device are stored in a central database like yieldHUB and each device is traceable, any potentially counterfeit device can be verified as such.

John O'Donnell, CEO and founder, yieldHUB

“Reliability and traceability are essential in AI running in the cloud, blockchain, automotive, 5G. You need very fast traceability aligned with easy extraction to run machine learning software, along with the ability to factor in whatever environmental variables you may have.”