Solving yield management problems is what we do best. Here are some of our most sought-after solutions:
Microchips have memory on board where a Unique Identification (UID) can be stored. yieldHUB’s traceability solution allows users to set up a recipe to interpret the data to build the UID information. This can include wafer lot ID, wafer ID, and wafer XY coordinates. With our platform, every chip is searchable instantly. If you get an RMA and serial number, we can find the entire manufacturing dataset! We give users the power to create wafer maps from final test data, allowing analysis such as bin flip or indexed correlation to identify changes in performance between wafer sort and final test.
In semiconductor manufacturing there are huge volumes of data that come from various different sources and need to be analyzed by different people. The yieldHUB data pipeline and VIPER (VIrtual Product Engineer) do not only automate the management of that data but also the cleansing. It is important to be using clean data for analysis otherwise the conclusions reached might not be reliable. The data cleansing includes both standard data formats such as STDF and custom data formats used by our customers.
Parametric / Bin analysis
Data analysis is at the heart of any yield management strategy. yieldHUB offers top quality capabilities and reports for analyzing parametric and bin data. The capabilities go beyond the basic analysis of these data types and include the ability to search for wafers or lots with particular bin or parametric signatures; as well as creating trends of bin or parametric performance across hundreds or thousands of wafers in seconds.
Characterization data can be far more complex than production data. This is because the test conditions are encoded in the data and each test can be tested under many conditions in characterization and therefore generates a lot of data. The yieldHUB platform has a sophisticated solution to automate the extraction of the conditions and provide a thorough characterization of each test with minimal setup.
Outlier detection is primarily aimed at the automotive and aerospace industries with such algorithms as Dynamic Part Average Testing (DPAT) and Static Part Average Testing (SPAT). Special algorithms such as GDBN (and for example Lot Norm in the case of aerospace) can also be made available. yieldHUB’s outlier detection capabilities are seamless and efficient. When the AECQ standard algorithms do not fit your data yieldHUB allows you to add your own algorithms so the possibilities are limitless.
WAT / PCM
We receive a lot of enquiries about Wafer Acceptance Test data (WAT) or Process Control Monitoring data (PCM). yieldHUB has parsers (smart code to ingest data) for WAT / PCM data from all major Fabs and has specialist charts and reports developed specifically for semiconductor WAT data.
Separation of engineering & production
yieldHUB allows both production and engineering data to be analyzed. Production data is protected, controlled, cleansed and automatically reported on. Engineering data is easily uploaded by any user, is flexible and editable. These are often managed separately but they can be compared and analyzed together.
Automated reports and notifications
Regular automated reports are great for monitoring. But when something does go wrong (e.g. a low yield, or a parametric shift) it is important to know as soon as possible. This is why getting notifications are key. On the yieldHUB platform, notifications can be set-up in seconds.
Integrated communication tools
With yieldHUB, users can communicate across time-zones and locations more easily. You can add and send comments through the system itself, share reports and send data at the touch of a button. yieldHUB enables you to communicate with your supply chain worldwide.
Frequently Asked Questions
Is there a limit on the number of tests you support?
We have customers with 20000 tests in a datalog. We have yet to see an upper limit on how many tests the yieldHUB systems can support.
Do you support test data from modules and systems?
Yes, we have several customers who have data from modules and systems and you can analyse this data in yieldHUB.
Will yieldHUB allow me to see which tests correlate with each other the most? I need to remove some tests to save on test time
Yes, you can generate a Highest Correlations report from any datalog and you will definitely find useful information to reduce your test time. You can also see which tests correlate the most with any given test.
Do you support OTP?
Yes, we support OTP/Fuse ID for traceability.
We are more than happy to be working with yieldHUB for the next five years. Their big data software and worldwide support has allowed us to manage our yields effectively over the years and their solutions have evolved with our needs.Julie Holland, Vice President, Corporate Operations of Diodes