Quality in 3D assembly — Is “Known Good Die” good enough?

There are numerous papers about 3D packages; however 3D has not been able to conquer the volume production status. One common issue is how to ensure quality across the overall and more complex supply chain. The Known Good Die (KGD) concept can only partially solve this. Advanced packaging processes not only require total reliability of the overall supply chain but also require advanced models to reduce the risk of unprofitable output caused by the packaging process itself. This also includes an additional perspective: process control. In traditional semiconductor production, component test (for KGD) and final test for ultimate functional quality assurance before shipment are sufficient to ensure quality and cost-efficiency. Advanced packaging methods such as 3D go beyond this. Currently, a distributed test flow is discussed in the industry, which compares the cost of test to the cost of non-testing, with special references to the packaging process. Opportunities and challenges of these new test strategies need to be considered to find the best models for these test distributions. Special equipment requirements need to be postulated. These new approaches will no longer be limited to the classic test areas - probing and final test. As many of the additional risks are related to the assembly process, this is an area where additional test insertions have to be considered. These test points will have different requirements than the established ones. The changed environment needs to be taken into account in terms of the test process itself. Secondly, the device to be tested will be very sensitive because it is not packaged in the end. Discussing the most appropriate equipment includes: What are the limitations of using probing tools or deploying final test equipment? Which strategy will offer the most synergy and reduce cost of test in the end? Using an analogy with the MEMS gives an interesting perspective on how to leverage the expertise that has been gained during the- last decade.