Test Strategies In The Era Of Heterogeneous Integration

Moore’s Law, the observation that the number of transistors on an integrated circuit doubles approximately every two years, is critical to advances in computing technology. For decades, fabs have managed to achieve exponential growth in digital capability and transistor density by making transistors smaller and smaller, but we’ve hit the physical limits of these processes. Today, new process technologies and advanced packaging solutions, like chiplets, are allowing the industry to continue the processing capabilities and digital scaling of Moore’s Law. As early as 1965, even Gordon Moore noted: “It may be more economical to build large systems out of smaller functions that are individually packaged and interconnected.”

Although they have become more ubiquitous in recent years, chiplets face a number of challenges in design, manufacturing, packaging and test. Because of this, it is imperative to optimize the design-manufacture-test loop to continue to reduce defect escape rates and cost of test while achieving the desired yield targets and quality levels.

Optimizing the total cost of quality is critical

When dealing with more complex test processes such as known good die (KGD) testing, final test, and system level test, strategies to optimize total cost of quality are critical. Key points to consider include:

• In the early stages of the design process, designers and test engineers need to use common tools to collaborate for chip verification and fault debugging
• Moving some tests earlier in the overall process reduces defects before KGD integration
• Reducing costs by deferring some tests to later in the manufacturing process
• As manufacturing processes mature and stabilize, applying analytics to adjust the test processes before and during high volume manufacturing will optimize the overall cost of quality

Defect escapes lead to exorbitant scrap costs

Compared with traditional monolithic devices, the design and manufacturing process for chiplets is significantly different. The scrap costs associated with manufacturing traditional monolithic semiconductor devices is basically linear, including single chip cost, packaging, and assembly costs.

Manufacturing processes for 2.5D/3D designs differ significantly in terms of the accumulation of scrap costs. Specifically, these costs increase geometrically from fabrication to assembly driven by scrap costs for multiple dies, multi-chip partial assemblies, and/or full 2.5D/3D packages.

Shift left or shift right?

While 2.5D/3D packaging is an enabler of the next generation of Moore’s Law, the economic viability of this approach requires reduced defect escape rates early in the manufacturing process in an effort to reduce scrap costs. Shifting tests, either left or right, in the test process is a strategy to achieve these goals and minimize the overall manufacturing cost of 2.5D/3D components. Shift left is the ability to increase test coverage earlier in the manufacturing process (e.g., during wafer inspection and partial packaging) to maximize KGD, while reducing future packaging costs. Additional tests can also be added to the process to identify new failure types or failure modes.

However, the benefits of shift left need to be weighed. For example, increasing test intensity early in the manufacturing process can positively impact known good devices but it can also lead to an increase in test costs that is not sufficiently offset by the optimizations, even after accounting for the resulting reduction in scrap costs.

Shift right means increasing test coverage later in the manufacturing process, expanding the ability to detect defects, and maintaining quality levels with the goal of reducing costs with higher parallelism testing.

Typically, a test item with a higher yield on wafer or mission pattern tests, or a high yield test that requires a longer scan test time is an ideal candidate for shifting right. These tests can be moved to final or system level test, or flexibly managed in between. The high level of parallelism achieved with system level test for example delivers on the promise of economic improvements that can only be achieved with multisite testing, further reducing costs while achieving quality goals.

The goal of shifting tests to the left or right is to achieve the optimal combination of quality and yield throughout the entire manufacturing process, ultimately optimizing the overall cost of quality. Specific strategies include:

• Minimizing scrap costs by reducing the defect escape rate in the wafer probe process
• Realizing mass production testing in the most efficient way to reduce test costs
• Implementing the closed-loop improvement of the entire manufacturing process through analytics, to increase the yield rate.

Teradyne’s FLEX Test solutions help you achieve this cost of quality with flexible solutions that allow you to shift tests between insertions. Our strong industry partnerships across EDA, fab/packaging and data analysis ensure this process is seamless.

Data analytics drives improved decision-making

In the face of the choice to shift left or right, optimizing the test strategy is a dynamic and continuous process, where analytics can play a key role in informing these decisions. Teradyne’s Archimedes analytics solution can help provide valuable data for adjusting test strategies throughout the chip manufacturing process.

Teradyne Archimedes analytics solution integrates technologies like data analytics, artificial intelligence and machine learning into your test solution, enabling a secure stream of real-time data with near-zero test time impact that can increase yield, improve quality, and reduce tester down time.

It’s an open development environment that supports both out of the box and custom solutions to ensure insightful learnings can be realized for 2.5D/3D packaged devices. Our tight integration with best-in-class analytics providers ensures you can choose the solution that enables you to achieve your goals for advanced devices.

Bridging the gap from design to test

With chiplets, reducing defect escapes is not the only concern, yield must also be taken into consideration. To achieve yield improvements, bridging the gap from design to test is key to improving engineering efficiency. New workflows require design, manufacturing and test engineering teams to work together seamlessly to accelerate device development and generate learnings. Not only do EDA and JTAG tools need to be enabled on ATE and SLT test systems, but it’s also useful to have a common set of libraries and debug tools that allow design and DFT engineers to seamlessly collaborate, sharing key insights, accelerating silicon development, and reducing learning curves.

Teradyne’s PortBridge is a common toolset that bridges the gap between design and test and can be deployed at any stage of the manufacturing process to identify, implement and verify opportunities for improving yield, including:

• Debugging faults in system level test
• Understanding faults in the final test insertion
• Enhancing test coverage during wafer inspection to reduce defect escapes
• Revealing inefficiencies in the production flow to improve device quality, reduce defects and increase yield

PortBridge works with Teradyne’s UltraFLEXplus and UltraFLEX testers and provides:

• Protocol Libraries for common protocols devices used today, and ones needed in the future. Usable all the way to production.
• Remote Connect, built-in support to connect EDA tools and custom bench environments to the ATE remotely. Gets the right people looking at the problem, using the tools and environment they are familiar with.
• Design File Support, which enables the use of standard design formats, like SVF, or custom formats, to eliminate conversion steps that waste time and lose valuable information.
• Host Debug Tools that deliver out-of-the-box, protocol-specific tools to expose the exact details needed while developing and debugging a test program.
• Production Enablement, the same protocol libraries can be used from debug to production to help with correlation, reduce overall effort and provide failure analysis with optimal test time.

With PortBridge, debug time is reduced from months to days with a platform- and software-optimized solution.

While 2.5D/3D packaging technologies offer a path to continue the next generation of Moore’s law, rapid identification of defects and quick implementation of optimizations is the key to cost-effective, high-volume manufacturing. A well-understood test process, where tests can be shifted left or right to reduce the defect escape rate, hence lowering scrap costs during the manufacturing process, is one strategy for minimizing the overall manufacturing cost of these components. This flexibility, coupled with the integration of capabilities in the design and test engineering domains, will facilitate the rapid identification, debugging and elimination of faults while achieving the best cost of quality. And while every stakeholder must do their part to increase efficiency, collaboration across all of the key stakeholders and really the entire industry is key to the success of achieving maximum operational efficiency.