Industry awareness of ultra HDI is growing, and a significant share of the market is looking into designing with these ultra-fine features. Why do you feel that this new technology is important to the industry?
The technology option selected for interconnecting a collection of IC devices will always be based on the lowest total cost of ownership that provides enough performance capability. When you have a new technology that might have a total cost of ownership higher than conventional approaches, the applications to go after are those for which the higher total cost of ownership is worth it. It is not a question of what can be done with the new technology, but what can be done with a lower total cost of ownership or what can be enabled with the new technology that just could not be done before.
I think there are two important killer apps for ultra-fine line substrate technology. The first is what we call chiplets today. We used to call this multi-chip modules. Over the last 40 years, multi-chip modules have become very common. Many high-end microprocessor modules use either ceramic or organic substrates with multiple bare die packaged into a single module. Chiplets are the next generation with multiple, high ball count flip chip devices routed together on a single substrate with very high-density interconnects.
This is one approach to continuing the advance of Moore’s Law. When the largest size die is already being used and you want to mix technologies in a module that cannot be done on die, like fast processors, electro optic TX and RX and large memory, the chips can be integrated together with near IC density on a single substrate.
The purpose of the high-density substrate is to first provide the connectivity required. The second purpose is to not screw up the signal or power integrity. Once the high-density interconnects are available, the next issue is how to design them to reduce the problems created by switching noise, reflection noise and cross talk. There still remain design challenges.
The second killer app is for the interface between high pin count devices and the routing paths in conventional circuit boards. Samtec coined the term “the final inch”. This is the interconnect region that integrates the fine pitch, usually grid array footprint of the IC or connector and the courser features of the circuit board.
In either chip scale devices or high pin count devices, the footprint is usually the routing bottleneck. This is a region where ultra-fine line technology may play a role as the killer app to enable the smallest size substrate with the fewest number of layers while providing the fan out in the final inch to all the high density or high pin count devices that must be routed on a board.
Industry awareness of ultra HDI is growing, and a significant share of the market is looking into designing with these ultra-fine features. Why do you feel that this new technology is important to the industry?
Ultimately, the value of a new technology is related to the problems that it solves. If it does not solve a problem or provide a lower total cost of ownership solution than alternatives, it may be a very cool technology but there is no business case for it.
What key information points are critical to shortening the cycle for adoption – to more quickly begin taking advantage of the benefits of ultra HDI?
When a new interconnect technology is introduced, you need to make sure commonly used design tools like Allegro, Mentor’s Pads, or Altium can design with the new design rules. This enables connectivity. Then you need to have established the design rules for acceptable signal and power integrity. Finally, you have to establish the DFM rules for fabrication, at high yield, assembly integration, and reliability. Whatever you end up with should also be compatible with production test equipment. This will make a new technology integrate seamlessly into the result of the manufacturing infra structure.
As a Design expert that has navigated through multiple waves of technology advancements,
what advice do you have for those who are looking at learning new techniques?
Find a killer app. Develop some example prototype vehicles. Develop some reliability test vehicles and develop a set of design rules that any customer can follow. As examples proliferate in the industry, other applications will appear that you had no idea about initially.
What are your outside interests – the hobbies that help you to reboot?
I write textbooks, teach undergraduates and graduate students how to become valuable engineers and also write science fiction.
