Intellectual Property (IP) Management in Electronic Design
BY HENRY POTTS, Mentor Graphics
Typically, when we hear the phrase intellectual property, we immediately think about theft and protection. We consider IP to be the most valuable resource of an electronics company and we want to protect it from compromise. What we often forget is that IP is of little value if it cannot be used efficiently: created, accessed by designers, and leveraged to improve a company’s competitiveness.
Management of IP is especially complex in today’s global environment where an electronics company may have several locations worldwide that they must leverage to bring a product to market. For example, a system-in-package (SiP) for an advanced product may require design of a custom IC, design of the SiP itself, and the design of the PCB on which the SiP is mounted. For the highest performing, most competitive product, it would be ideal for these three elements be designed in parallel versus serially, thus adding even more complexity to the process and more need for efficient IP management.
A company’s IP at the very highest level is knowledge. That knowledge can be broken down into three elements: data, intent, and know how. Missing any one of these elements leaves a hole in a company’s ability to produce their next product. These elements must be easily created, protected, and made available to the extended members of the design team.
The design data of the ICs, the SiP, and the PCB themselves involves designing the IP elements, and this database expands as the design proceeds. This work-in-progress design data must be controlled and accessed by many members of the design team. For example, while layout designers are completing placement and routing, an engineer should be verifying that routes meet performance specifications of the product. That engineer must be able to access the design data but not have permissions to change the routing.
This scenario becomes even more complex if the electronics company wants to implement collaborative design of the IC, SiP, and PCB. Here, the IP needs not only to be shared, but at times open for negotiation. For example, the input/output (I/O) assignment on the SiP is variable and a good assignment can have a significant positive impact on the PCB routing and the end product’s performance. So, during the SiP design, collaboration (and access to a good design tool) between the SiP designer and the PCB designer and sharing their IP is necessary.
Component libraries are another data element of high-value IP. Companies invest highly in qualifying parts and then entering them into libraries accessible on a global basis. In the case of an SiP design, these parts may be packaged discretes or even bare die. Having a central component library system that enables productive input of part data, control of that data, and access to the data on a global basis by various design teams can significantly improve both the time-to-market and the quality of the product.
The creation and access of design constraints, also called design intent, is another IP element. Intent comes in several forms. Again, some (e.g. manufacturing rules and best practices) are created prior to the start of a design, some (e.g. high speed rules) as the design progresses. Then we are faced with the challenge of having multiple disciplines responsible for creation and use of the IP. Manufacturing engineers create both the hard manufacturing rules and the design-for-high-yield manufacturer-specific best practices. Engineers pre-analyze the design and create delay, SI, and EMI rules. These rules are then made accessible to the layout designers who must follow them. As you can imagine, this requires a well-integrated systems design flow from the design automation vendor. These constraints have to be defined and used in languages familiar to the multiple disciplines involved. For example an engineer may think of delays and edge rates in terms of nano- and pico-seconds while a designer must implement them in terms of inches of routing.
Many of today’s designs are, in fact, upgrades or re-spins of previous designs. A company can improve their design cycle times and product quality by re-using part or all of previous products. So when a design is complete, saving the design (data), the constraints (intent), etc. in a well organized IP management system can give the next product a jump start. But keep in mind that the design team for the second product may be completely different; different people, different locations, different skill level, and so on. So there must be a transfer of know-how along with the basic re-useable information.
Let’s return to the original thought of IP protection. In today’s global environment, we are often utilizing third parties for manufacturing and even parts of our design. In this situation, it is important that our IP not be compromised so having an IP management system that only shares the subset of IP necessary for the third party to perform efficiently is extremely important.
Henry Potts, VP and general manager, systems design division, Mentor Graphics Corp. For more information go to www.mentor.com/pcb/expedition.