Last month’s column identified the significant challenges today’s electronic systems engineering teams face, and how a digital transformation of the entire design and manufacturing flow promises to resolve them by confronting product, process, and organizational complexities. One of the areas where we see opportunities for improvement is in the reuse of critical IP as part of a team’s data management strategy: whereas the data is already digital, it’s not always leveraged efficiently from one design to the next.
IP reuse isn’t new; it’s been used extensively in software and IC design. It’s estimated that 68% of an IC design is built on data reused from in-house or third-party IP. In both cases, reuse was instituted to support concurrency, reduce cost, improve quality and reliability, and help manage complex IC designs and software projects. The complexity of electronic systems is driving PCB design toward this same hierarchical “building block” approach.
Many new products are variations of existing products, so it only makes sense to reuse data from those previous products whenever possible. Leveraging “known good” verified circuits in new PCB designs accelerates new product development and eliminates repeated creation of the same circuitry. For example, once a power supply circuit is built and verified, why recreate it for every design that has the same power supply requirement?
Reuse also facilitates component consolidation and ensures continuous circuit improvement as qualified data from the consumers of the reuse blocks are incorporated into the source circuit blocks. Finally, for designs containing sensitive or restricted components or circuitry, formal management of the source circuit blocks allows for proper tracking of critical information, such as the IP owner, ITAR classification, and any export control requirements.
In PCB systems, reusable IP comes in three primary forms:
- Component libraries: Most are familiar with a central library of part models—in the form of symbols, 2D/3D footprints, simulation models, etc.—that get reused across multiple designs. The current focus here is on team synchronization (ensuring everyone is always using the same library data) and library completeness (measured by the number of parts and the inclusion of newer elements like 3D representations and multi-physics simulation models).
- Boards: Teams often reuse internal or COTS boards across many multi-board systems. Incorporating these boards into new projects is critical so that the complete system can be modeled without data redundancy.
- Circuit blocks: These fit between #1 and #2 in design hierarchy and consist of multiple components and/or the interconnects between them. The data could be logical (schematic-based), physical (layout-based), or a combination of the two. They could be developed internally (e.g., power supply, antenna, or standards-based interfaces between processor and memory), or leveraged from external sources (e.g., reference designs from IC vendors).
There are a lot of lessons to be learned from reuse in IC and software design that can now apply to PCB design. These include:
- Reuse requires a systematic, new process to verify known-good data, store the design data and associated verification results where they can be found, and then ensure that searching for existing IP is part of the process for new designs. It’s a classic case where a little added process complexity resolves much more significant challenges, but at the time it’s just another hurdle to overcome.
- Once IP is being reused, it becomes critical to track it in case changes are required. The impact of changes must be assessed, and the appropriate teams notified if it’s executed.
- Certainly, design software tools are critical to simplifying an IP reuse process, but a new methodology requires a commitment from management to ensure compliance with the new process and accept some short-term delays as it’s implemented.
- There is a pervasive reluctance to use circuits designed by others because the quality is unknown, it feels like plagiarism, or is due to the classic “not-invented-here” syndrome. This reluctance needs to be overcome, first by a commitment from management and then through the benefits that teams will see from practicing reuse.
If deployed across an engineering organization, IP reuse can deliver significant benefits:
- Getting to market first by jump-starting new designs through the reuse of existing IP, thereby accelerating the product development process. Collaboration and efficiency are increased through a “building block” design process that fosters concurrent development, enables knowledge-sharing, and helps manage the increasing complexity of today’s PCB systems.
- Reducing product development cost by eliminating repeated creation of the same circuitry, consolidating components to increase purchasing power, and reducing the number of prototypes due to inherent lessons learned.
- Improving design quality by leveraging past successes—which also reduces risk in new products, reducing verification time by “certifying” known good circuits, and ensuring continuous circuit improvement.
- Ensuring compliance with corporate and industry requirements by leveraging certified IP, managing access rights for IP security, and reducing an organization’s overall risk.
Effective reuse is ultimately a combination of corporate culture, efficient methodology, and enabling EDA technologies. I would encourage engineering teams to evaluate their current processes for effective IP reuse; a little process improvement pain will lead to extensive time, cost, and quality gains.
By digitally transforming the reuse of design IP, electronic systems engineering teams will reap the benefits and rewards of being the first to deliver tomorrow’s designs to market today.
This column originally appeared in the November 2021 edition of Design007 Magazine.