The foundry model: Round two is engineering
November 17, 2020
November 17, 2020
It’s been more than 30 years since the foundry model revolutionized the semiconductor industry by restructuring chip manufacturing. Now the industry’s ready for round two, this time they are expanding their success in other engineering functions.
Semiconductor companies have three distinct engineering capabilities: core product design and engineering, design and technology enablement (or infrastructure engineering), and operational engineering (process and manufacturing engineers for yield, test, product and quality). Companies have long been structured to use these capabilities to deliver very specific semiconductor functionality for a small number of customers. But the market is changing.
A growing volume of customers have diverse product and program needs as they push to create unique end products. This trend will only gain steam as 5G becomes more pervasive, enabling companies to capitalize on applications that weren’t possible with previous generations of networks. That means semiconductor companies must spread their product design and engineering capabilities across many different industries, customers, and products. Meanwhile, differentiation is tough, since the technology continues to commoditize as the industry matures. And, there’s also the ongoing challenge of trying to compete for semiconductor engineering talent against the big, enticing platform companies.
Bring this all together and suddenly semiconductor companies find themselves trying to serve more customers and more unique needs with an internal infrastructure built to design products to service a specific market while trying to differentiate themselves and deal with
customers that have differing breadth and depth, all while struggling to recruit talent.
<<< Start >>>
<<< End >>>
Product design and engineering is a semiconductor company’s core competency and primary differentiator. And it’s now time to establish an organizational structure that allows these businesses to do what they do best. That’s what the foundry model was all about. As the first dedicated foundry, TSMC started the movement to outsource chip manufacturing, which freed semiconductor companies to focus on design and engineering infrastructure. Now market dynamics are adding further pressure on that engineering infrastructure. Chip companies need to move up the engineering stack to create a similar foundry model for infrastructure and operational engineering, enabling them to get much closer to customers.
As markets evolve and diversify, it becomes more important to delineate infrastructure from core competencies. Pulling in outside organizations to handle infrastructure drives scale and creates the ability for infrastructure roles to support and grow in a flexible way across different design cycles, customer requirements, and market opportunities. By partnering with another organization, companies can attack new businesses and industries with new chip designs, and gain perspective on whether new ventures are successful without having to ramp up the in-house capabilities they will be stuck with if it wasn’t the right play. Conversely, if it was the right move, they can scale very quickly. The organization retains its key talent and has the flexibility to adapt infrastructure with a rapidly changing market. It’s a scalable, agile approach that de-risks the business and saves cost. Most of all, it enables chip companies to focus on their core product design and engineering—the bread and butter of the business.
This trend is already evolving in the operational engineering space. Tessolve, a fast-growing semiconductor engineering solutions company, now provides yield analytics for many semiconductor companies. Tessolve is consolidating engineering horsepower into a scalable and flexible organization that can provide, in this case, yield analytics for multiple businesses. There’s no reason why this model won’t work for other capabilities in the engineering stack and allow semiconductor companies to divest themselves of inflexible and costly infrastructure and operational capabilities.
<<< Start >>>
It’s now time to establish an organizational structure that allows these businesses to do what they do best. That’s what the foundry model was all about.
<<< End >>>
If your revenue makeup is shifting, indicating your customer base is expanding or becoming more complex, that’s a sign to look into what parts of infrastructure and operations engineering that other organizations can participate in. This exploration needs to, first and foremost, be supported by top executives, as it's a pretty big change in how the business is organized. It’s also an approach that will only work if it’s aligned with the organization’s strategy and goals.
The best starting point is to assess which parts of your engineering capabilities are fungible, meaning those that require industry skills but not product-specific skills – skills such as data analytics or silicon design verification. If you can separate fungible skills from product-specific skills, those fungible skills are candidates for co-engineering partnership. They may still be hard functions, but product-agnostic enough to be able to trust a third party to execute. If they’re repeatable skills, they may also be candidates for automation and AI by either your company or your third-party partner. Of course, it’s important to find the right partner – one that can demonstrate how they’re building their “engineering foundry” and the capabilities that will enable them to scale up and down at the speed the organization may need.
As the market for semiconductors becomes more complex and diverse, companies that can effectively contract for infrastructure and operational engineering will free themselves to focus on their customers and product designs, meeting a much broader range of needs and applications. And, with partnered capabilities behind them, they will have the flexibility, scale, and cost base to move with the market—continuing to design the chips to power a growing range of innovative high-tech products.
The foundry model revolutionized how chips are manufactured. Let’s revolutionize chip design.
Disclaimer: This document is intended for general informational purposes only and does not take into account the reader’s specific circumstances and may not reflect the most current developments. Accenture disclaims, to the fullest extent permitted by applicable law, any and all liability for the accuracy and completeness of the information in this presentation and for any acts or omissions made based on such information. Accenture does not provide legal, regulatory, audit, or tax advice. Readers are responsible for obtaining such advice from their own legal counsel or other licensed professionals. This document may contain descriptive references to trademarks that may be owned by others. The use of such trademarks herein is not an assertion of ownership of such trademarks by Accenture and is not intended to represent or imply the existence of an association between Accenture and the lawful owners of such trademarks.