(9 pages)
The semiconductor industry is booming, with expected average annual growth of 6 to 8 percent through 2030 and yearly revenues forecasted to reach $1 trillion. The industry will have to double semiconductor production to keep pace with future demand, but most fabrication plants, commonly called fabs, are already operating at capacity. To increase supply, many companies have announced plans to build new fabs and some are already in the construction phase. And in a coordinated push to achieve a microelectronics resurgence, the United States is becoming a hot spot for fab construction.
About the authors
This article is a collaborative effort by Katy Bartlett, Ondrej Burkacky, Loraine Li, Rutger Vrijen, and Bill Wiseman, representing views from McKinsey’s Semiconductors Practice.
The value of US-based semiconductor projects that are under way, announced, or under consideration totals $223 billion to over $260 billion through 2030.1These figures are current as of January 2023. More announcements may be forthcoming. But it’s not just private companies that want to move more manufacturing to the United States. Semiconductor production is getting more funding from the federal government, which recently approved $54 billion in grants for domestic semiconductor manufacturing and research through the US CHIPS and Science Act.
The value of US-based semiconductor projects that are under way, announced, or under consideration total $223 billion to over $260 billion through 2030.
The current economic outlook is prompting several semiconductor companies to slow their capital deployment, however. Combined with the strong outlook for future demand, this creates a dilemma: companies may decide to invest in the through cycle to avoid another supply shortage down the road while simultaneously attempting to manage cash flow limitations. Building fabs in the United States may also present different challenges than those encountered in other countries. Some projects have already experienced delays, including those related to labor and material shortages. What’s more, volatile prices for raw-material commodities are injecting another element of uncertainty into the construction process. The potential solution for companies that want to continue building US fabs amid all this uncertainty? A mix of creative financing, more thoughtful design, greater prefabrication, and better strategies for negotiations, scheduling, and cost control.
The building boom for US fabs
For many years, chip manufacturing has been consolidated in Southeast Asia and China. Semiconductors manufactured in the United States now account for only about 12 percent of the global total, down from 37 percent 30 years ago, according to a recent White House statement (Exhibit 1).2“Fact Sheet: Biden-Harris administration bringing semiconductor manufacturing back to America,” White House briefing room statement, January 21, 2022. The amount varies by node size, and leading-edge chips—those with node sizes under ten nanometers—are largely produced in East Asia, although some fabs outside this region are now ramping up production (Exhibit 2).
1
2
When supply chains were functioning well, companies had little incentive to build new fabs outside Southeast Asia. But chip output and distribution have recently faced challenges because of the COVID-19 pandemic and consequent supply chain disruptions, with a 2021 drought in Taiwan and recent geopolitical issues compounding the problems. These considerations have prompted companies to take a new interestin diversifying their fab locations and exploring US sites. The availability of subsidies is one of the main considerations when evaluating potential new locations.
Where are US fabs being built?
A significant portion of the $223 billion to over $260 billion being invested in US fabs is concentrated in geographic clusters. The exhibit shows how US fab locations, both current and announced, are distributed.
Of the total $223 billion to over $260 billion investment in new US fabs, about $183 billion is attributed to projects that are ongoing or announced; the remainder relates to projects still under consideration (Exhibit 3). Most investment is flowing to specific geographic clusters. Arizona and Texas, for instance, are attracting investment because they already have fab ecosystems, and their local governments have historically provided incentives and helped to coordinate the process. Adding to the large incentives play: Ohio is emerging as a desirable site, with over $20 billion in investment announced for Columbus-based fabs, and New York is offering incentives to encourage fab construction. Other states attracting investment include Indiana, New Mexico, Oregon, Utah, and Virginia. (For more information on geographic clusters, see sidebar “Where are US fabs being built?”)
3
Key challenges for US fab construction
Semiconductor fabs are complex, capital-intensive projects in any location. But in the United States, where fab building has been uncommon and construction talent is in high demand across sectors, semiconductor companies may face even more challenges than usual. Many of the issues they face fall into one of the following categories.
A shortage of construction talent
It’s more than construction: The ongoing talent challenges for semiconductor companies
Once a semiconductor company nears completion of a new fab, it faces another talent challenge: recruiting, training, and retaining the engineers and skilled technicians needed to operate it. Research suggests that the United States will have about 300,000 fewer engineers than needed across all industries over the next eight to ten years; for skilled technicians, the gap will reach about 90,0000. Robust demand for workers with technology skills has already increased competition for talent, and semiconductor manufacturers have had less success than technology companies in acquiring the best employees. With engineers and other talent in short supply, fab operators may need to offer highly competitive compensation or consider recruiting from nontraditional talent sources (for example, other industries, residential construction, military veterans).
Large-scale fab construction has not occurred in the United States in more than 20 years, and few builders within the country possess the experience, capabilities, and expertise required to deliver these specialized projects. Compounding this issue, semiconductor players must compete with companies across multiple sectors, including residential housing, for construction workers of all types—from earthwork specialists to skilled electricians—in an already tight labor market. Once fabs are completed and open, they will also face competition for a very different type of talent: the technical employees required to operate them. (For more on this topic, see sidebar “It’s more than construction: The ongoing talent challenges for semiconductor companies.”)
A greater emphasis on sustainability
There is a greater emphasis on sustainability within the semiconductor industry because many of the most important customers want to reduce emissionsalong their supply chains. Many semiconductor companies have not yet clearly articulated their sustainability goals, however, with only about 60 out of approximately 2,000 companies committing to emissions targets.
Semiconductor manufacturing may attract particular scrutiny from end customers because it is responsible for a high level of emissions associated with end products. For example, more than 70percent of the lifetime emissions associated with some mobile phones are related to manufacturing the phone itself and the chipset. As end customers increasingly focus on achieving net-zero emissions, more semiconductor companies are expected to commit to more aspirational and actionable emissions targets.
Fab owners may also consider accessibility to renewable-energy resources more carefully as they decide where to build new main fabs, since 45percent of emissions of a typical fab are electricity related, excluding Scope 3 downstream emissions.3Scope 1 emissions are those from direct or controlled sources; Scope 2 emissions are from the generation of purchased electricity, steam, heating, and cooling equipment; Scope 3 upstream emissions include all other indirect emissions in a company’s value chain; and Scope 3 downstream emissions are related to the use of products containing semiconductors.
Supply chain complexity
The typical semiconductor production process could involve steps in more than five countries and three or more shipments across the globe. Regional bottlenecks exist at nearly every step in the value chain because of industry consolidation, labor cost dynamics, and technical complexity. To boost supply chain resiliency, semiconductor companies might consider building or moving pieces of the value chain closer to new fab sites. Strategies for material sourcing for construction and operations may require updating. The capital spend will extend beyond the fab build because creating a true ecosystem will necessitate additional infrastructure and suppliers.
Federal and local incentives
The CHIPS and Science Act will provide $54billion in federal incentives for semiconductor manufacturing and R&D, but companies will have to navigate a range of eligibility requirements in order to qualify for funding. Subsidies are critical to shortening the payback period for fab construction, and companies may favor sites that allow them to obtain state or local incentives. Negotiations with government stakeholders before and during fab construction may influence both funding and incentives; it is clear that companies that receive CHIPS program management funding will have to meet certain requirements.
Well before they begin fab construction, semiconductor companies will need to consider future negotiations with stakeholders at all levels of government. They could also improve site selection by looking at other factors that will affect costs and timelines. Other site selection criteria (for example, operations costs, capital efficiency, infrastructure, labor, carbon emissions, water, energy, land, societal) will need to be fully defined, analyzed, and weighed to ensure a holistic decision-making process.
Difficulties with performance management and execution
Delivering large or mega capital projects on time and on budget is difficult under normal circ*mstances. The current disruptive forces—commodity price volatility, inflation, supply chain disruptions, an overheated labor market—make projects even more complicated, and semiconductor companies may encounter obstacles even if they follow construction best practices across a project’s life cycle. For instance, too many recent projects have suffered from delivery delays for key long-lead mechanical, electrical, and plumbing systems, even when procurement orders were placed well in advance and confirmed by suppliers. What’s more, the overheated labor market may reduce productivity because companies cannot find enough qualified construction employees, making it even more difficult to deliver facilities on time and on budget.
An opportunity to transform fab construction
The increasing demand for chips could present a major opportunity for semiconductor companies, provided that they can build new fab plants efficiently. Challenging economic times may not be an insurmountable obstacle to well-functioning construction projects—companies that built fabs during the 2008–09 recession, for example, often remained on track and within budget.
As semiconductor companies contemplate major capital investments in US-based fabs, their leaders may want to consider current challenges and implement the following cutting-edge practices during construction.
Creative financing
Government incentives alone will not cover the enormous capital investment required to build leading-edge fabs, and semiconductor companies may want to explore alternative funding strategies beyond debt or small capital to fill the gap. For instance, one leading semiconductor company recently undertook a joint venture with a large asset manager that provided private capital. Such arrangements allow companies to conserve cash and preserve their debt capacity, and other semiconductor companies may find similar opportunities to form partnerships with private equity megafunds, including those focused on technology and infrastructure. The terms of these agreements state that alternate financing may represent equity investments by partners that entitle them to a profit share down the road. Fab players could also consider large private-asset managers or insurance companies with long-dated liabilities in their funding strategy.
Modular design and prefabrication
Prefabrication and modular (PFM) solutions can allow companies to complete certain elements of construction off-site in more controlled environments. Done well, these solutions can confer multiple benefits, including shorter project timelines and lower costs. For instance, companies may be able to undertake PFM building in locations where costs have traditionally been lower. Safety might improve because the need for on-site labor is lower, and quality could see a similar boost because manufacturing occurs in a more controlled environment. Sustainability—a growing concern—may increase because the amount of waste materials may decrease, and companies may be more likely to access renewable energy during off-site manufacturing.
Before embarking on projects, semiconductor companies must decide where they want to apply PFM. Some may advocate for modular, preassembled clean rooms, for example, while others may use skid units for facilities or central utility buildings. In other cases, they may simply apply PFM principles to architectural or civil structural architecture components by using elements such as precast concrete and prefabricated steel.
Although PFM solutions offer many benefits, return on investment can be slow, partly because the process, which is highly complex, can be difficult to get right. Companies must manage multiple small construction and manufacturing sites, more complicated logistics, and more quality-control issues, among other concerns. To minimize potential errors, fab operators could benefit from introducing PFM concepts early in the construction process and standardizing them across projects. If companies hesitate, they may find it more difficult to introduce PFM concepts at a later stage. On the organizational side, companies could benefit from new ways to manage PFM projects and increase transparency about the process. Groups that may need to be involved include contracting, logistics, testing and quality, supply chain, and maintenance. For instance, certain assets that need upkeep may be stored or kept at an off-site location, and companies could specify both the process for maintaining them and the responsible parties.
Strategic negotiations, contracting, and procurement
Given the current labor shortages, supply chain disruptions, and economic uncertainty, skilled procurement and contracting, including negotiation skills, are more important than ever. Some engineering, procurement, and construction firms and general contractors have already adapted their contracting strategies to reduce risk. For instance, some have added additional contingencies or allowances or included contract language that allows them to increase prices if certain commodities used in construction become more expensive.
Some fab operators are also becoming more involved in procurement, either by moving more elements into the owner-furnished, contractor-installed category, or by adopting other purchasing models (for instance, engaging in joint negotiations and then having the contractor complete the purchase). These shifts allow fab operators to reduce contractor overhead, ensure on-time delivery, and bolster supplier relationships.
Fab operators can further improve their current capabilities and sourcing strategies to include real-time market intelligence analytics, cleansheets, and robust procurement tracking and tendering. They could also engage in value engineering or pursue alternate solutions that improve their ability to anticipate and control material and labor costs. Together, these new skills may help them set better prices and realize margin that would otherwise go to contractors.
Generative scheduling
Semiconductor companies might benefit from implementing a generative-scheduling approach, which involves building a data-driven model that considers the physical and spatial constraints that govern how work needs to be done on-site. Once the model is created, companies can use “what if” questions about labor, equipment, installation rates, access, start-up sequence, productivity, and other metrics. These questions allow them to compare different schedule and resource combinations in real time. The output is an optimized construction sequence and resourcing plan for all phases of a fab project, from preconstruction through execution, that minimizes cost and schedule overruns. Companies can do new generative-scheduling analyses at any time, especially if circ*mstances change.
Project control towers
Within complex megaprojects, transparency about progress is essential but often absent, making it difficult to recognize and resolve problems proactively. A centralized project delivery hub—basically a digital platform that serves as a control tower, accessible to all relevant staff—can help management gain greater visibility into potential obstacles in all project areas. On a recent fab project, establishing key project management solutions, particularly a control tower to monitor project performance and risk areas, increased transparency significantly and accelerated progress, allowing the company to reduce the project timeline by eight weeks.
Semiconductor companies could be opening a new era in which more chip manufacturing moves to the United States. But this major shift may also require big changes in the process for planning and constructing fabs, especially since the strategies and processes that work well elsewhere may not work in the United States. The stakes are high, given the significant investment required to build a fab, but the potential gains are large as well. Considering new strategies during the planning phase, before the ground is broken, may give fab operators the greatest chance of completing fabs on budget and on time.
Katy Bartlett is an associate partner in McKinsey’s Denver office, Ondrej Burkacky is a senior partner in the Munich office, Loraine Li is a consultant in the Boston office, Rutger Vrijen is a partner in the Bay Area office, and Bill Wiseman is a senior partner in the Seattle office.
The authors wish to thank Jim Banaszak, Justin Dahl, Sebastian Göke, Rawad Hasrouni, Christine Herrmann, Henry Marcil, Praveen Matta, George Naughton, Mark Pitcher, and Matt Schrimper for their contributions to this article.
Explore a career with us
Search Openings
