Many industry experts believe the semiconductor industry is entering a new golden age. By many factors, it is certainly entering a growth stage.
First and foremost, semiconductors are a critical component of nearly all major technology trends, a fact illustrated by the recent breakout of sophisticated conversational AI tools like ChatGPT, which is fueling a new AI arms race, dependent on greater computing power and connectivity. Yet the increased chip demand, in combination with ongoing global supply chain issues originating from the COVID-19 pandemic, has resulted in a semiconductor chip shortage for the past two years. This shortage is expected continue through 2023, leveling off slowly as production capacity returns. Despite the eventual recovery of global supplies, the recent crisis has kick-started a resurgence of semiconductor manufacturing and design within the United States
U.S. leaders have long recognized that the reliance on other nations made semiconductors a national security issue because U.S. defense systems and platforms rely on semiconductors. The crisis has extended such security concerns to the broader economy. As the general public’s dependence on semiconductor-powered devices to make, sell, or operate goods and services in virtually every industry, chip shortages threaten general economic performance and stability.
The U.S. government’s new policy response is epitomized by the passage of the CHIPS and Science Act of 2022, which aims to accelerate investments, research and development (R&D), and commercialization of new and cutting-edge semiconductor technologies.
In the wake of the CHIPS Act and complementary industry-led initiatives emphasizing domestic production, the U.S. is poised to enter an extended era of growth in semiconductor manufacturing and design.
Trends and Expectations
The U.S. has become heavily dependent on semiconductors, and demand is increasing domestically. According to data from McKinsey & Company, three technology verticals in particular are driving the demand for semiconductors. Computing and data storage, wireless communication, and automotive electronics account for 70% of the overall semiconductor industry growth.
Despite their unrelated appearance, these verticals share a good deal in common. For example, many consumer electronic devices are connected to networks using wireless technology from wide-area 5G cellular communications to local-area Wi-Fi. Those devices connect to remote data centers where information is organized and processed, nearly in real time, made possible by high-performance chips in many different communications and storage devices throughout the electronic ecosystem. Ever-increasing data loads fuel demand for faster and more efficient semiconductors in a quickening cycle of technological innovation. Data networks that once struggled to service mere calls and texting, and now support social media and Netflix streaming, will soon support users engaged with AI avatars and extended reality environments.
Additionally, the emergence of edge computing has spurred the demand for fast and efficient computer systems distributed around the “edge” of large networks, as businesses and consumers employ ever more sophisticated apps and services in mobile devices. Accessing the cloud suffers a time lag due to data traveling back and forth from centralized sources. As semiconductors that are both powerful and energy efficient become increasingly available and affordable, more data processing is performed by local devices, cutting down on the lag.
Similarly, automotive vehicles have become more computerized and connected because the onboard systems are performing more data-intensive processing, especially with autonomous driving features. Modern vehicles are built with the ability to navigate highways, cruise on autopilot, self-park, and brake automatically in response to perceived external threats. As more vehicle features have become computerized – from the entertainment system to seat and comfort settings to engine controls and diagnostics – the modern vehicle has essentially transformed into a computer network on wheels. And, as if vehicle computerization wasn’t advancing fast enough, the rise of electric vehicles presents yet another driver of semiconductor demand with additional chip-laden sensors, complex battery management systems, and fully electronic engines.
The semiconductor sector permeates virtually every industry, but its demand is disproportionately driven by these three technological trends.
Global Challenges
Despite the U.S. accounting for 48% of global chip purchases, U.S. chip suppliers currently reflect only a 12% global market share of the semiconductor industry – compared to 37% in 1990, at the height of its former dominance.
Over the past several decades, the semiconductor supply chain became truly international, with elements spread across nations and continents, gaining significant presence in Taiwan, China, South Korea, and Japan. The global semiconductor industry sourced elements from all over the world, and different nations developed specialties in complementary processes related to chip production. The collaborative effort worked remarkably well for years to sustain a robust market, suffering only occasional shortages or inventory gluts as competition played out. The start of the COVID-19 pandemic, however, proved to be the beginning of the end of that international arrangement, as supply chain instability became painfully apparent in the wake of extended shutdowns. Governments were under increasing pressure to obtain scarce chip resources to benefit domestic businesses and citizens.
The U.S is no exception. By the time COVID-19 hit, many U.S. firms had become critically dependent on chips made internationally. According to the U.S. Department of Commerce, the resulting semiconductor shortage caused an estimated $240 billion reduction in U.S. GDP. The auto industry alone produced 7.7 million fewer cars in 2021 due to a lack of chips that were delayed from Asian producers impacted by factory shutdowns and shipping bottlenecks.
Over 75% of global chip manufacturing is concentrated around Asia, having long benefited from access to inexpensive labor and generous state-sponsored incentives. In contrast, a new fabrication plant in the U.S. costs 30-50% more to build and operate over 10 years than in several Asian countries, and the labor cost is several times higher due to a smaller skilled labor pool. Such disparities discouraged the expansion of what chip manufacturing remained in the U.S.
In the wake of the COVID-19-based supply shock, the decline in U.S. competencies in microelectronics raised legitimate concerns about national security and business competitiveness. Many business and government leaders called for securing the value chains of semiconductor design, manufacturing, packaging, and distribution as one of the foremost economic and domestic security issues of recent times.
Innovation Leads to Investment
The same McKinsey & Company report indicates international chip demand will continue to grow as the global economy recovers from the effects of the pandemic, leading to a $1 trillion semiconductor industry by the end of the decade. The opportunity for U.S. expansion is far better now than in recent years, particularly with supply chain problems being addressed by building locally.
In 2021, VC funding for semiconductor startups more than tripled year-over-year, according to PitchBook data. Despite venture funding falling in 2022 from the record highs of 2021, we’re not likely to see any notable cooldown. While the overall dollar amount of VC investments dipped in 2022, the number of deals was down much less by comparison – and still higher than in 2020, with nearly one-third of VC investments going to chip-related startups.
Further, 2022 saw the demand for semiconductors coexist alongside critical supply bottlenecks and chip shortages that continued since the earliest days of the pandemic, even with an increase in supplies from other sources.
A study from Deloitte found VC firms generally are not investing in building new chip fabricating plants or foundries. Rather, VC dollars will be directed to “fabless” semiconductor companies, which make no physical chips. Instead, they produce electronic designs for proposed chips and work with third-party foundries to manufacture the chips, test and process the designs, and eventually package them up once perfected. These fabless companies receive not only VC investments, but also funding from larger chip companies that view them as strategic investments.
CHIPS Act
For key reasons noted above, the U.S. administration has recognized the value in expanding U.S. semiconductor production and is bolstering efforts via the CHIPS Act.
The CHIPS and Science Act of 2022 will direct $280 million in spending over the next 10 years to boost domestic research and manufacture of semiconductors, predominantly to counter China’s dominance of the industry. The majority of the funds, roughly $200 million, is allocated to R&D and commercialization. In addition to government funding, the CHIPS Act establishes a tax credit of approximately $24 billion to spur private investment in the semiconductor industry.
The legislation also expands the technological ecosystem, along with jobs and related industry, to regions that previously had not been known for technology creation. As part of the CHIPS Act, $10 billion is being allocated to create 20 regional innovation hubs, designed to develop technology, create jobs, and encourage innovation. While exact locations are yet to be determined, at least three new hubs will be based in each of the Economic Development Administration’s six regions that span the continental U.S., with $1 billion going directly to “persistently distressed communities.”
Given the accelerated timeline, areas with existing infrastructure and skilled workforce will likely have an advantage, but we can expect to see the semiconductor reinvigorating local and regional economies as this takes effect.
These government incentives add needed fuel to the embers of current domestic semiconductor investment. As more facilities become available domestically to design and produce chips, the U.S. semiconductor industry will experience levels of growth not seen in several decades.
Collaboration Is Key
Consistent with the shift to fabless chip design, we are entering a period of specialization and collaboration among U.S. firms, focusing efforts on core competencies, often specialized applications, and licensing all other semiconductor design and manufacturing needs from others. We’re more likely to see firms working together to license and borrow specialized elements, instead of wasting resources attempting to recreate the wheel. This sharing of chip-related intellectual property (IP) will become a commonplace and necessary aspect of rebuilding the new U.S. semiconductor ecosystem.
Because U.S. innovation will rely on cooperation and sharing among firms, the system will become even more dependent on reliable patent protection. Relying on trade secrets, in contrast, will diminish in effectiveness where fewer companies possess the desire or capability to perform all or most of the chip design alone. Patents encourage the sharing of chip-related IP by helping to ensure each firm is compensated for its contribution. As further incentive, more companies will have patent assets to use as collateral when they seek funding. As a result, a key indicator of the success of the CHIPS Act will be a significant increase in the number of new U.S. patents directed to the design and use of semiconductors.
While we are likely to see companies of all sizes benefiting from a resurgence of U.S. semiconductor manufacturing, there will be a significant increase in growth-stage companies making and using chips in innovative ways across many industries. In this sense, a rising tide lifts all boats. The rise in semiconductor funding and collaboration will lift not only U.S. semiconductor firms but also the technology firms of leading industries in the U.S. whose growth is tied to chip access and development.