Summary

OVERVIEW

Microelectronics are a vital part of daily life today, either embedded within the digital devices that people depend on or used to produce, measure, or monitor the non-digital world. Since the first semiconductor integrated circuits were produced nearly simultaneously by Texas Instruments and Fairchild Semiconductor in 1958, firms in the United States have played leadership roles in the global semiconductor industry. Advances in semiconductor technologies have conferred upon the United States incalculable economic and national security benefits. The Department of Defense (DoD) played a critical role in creating the semiconductor industry almost from the very start, funding key innovations and (together with the National Aeronautics and Space Administration [NASA]) acting as a major customer in the early days of the industry. DoD systems have enjoyed technical superiority for decades due in part to the steady cadence of advances in U.S. microelectronics being incorporated into its communications, weapons, and deterrence systems.

Today, however, geopolitical and techno-economic factors have resulted in a hollowing-out of the U.S. semiconductor manufacturing sector, and U.S. semiconductor technology leadership is in question. Fewer and fewer firms can afford the capital expenditures necessary to build and operate the most advanced semiconductor chip fabrication facilities, which now cost tens of billions of dollars, resulting in a concentration of manufacturing capability among a small number of companies. In response to rising costs, many U.S. semiconductor firms have moved their manufacturing and assembly, testing, and packaging facilities overseas,

seeking lower costs, reduced operating expenses, robust supply chain ecosystems, and, often, generous support from host nations. Today, the domestic U.S. share of global semiconductor manufacturing value by sales is just 12 percent, down from 37 percent in 1990. World-class contract manufacturers (so-called “foundries”) have emerged in Taiwan and South Korea—home today to the leading firms capable of high-volume, leading-edge, logic, and memory chip manufacturing. While U.S. firms remain leaders in important sub-markets like electronic design automation software tools, integrated circuit design, and semiconductor manufacturing equipment, the global semiconductor supply chain has shifted decisively toward countries in the Asia-Pacific region, partly because of supportive government policies and targeted investments in those nations. The bottom line is that U.S. technology leadership in commercial microelectronics is no longer assured, and much of the vast, global supply chain is outside of its span of control.

Against that backdrop, securing access to the very best microelectronics has become a defining challenge for today’s DoD. In 2022, DoD released “Microelectronics Vision,”¹ articulating a goal that “DoD will obtain and sustain guaranteed, long-term access to measurably secure microelectronics that enable overmatch, increased operational availability, and support warfighter combat readiness” (p. 1). DoD needs a wide variety of microelectronics to support its missions, ranging from legacy chips required to sustain existing systems, to custom chips with the highest security measures, to cutting-edge chips manufactured with older technology, to state-of-the-art (SOTA) commercial-off-the-shelf (COTS) chips to be embedded in other systems and for the most advanced high-performance computing (HPC) and data centers. Because semiconductor design and SOTA fabrication is extremely expensive today, neither DoD nor its systems suppliers can maintain dedicated SOTA facilities to support DoD’s low-volume, high-mix needs; therefore, it is critically dependent on finding ways to access the commercially driven SOTA manufacturing base. DoD’s Trusted Foundry (a Program of Record) and Microelectronics Quantifiable Assurance (MQA) programs are two approaches that support a vision for engagement with commercial suppliers, but a comprehensive approach is lacking, and hurdles remain.²

In decades past, DoD demand dictated the cadence of innovation in the semiconductor industry, and DoD was seen as an attractive customer. Today, leading semiconductor firms do not seek (and sometimes actively avoid) doing business

___________________

¹ Department of Defense (DoD), 2022, “Microelectronics Vision,” Defense Microelectronics Cross-Functional Team, May, https://media.defense.gov/2022/Jun/15/2003018021/-1/-1/0/department-of-defense-microelectronics-vision.pdf.

² MITRE Corporation, 2023, “Microelectronics Quantifiable Assurance (MQA) Independent Assessment,” https://www.cto.mil/wp-content/uploads/2023/07/MQA-Assessment-Briefing-for-Release-Distro-A.pdf.

with DoD, focusing instead on lucrative consumer markets. DoD’s long timelines for systems development, complex procurement processes, stringent security requirements, export restrictions, silos of expertise, and small volumes make it an unappealing customer for commercial semiconductor firms, which act quickly and decisively to fund and field each generation of technology, engaging in worldwide production for the global consumer market. That is, over a period of decades, there has been a marked shift in the relationship; today, the semiconductor industry no longer seeks support from DoD, but DoD needs the semiconductor industry. DoD will not be first mover at new technology nodes owing to the market complexities and economics of scale. However, it cannot be years (or decades) behind the technology forefront either, which is the direction that DoD concerns over supply chain security have led us toward. DoD must stay close to the early-stage development of new technologies, helping to shape directions and preparing to quickly incorporate new advances into their systems.

The U.S. government, led by the Department of Commerce (DOC), is currently stewarding a once-in-a-generation effort to promote and protect U.S. semiconductor industry leadership, following passage of the CHIPS and Science Act of 2022 (hereafter “the CHIPS Act”). The CHIPS office is in the process of distributing $39 billion of incentives to increase domestic semiconductor manufacturing in the coming months and years. In parallel, in October 2022 and October 2023, DOC imposed broad and novel export controls to restrict China’s ability to both purchase and manufacture certain high-end chips critical for military advantage³—the latest in ongoing technology protection actions. These efforts to promote and protect the semiconductor industry are reminiscent of U.S. government actions in the 1980s and 1990s, which targeted then-ascendent Japanese semiconductor firms with tariffs and voluntary export restraints and established a variety of public–private partnerships (PPPs), notably SEMATECH and the Semiconductor Research Corporation (SRC), to help sustain U.S. semiconductor industry leadership—efforts that were largely deemed successful. Policymakers today are hoping that history repeats itself. The CHIPS Act calls on all government agencies to partner with industry to protect and promote the U.S. semiconductor industry to ensure economic and national security.

This study was conducted at the request of Congress to address the challenges that DoD is experiencing as it engages with the global microelectronics sector, and to help it better understand how to engage with PPPs to support assured production

___________________

³ Bureau of Industry and Security, 2023, “Commerce Strengthens Restrictions on Advanced Computing Semiconductors, Semiconductor Manufacturing Equipment, and Supercomputing Items to Countries of Concern,” Press Release, October 17, https://www.bis.doc.gov/index.php/documents/about-bis/newsroom/press-releases/3355-2023-10-17-bis-press-release-acs-and-sme-rules-final-js/file.

and innovation in the semiconductor industry. Specifically, the study addresses four primary questions:

1. What is the competitive position of the United States in the global semiconductor ecosystem? And what are the barriers to sustainable and resilient production of semiconductors in the United States?
2. How can PPPs strengthen semiconductor manufacturing?
3. When partnering with the private sector for semiconductor production, what unique challenges and opportunities exist for DoD to support sustainability and resilience in the U.S. semiconductor ecosystem?
4. What policies for PPPs could be adopted to accelerate the development and adoption of disruptive technologies in the United States that benefit DoD and dual-use needs?

FINDINGS AND RECOMMENDATIONS

The vibrant national discussions regarding CHIPS Act funding allocations and the latter stages of this study occurred largely in parallel. The committee approached the study and the resulting report with a goal of incorporating the concurrent discussions happening in the relevant various communities of practice—among defense policymakers, industry, and analysts, as well as experts in science and technology policy, workforce development, and international policy. The Committee on Global Microelectronics: Models for the Department of Defense in Semiconductor Public–Private Partnerships found that DoD’s microelectronics challenges are significant and simply cannot be addressed in isolation. There is a significant risk that the most advanced chips in the world will continue to be produced offshore, as well as by foreign-owned companies, for the foreseeable future—chips that DoD may be unable to do without if it wishes to sustain a competitive stance. As a result, the committee urges DoD to (1) pursue a strategy of sustained and comprehensive engagement with DOC and other agencies to ensure alignment with the CHIPS Act implementation across the government in an effort to reshore and friend-shore these leading-edge capabilities and (2) simultaneously find new ways to engage with commercial semiconductor firms without expecting them to conform with DoD’s complex bureaucracy and procurement rules. DoD can benefit greatly from learning to act as a fast follower (Box S-1), nimbly leveraging the commercial success of semiconductor firms in the United States and allied nations, as it seeks to achieve its stated microelectronics vision.

The committee notes that DoD’s current efforts to engage with commercial suppliers, which include the Trusted Foundry program and demonstration projects such as Microelectronics Quantifiable Assurance (MQA), Rapid Assured Microelectronics

Prototypes (RAMP), Rapid Assured Microelectronics Prototypes-Commercial (RAMP-C), State-of-the-Art Heterogeneous Integrated Packaging (SHIP), and SHIP 2.0, are steps in this direction.⁴ DoD will need to accept, and mitigate, more risk than it has historically tolerated in these and other new arrangements as it builds the needed substantive partnerships with the commercial sector to secure timely access to the microelectronics technologies it needs to support its missions.

In all its activities, DoD is advised to emphasize long-term (i.e., well beyond 5 years) strategic coordination, investment in emerging technologies, leveraging of commercial advancements, and a modernization strategy that is nimble enough to incorporate emerging technologies and be responsive to global competition. The committee has recommendations related to each of these priorities and areas of emphasis. These recommendations, which emerge throughout this report, are summarized here according to a thematic grouping.

Develop a Comprehensive Department of Defense Microelectronics Strategy

DoD currently lacks an overarching microelectronics strategy, meaning different service branches and program offices are seemingly pursuing uncoordinated efforts related to microelectronics research, development, procurement, sustainment, and modernization. The committee recommends that DoD develop such a strategy with buy-in from relevant stakeholders. The committee notes that a

___________________

⁴ The DoD Trusted Foundry, Microelectronics Quantifiable Assurance (MQA), Rapid Assured Microelectronics Prototypes (RAMP), and State-of-the-Art Heterogeneous Integrated Packaging (SHIP) initiatives are addressed in more detail in the report.

comprehensive strategy does not mean that there is a centralized approach to working with industry but rather that the strategy and its implementation are guided by several distinct pillars, with specific lines of effort under each pillar. (See Recommendation 5.16.)

Develop a 21st-Century Research Agenda for Semiconductors

The committee recommends that DoD prioritize investing in disruptive, sometimes called “leap-ahead,” semiconductor technologies⁵ as key drivers of future technological advancement. This commitment should encompass both funding of and active involvement in long-term semiconductor research and development, including prototyping and testing, to ensure DoD-specific needs are being addressed. In coordination with the CHIPS interagency agenda, DoD (led by the Office of the Under Secretary of Defense for Acquisition and Sustainment) is encouraged to establish a long-term research agenda focused on post-complementary metal-oxide-semiconductor (post-CMOS) materials and devices. In addition, the committee recommends that DoD accelerate its artificial intelligence and machine learning (AI/ML) implementation strategy as these technologies are driving disruptive applications across many operational fronts and are dependent on access to very large numbers of SOTA commercial chips, which today come from primarily foreign-owned, offshore manufacturing companies.⁶ (See Recommendations 5.2, 5.3, and 5.4.)

Develop a Chip Design Initiative

Across the microelectronics sector, the cost to design leading-edge chips has been rising swiftly as the complexity of the designs has increased, challenging the commercial sector and the defense sector alike. The committee recommends that DoD organize and lead a new initiative with industry that would seek to utilize AI/ML tools to substantially accelerate new application-specific integrated circuits design and development, and thus reduce chip design costs.

Furthermore, although DoD does not ordinarily design its own custom chips (relying instead on contractors to do this work), it is important that DoD sustains in-house expertise for modern design to enable it to tap into commercial supply

___________________

⁵ DoD, 2022, “DoD Microelectronics Commons: A National Network for Defense Microelectronics Innovation,” Defense Microelectronics Cross-Functional Team, https://www.cto.mil/wp-content/uploads/2022/11/DoD_Microelectronics_Commons.pdf.

⁶ DoD, 2023, “Data, Analytics and Artificial Intelligence Adoption Strategy,” https://media.defense.gov/2023/Nov/02/2003333300/-1/-1/1/DOD_DATA_ANALYTICS_AI_ADOPTION_STRATEGY.PDF.

chains for manufacturing, and partner effectively with experts in chip design, in order to sustain competitive advantage.

This might be accomplished in several ways. DoD could create a center of excellence in chip design and security that DoD program managers can access as needed. Alternatively, DoD might consider embedding design teams within commercial companies that are producing modern chip designs. By engaging with the skilled design teams in industry, DoD would better understand the design processes that support the most advanced technologies. (See Recommendations 5.5, 5.6, and 5.7.)

The Department of Defense Needs Assured Access to the Most Advanced Chips

Many new and emerging defense technologies, including those which leverage artificial intelligence, machine learning, and data analytics, require very large volumes of the most advanced commercial chips. Most of these chips are currently manufactured offshore and by non-U.S.-owned companies, a fact that prevents DoD use under current policies. To ensure access to these chips, which is vitally important, the committee recommends that DoD develop a long-range strategy, to be supported beyond the timeframe of the CHIPS Act. Within this strategy, the committee recommends that DoD coordinate closely with other government agencies and manufacturing companies and explore ways to link semiconductor capital financing efforts by the U.S. government with agreements and programs to ensure ongoing DoD access to leading-edge chips.

The committee further recommends that DoD collaborate with the National Institute of Standards and Technology (NIST) to ensure that university researchers and start-ups have ready access to advanced facilities for prototyping and scale-up of new chip technologies, which in turn may allow for U.S. chip manufacturers to disrupt the ecosystem and regain clear manufacturing leadership once more. (See Recommendations 5.2 and 5.13.)

Strengthen Private-Sector Engagement

The committee believes that the success of DoD’s microelectronics strategy will be largely dictated by DoD’s ability to constructively partner with the private sector, especially with leading-edge semiconductor companies, including “fabless” companies (without fabrication plants) that utilize foundries and integrated device manufacturers (IDMs) with their own fabricators. Currently, DoD’s microelectronics challenges are exacerbated by a series of self-imposed regulatory and administrative barriers, which make doing business with DoD undesirable in the eyes of many firms, and unnecessarily slow and difficult for those that do. The committee

recommends that DoD’s microelectronics strategy prioritize private-sector engagement, focusing specifically on the following:

• Reviewing and revising policies that limit DoD from manufacturing custom chips in commercial facilities, and simplifying the procurement processes;
• Developing policies to delegate more decision-making authority to program managers, given inputs from appropriate experts on security risks versus rewards;
• Pursuing export control reforms to reduce the International Traffic in Arms Regulations and Export Administration Regulation barriers to doing business with DoD;
• Continuing development of the Microelectronics Quantifiable Assurance program, which seeks to improve microelectronics sourcing flexibility; and
• Learning how to be a fast follower, to leverage the success and innovation of commercially competitive U.S. semiconductor firms.

(See Recommendations 5.8, 5.9, 5.10, 5.14, and 5.15.)

Engage Robustly with CHIPS Act Interagency Efforts

DoD microelectronics efforts should be aligned with broader interagency CHIPS Act efforts to the extent possible. For example, the committee recommends that DoD’s microelectronics workforce initiatives should be coordinated with the CHIPS Act workforce initiatives being stewarded by NIST and the National Science Foundation (NSF). The committee recommends that DoD coordinate with interagency efforts to advance immigration reforms that will increase access to talent, especially those actions recommended by the 2022 President’s Council of Advisors on Science and Technology (PCAST) report.⁷ DoD’s Microelectronics Commons⁸ and Next-Generation Microelectronics Manufacturing⁹ (NGMM) programs should be closely aligned with the CHIPS Research and Development Office’s work, especially that of the National Semiconductor Technology Center

___________________

⁷ President’s Council of Advisors on Science and Technology (PCAST), 2022, Report to the President on Revitalizing the Semiconductor Ecosystem, White House, September, https://www.whitehouse.gov/wp-content/uploads/2022/09/PCAST_Semiconductors-Report_Sep2022.pdf.

⁸ Microelectronic Commons, 2023, “The Microelectronics Commons: A National Network of Prototyping Innovation Hubs,” https://microelectronicscommons.org.

⁹ DARPA (Defense Advanced Research Projects Agency), 2023, “Next-Generation Microelectronics Manufacturing Aims to Sustain R&D Ecosystem: DARPA Names Eleven Teams to Kick Off Groundbreaking U.S. Chips-of-the-Future Effort,” https://www.darpa.mil/news-updates/2023-07-20.

(NSTC).¹⁰ Furthermore, the committee recommends that DoD collaborate closely with CHIPS for America¹¹ on incentive funding decisions and explore long-term sources of funding that extend beyond the 5-year timeframe stipulated in the CHIPS Act. Finally, DoD should coordinate with the Departments of Commerce, Energy, and State, as appropriate, to explore opportunities for increasing domestic and international semiconductor supply chain resilience. (See Recommendations 4.2, 4.3, 4.8, 5.1, 7.1, and 7.2.)

Address Microelectronics Obsolescence, Modernization, and Supply Chain Resilience

DoD is unique among nearly all other consumers of chips in that it continues to need microelectronics at technology nodes¹² that are decades old to sustain systems that have been in continuous service for that length of time. There is a particular need for approaches that address DoD’s unique microelectronics obsolescence challenges and modernization needs. The committee recommends that DoD sustain a concerted focus on supply chain resilience, to include modern design technologies, critical material vulnerabilities, radiation-hardened microelectronics, and advanced packaging. (See Recommendations 4.2, 4.5, 5.15, and 5.16.)

Strengthen Workforce Education Initiatives

The semiconductor sector in the United States will only be as strong as its workforce. DoD has the opportunity to help the nation address the gaps that have emerged in both the science and engineering professional workforce and the technical workforce for semiconductor manufacturing. The committee recommends that DoD work closely with NIST and NSF as they form and implement workforce education programs for the semiconductor sector under the CHIPS Act, as well as form and grow sustainable semiconductor education programs through its own efforts. In addition, DoD should support ongoing efforts to ensure federal government salaries for science, technology, engineering, and mathematics (STEM) employees are sufficient to attract the needed talent to incorporate current and future disruptive technologies, such as those based on AI, ML, quantum information

___________________

¹⁰ National Institute of Standards and Technology (NIST), “Establishing U.S. Leadership in Future Semiconductor Technologies,” National Semiconductor Technology Center, https://www.nist.gov/chips/research-development-programs/national-semiconductor-technology-center.

¹¹ NIST, “Chips for America,” https://www.nist.gov/chips.

¹² In semiconductor manufacturing, a “node” is a measure of the smallest feature size that it is possible to fabricate on the wafer, with the current lithographic tools and processes. Currently, the most advanced logic chip manufacturing is at node dimensions on the order of a few nanometers. DoD today requires some highly advanced chips that are fabricated at 130 nm node or even larger.

systems (QIS), and data science methods, into DoD practices, systems, and architectures.¹³ Given the prevalence of non-U.S. nationals in the semiconductor sector, DoD should press for reforms to the immigration system as recommended by PCAST in a 2022 report,¹⁴ including the award of lawful permanent resident status for all individuals with advanced STEM degrees working in semiconductor and other advanced technology areas. (See Recommendations 7.1 and 7.2.)

Adhere to Best Practices for Public–Private Partnerships

The committee reviewed a great many existing and prior PPPs related to semiconductor technologies in which DoD has participated, and generally found that they added value for DoD, especially in the early years of the PPP’s operation, presumably when the urgency of the problem was most acute. The committee notes that during the committee’s work, DoD has established eight new PPPs related to semiconductor supply, collectively referred to as the Microelectronics Commons, funded for 5 years under the CHIPS Act. Each new PPP requires a robust and sustained investment to attract the needed participants to engage as well as a leadership and administrative structure. Against that backdrop, the committee did not find any single challenge that could only be solved by creating a specific new PPP, rather than potentially leveraging an existing entity. If DoD nonetheless determines that one or more new PPPs is advisable to support its semiconductor needs, the committee has several recommendations about its structure and operations.

As for those within the Microelectronics Commons, the PPP should focus on supporting teams to move new technologies from demonstrations of prototypes in research laboratories to validation of components within systems and in relevant environments—the so-called “lab to fab” stages of technology development (i.e., technology readiness levels 4–6¹⁵)—and have substantive industry engagement supported by a mix of public and private funds for an extended period of time. This PPP needs well-articulated goals, metrics for success, and intellectual property (IP) treatment agreed upon by all members at the outset. Regular reviews of the PPP (e.g., quarterly or biannually) should be undertaken.

This PPP should be fully coordinated with the CHIPS for America office to avoid duplication and promote exchange of promising ideas, and this PPP should

___________________

¹³ G. Dille, 2023, “Pentagon Approves Pay Raise for STEM, Cyber Roles at IC Agencies,” Meri-Talk, August 23, https://www.meritalk.com/articles/pentagon-approves-pay-raise-for-stem-cyber-roles-at-ic-agencies.

¹⁴ PCAST, 2022, “Revitalizing the U.S. Semiconductor Ecosystem,” https://www.whitehouse.gov/wp-content/uploads/2022/09/PCAST_Semiconductors-Report_Sep2022.pdf.

¹⁵ DoD, 2023, Technology Readiness Assessment Guidebook, DoD Office of the Executive Director for Systems Engineering and Architecture, June, https://www.cto.mil/wp-content/uploads/2023/07/TRA-Guide-Jun2023.pdf.

explore partnerships with existing manufacturers to make use of these facilities and to develop technology transition plans at the outset.

Finally, any IP developed by this PPP should be carefully managed based on factors such as technology readiness level and the intended end use of any technologies created within the PPP. For any developments beyond early-stage research (where a consortium model allowing for sharing of information may be appropriate), DoD should recognize private ownership of IP created during the partnership and restrict government ownership narrowly to those extreme situations where chips are being developed solely for use by DoD and where the chips have no commercial application. With this approach, companies will be more willing to fully engage with the PPP. (See Recommendations 4.1, 5.1, and 5.11.)

CONCLUSION

There are a great many challenges facing DoD as it seeks to ensure a supply of microelectronics to both sustain its current capabilities and make certain it has technological advantage over any potential adversaries in the future. The committee is confident that many of these problems are solvable with a mix of targeted investments and changes in policy, process, and practice. However, the committee recognizes that DoD is a large and wide-ranging enterprise, and change is not trivial within such a structure. Among the committee’s recommendations, the single most important element is that of building close, trusted, working relationships with the commercial entities that lead innovation in the semiconductor industry to permit DoD to rapidly adopt state-of-the-art technologies, with an urgent focus on artificial intelligence and machine learning technologies which may have transformational impacts in national security spanning many technology domains.

To review the report’s structure, Chapters 2 and 3 provide historical background and the foundation for the recommendations that follow in Chapters 4 and 5. Chapter 6 provides a general overview of directions the committee suggests DoD follow. Chapter 7 reviews the workforce education issues in the semiconductor sector, both at professional and technician levels, and policy steps for consideration.

This page intentionally left blank.