6


General Principles for the Department of Defense (DoD) to Follow for Accelerating the Adoption of Disruptive Technologies to Benefit DoD and Commercial Needs

Given previously described barriers and challenges and corresponding specific recommendations, this chapter will provide general and enduring principles for the Department of Defense (DoD) to drive change. This chapter supplements the report’s findings and recommendations with general principles to guide DoD efforts beyond the specific recommendations in Chapters 4 and 5.

PRINCIPLE 1: BE A FAST FOLLOWER

DoD should strive to be a fast follower in rapidly adopting and incorporating into systems new microelectronic technologies developed by commercial industry. Here the committee is considering mainstream logic, memory, and analog integrated circuit process technologies that are advanced by industry leaders such as Intel, GlobalFoundries, Micron, Texas Instruments, Analog Devices, Nvidia,

and others. DoD cannot hope to match the capabilities of those companies but should instead strive to adopt those technologies for DoD use very shortly after they are available for commercial purposes. In some cases, DoD may need to work with those companies to develop derivative versions of their technologies to better match DoD needs. Even for the development of emerging mainstream technologies such as artificial intelligence (AI) and quantum computing, there are many corporate laboratories that can be engaged by DoD. In general, then DoD should strive to rapidly adopt leading-edge technologies developed by industry and try to minimize developing custom technologies for DoD use. (See Recommendation 5.15.)

PRINCIPLE 2: SIMPLIFY PROCUREMENT

An impediment to being “fast” is DoD’s long and complicated procurement process. A simpler and easier procurement process would make more suppliers willing to work with DoD and would enable DoD to more quickly adopt the latest technology. To briefly summarize a complex area, DoD’s procurement process needs to be simplified, using a Kaizen process or other similar formalism, so that suppliers will not have to deal with a lot of bureaucracy and thus will be more willing to be a supplier to DoD. (See Recommendation 5.9.)

PRINCIPLE 3: ESTABLISH UPGRADE SCHEDULES FOR MICROELECTRONICS SYSTEMS

One way to overcome the impediment presented by long and complicated procurement processes is for DoD to mandate a periodic and regular upgrade schedule for some of its electronic hardware that is vulnerable to obsolescence because of the rate at which commercial microelectronics advances. There may be opportunities for DoD to learn these practices from cell phone manufacturers (e.g., Apple) that adhere to regular product upgrade cycles. As one example, signal processing electronics in radar systems could be upgraded more frequently so that these systems are only one or two generations behind commercial state-of-the-art (SOTA) microelectronics. Mandating upgrade schedules would provide DoD with an incentive to keep abreast of industry’s advances. Furthermore, the predictability may make industry more willing to work with DoD because of the anticipated opportunity to do business.

As detailed in the Chapter 3 discussion of modernization and subsequent recommendations, DoD should implement a regular upgrade schedule for integrated circuits to ensure that manufacturing capacity is always available and manufacturers are more incentivized to be suppliers to DoD. (See Recommendation 4.2.)

PRINCIPLE 4: COORDINATE CLOSELY WITH THE DEPARTMENT OF COMMERCE AND OTHER AGENCIES

The Microelectronics Commons (ME Commons) may be an effective way to fund research and development (R&D) on disruptive technologies for DoD use, but certain provisions are important to ensure that useful results ensue. To start with, most members on the committee with experience in developing advanced semiconductor technologies feel that a 5-year funding timeframe is very likely too short for a disruptive technology to make it all the way through the research, development, and prototyping phases to being ready for manufacturing—a process that usually takes 10 years or more. What will happen when disruptive technologies get halfway through the “valley of death” and then funding stops? If DoD and the Department of Commerce (DOC) collaborate in setting up and running new capabilities such as the ME Commons and the National Semiconductor Technology Center (NSTC), then there is a greater chance of being granted longer-term funding from Congress and achieving ultimate success in bringing innovative technologies from the research phase to manufacturing. DoD should identify down-selection criteria for new project and center proposals to better ensure that centers invite the strongest partners to participate. Another important provision is that projects have clearly defined deliverables that are measurable. And when projects are not making good progress toward their deliverables, then those projects should be terminated to free up funds for more promising projects and hubs.

The intention of ME Commons to set up core facilities (laboratories and fabrication plants—labs and fabs) is certainly important in being able to demonstrate the disruptive technologies needed by DoD, but these laboratories and fabs can be very expensive and time-consuming to set up. The high capital cost of semiconductor manufacturing facilities is the main reason why the industry is now concentrated in relatively few locations and few companies. The ME Commons should try to use existing facilities at universities, such as the Albany NanoTech Complex, the Interuniversity Microelectronics Centre (IMEC), the National Institute of Standards and Technology (NIST), and government laboratories, as a way to get quick access to processing facilities. But where new equipment or facilities are needed, ME Commons should not try to do this on its own. ME Commons should use its money to expand existing government facilities. It makes no sense to duplicate the facility efforts of NSTC and the National Advanced Packaging Manufacturing Program (NAPMP) when most of this expensive equipment can and should be used for both DoD and DOC purposes.

In pursuit of developing disruptive technologies for DoD use, it might sound like a simple matter for DoD to provide R&D funds to the appropriate institutions,

whether they be commercial, academic, or start-ups and small firms. But challenges that DoD can face in getting these institutions to commit to taking on these projects fall into the following two categories: (1) commercial companies may not want to be distracted by R&D projects that are too far from their main business, and (2) academic researchers and small to medium-sized companies may be very interested in the project but do not have access to the SOTA lab/fab facilities needed to do the research and bring the technology to full manufacturing. The NSTC and NAPMP facilities being planned by DOC are likely good options for supporting disruptive technology projects at universities and small to medium-sized companies as they evolve from research to prototyping phases.

As detailed in Chapters 3, 4, and 5, DoD and DOC have a common interest in supporting the U.S. semiconductor industry and developing advanced technologies. Thus, DoD and DOC should collaborate in setting up and running the expensive R&D facilities needed to do this. (See Recommendation 4.4.)

PRINCIPLE 5: COLLABORATE WITH LEADING COMPANIES FOR CUSTOM NEEDS

Convincing a commercial company to take on a disruptive technology project should start with DoD partnering with companies that already have a technology or products that are not significantly different from the new technology or would be a technology that would be commercially interesting to them. And if the disruptive technology is expected to be a derivative version of commercial technology, then DoD should strive to be a fast follower (Box 6-1).

DoD should offer intellectual property (IP) rights that are not too restrictive for the commercial company and ensure that red tape associated with the procurement process is not overly complicated and onerous.

To summarize this point, and as discussed in earlier chapters, when DoD needs a customized process to meet its needs, it should collaborate with a leading-edge manufacturer in a way that minimizes red tape and has clear financial benefits for the manufacturer.

PRINCIPLE 6: ENSURE UNIVERSITY RESEARCHERS AND START-UPS HAVE ACCESS TO ADVANCED EQUIPMENT

Academic researchers and start-up companies may have much to offer DoD in terms of out-of-the-box thinking and an eagerness to take on revolutionary projects. But their lack of access to modern lab/fab facilities may prevent their ideas from ever reaching the prototype phase or manufacturing readiness. This is a common problem for academic researchers and start-up companies, whether the technology being researched is for DoD or commercial application. To address this need, the CHIPS and Science Act of 2022 (the CHIPS Act), under DOC/NIST direction, is proposing that NSTC provide access for these smaller institutions to a modern 300 mm wafer processing facility. Such a facility will be very expensive to set up and run and calls out for DoD and DOC to both contribute funds to operate this facility in a collaborative manner. In some cases, universities have access to process tools needed for the research, but the lack of a professional staff to operate and maintain these tools significantly limits their availability and usefulness. Thus, some money from DoD or DOC can be valuable when used to support professional staff at university facilities.

DOC’s NSTC program is intended to be set up to support the advanced semiconductor process needs of universities and start-up companies. DoD (including through the ME Commons) should collaborate with DOC to ensure that NSTC can support the academic projects and start-up companies that are of potential value to DoD. (See Recommendation 5.2.)

PRINCIPLE 7: SUPPORT ACCESS TO SCALE-UP CAPITAL

When the proposed NSTC facility becomes operative, it will certainly be very helpful for start-ups and university groups that need prototyping facilities. However, most hardware innovations, such as those that would result from semiconductor start-ups, require sustained funding for much longer periods of time than private investors are willing to provide funding. In the long run, when such technologies have potential to make a difference in DoD operations, DoD should consider nurturing them by playing a role akin to a patient venture capitalist. There are existing organizations that could help with this activity. DoD and the

intelligence agencies have similar aims in this area, so the Director of National Intelligence’s (DNI’s) In-Q-Tel could assist in this aim. The Defense Innovation Unit (DIU) and DoD’s new Office of Strategic Capital could also assist. Focusing on DoD needs and its objective of sustaining an engagement with the semiconductor industry would be critical to making sure that DoD does not fall behind and yet is simultaneously contributing to supporting innovations that will keep the U.S. semiconductor industry ahead of competitors.

To summarize, DoD and DOC should both be looking for long-term sources of funding to support revolutionary integrated circuit projects that will need more than 5 years to reach manufacturing readiness. (See Recommendation 5.3.)

PRINCIPLE 8: SUPPORT THE TRANSITION FROM PROTOTYPING TO MANUFACTURING

The Flemish Government helped to establish and has continued to fund IMEC in Belgium for four decades, and the European Union (EU) has also made investments. The amount of government funding has been rather stable, with the fraction of the total budget from government sources gradually decreasing as IMEC became more successful in attracting funding from commercial companies. IMEC is a remarkable source of advanced 300 mm processing capabilities, for large and small companies, although the IMEC infrastructure is not complete enough to provide a path to manufacturing readiness. Ideally, with DOC and DoD support, NSTC could provide IMEC-like capabilities for U.S. companies doing R&D on advanced technologies. But it must be recognized that IMEC and the Albany NanoTech Complex are excellent facilities for supporting research, development, and prototyping, but not for bringing projects to full manufacturing readiness. Projects need to transition to a commercial manufacturing facility to have any reasonable hope of reaching manufacturing readiness. Only commercial manufacturers have the tools, personnel, and experience to reach these later stages.

There may be several semiconductor facilities that can be used for early R&D, but thought needs to be given up front to how technologies coming through these organizations can be transitioned to a commercial manufacturer. The sooner a commercial manufacturer is involved, the better, and government support of this transition will accelerate innovation.

PRINCIPLE 9: BUILD INTERNATIONAL COLLABORATIONS FOR ACCESS TO THE INTERNATIONAL SEMICONDUCTOR ECOSYSTEM

It might seem ideal to return to the 1960s–1970s when the United States had a dominant position in all parts of the semiconductor ecosystem: logic technology, memory technology, wafer fab equipment, semiconductor manufacturing chemicals, and innovative circuit design. However, those days are long past and highly

unlikely to return. Today there is a global semiconductor ecosystem with strong players in Asia, Europe, and the United States, and the United States has benefited from being part of this global ecosystem. The CHIPS Act is a good step in strengthening the U.S. semiconductor industry, but in sectors where the United States does not have a leadership position, multiple suppliers from friendly countries should be set up. To this end, DoD, DOC, and the Department of State should collaborate on identifying and nurturing friendly countries that are home to key suppliers to the semiconductor ecosystem. The goal should be to strengthen both the United States and its global partners. DoD should also consider doing collaborative R&D projects with strong non-U.S. companies in friendly countries where no good alternative exists in the United States.

International competitors are investing heavily in their own semiconductor capabilities to build strong industries. For example, China has announced investments of some $150 billion. The European Union is investing $43 billion, India $30 billion, Japan $6.5 billion, and South Korea has announced its own significant investments, as discussed in Chapter 2. In sum, the United States is entering with sizable investments but is somewhat late to the race.

The United States needs to establish strong working relationships with semiconductor organizations in allied and friendly nations. To ensure a geographically diverse and robust integrated circuit supplier base, DoD, DOC, and the Department of State should collaborate to identify strong suppliers in friendly countries that can augment the U.S. supplier base. (See Recommendation 4.3.)

PRINCIPLE 10: STREAMLINE ENVIRONMENTAL REVIEWS

Another area of consideration for DoD concerns environmental regulations and controls that are applied to the construction of semiconductor facilities. DoD and DOC should collaborate to ensure that the National Environmental Policy Act and other environmental reviews are streamlined and do not pose a significant delay in the construction of advanced semiconductor fabs.

PRINCIPLE 11: FOSTER INFORMATION EXCHANGE

An alternative mechanism for DoD to keep abreast of cutting-edge developments in microelectronics that may significantly impact its operations that depend heavily on electronics is to work with industry to develop a framework that fosters early information exchange on leading-edge microelectronics technologies. Such information exchange has the potential to be a win–win proposition for the companies involved and for DoD. Companies acquire a willing and ready customer, and DoD gets access to state-of-the-art technology. The critical challenge is finding a workable framework that is agreeable to both parties. While it is common today

for private contractors to be embedded in DoD facilities to improve efficiency of operations, the reverse process of embedding DoD employees can be more difficult. DoD may therefore have to incentivize companies to assign technical liaisons, who are full-time employees of the companies but whose special role is to look out for how cutting-edge microelectronics could benefit DoD. These individuals would be responsible for interfacing with DoD counterparts, and for making DoD aware of developments that should be considered for upgrading or replacing electronic systems.

DoD should be kept aware of leading-edge technologies developed by commercial industry by employing their own team of semiconductor technology experts who are continually in touch with industry leaders and can then make recommendations to DoD program managers in selecting the most appropriate integrated circuit technologies.

All these principles are general, with specifics captured in Chapters 4 and 5 in the form of recommendations. These principles aim to provide general guidance for DoD in its effort to assure access to the microelectronics it needs for success.