The Defense Advanced Research Projects Agency (DARPA) must ensure that its federal research and development expenditures bolster U.S. defense capabilities and commercial sector and not those of potentially adversarial nations. To support DARPA’s Microsystems Technology Office (MTO) in its technology transfer efforts for the Electronics Resurgence Initiative program, the Potomac Institute for Policy Studies investigated the global proliferation of select technologies that were strongly reliant on initial DARPA funding to catalyze their development. DARPA Spintronics, which launched commercial efforts related to magneto random access memory (MRAM), was used as a case study, as were the DARPA efforts towards silicon-based RF technologies [i.e SiGe bipolar complementary metal-oxide-semiconductor (BiCMOS)-based phased arrays].The objective of this study was to collect and analyze data to create a report tracking the proliferation of these select technologies, starting from the initiation of U.S. federal investment via DARPA and carrying through the following period of commercial R&D efforts, industrial partnerships, and mergers and acquisitions. To execute this study, information was gathered from interviews with Program participants and individals involved with subsequent commercialization efforts. Reputable internet sources throughout the microelectronic community were also utilized.
Our research found that investments by DARPA were critical to the development of highly capable MRAM memory and SiGe RF technologies. Industry partners were highly reluctant to conduct initial research and development activities- it was not clear that the technologies were viable nor did markets exist to motivate early internal funding. However, as a result of DARPA funding, these technologies have found profitable transitions in the commercial sector and the defense-industrial base. Additionally, the rapid development of both MRAM and 5G-enabling mmWave technologies were greatly assisted by the highly specialized workforce that was trained via DARPA funding. The graduate students who had worked on DARPA projects are now currently employed throughout the domestic and foreign technological ecosystems.
Our research also found that past investment in MRAM and SiGe BiCMOS-based phased array technologies has resulted in commercial sectors strongly reliant on foreign-owned manufacturing capabilities. Despite a significant amount of domestic research and development efforts, there are no MRAM or SiGe-capable mass production foundries owned and operated by U.S.-based firms and no U.S.-owned firm has plans to build one. Indeed, high-volume low-cost semiconductor manufacturing is largely located in the Asia/Pacific region. The limited availability of domestic fabrication options raises barriers that discourage domestic commercialization of USG-funded microelectronic technology.