When NSPM-33 was signed in February 2018, it was framed as a decisive step in securing critical technologies, with domestic semiconductor manufacturing placed squarely at the heart of the national security conversation. For many in the industry, this presidential memorandum marked the culmination of years of quiet concern. But for others—particularly those more focused on design innovation than supply-chain logistics—the directive felt abrupt, as if the underlying vulnerabilities had materialized overnight. Of course, they hadn’t. The fragility of semiconductor supply chains, especially in their reliance on rare-earth elements and complex international processing routes, had been building for decades, hidden in plain sight.
For chip designers, the memorandum triggered a period of introspection, coupled with the immediate practical task of mapping wafer-supply chains in unprecedented detail. The industry’s previous tolerance for opacity in upstream sourcing was no longer tenable. Designers, whether they liked it or not, now had to engage directly with questions of material origin—questions that had once been the domain of procurement specialists or, at best, distant suppliers. In this regard, the U.S. Geological Survey’s open data on rare-earth processing sites emerged as an unexpected but invaluable tool. The datasets, though originally developed for broader geological and economic monitoring, provided a starting point for firms looking to trace the origins of key inputs.
Yet, integrating this data into meaningful supply-chain maps was not as straightforward as some may have hoped. The USGS data was comprehensive, yes, but also generalist by design. It identified processing sites, mining operations, and trade flows, but did not align neatly with the part numbers, lot codes, or procurement records that chip designers worked with daily. Bridging this gap required a concerted effort: designers found themselves working alongside supply-chain analysts, data scientists, and, in many cases, external consultants tasked with translating geological data into actionable supply-chain intelligence. The process was iterative and, at times, frustrating. Many designers remarked that the very act of trying to build these maps exposed just how little visibility they had had into critical material flows before NSPM-33 brought the issue to the fore.
The development of annual “critical microelectronics” supply-chain risk assessments became an essential component of this new transparency regime. These assessments, intended not merely as internal documents but as inputs into broader national security reporting structures, required firms to balance technical precision with a degree of narrative clarity that did not always come naturally to engineering teams. The task was complicated further by the inherent uncertainty in many upstream supply chains. For all the data available through sources like the USGS or trade monitoring services, gaps persisted—especially where materials passed through intermediary refiners or traders operating in jurisdictions with limited reporting obligations.
Some firms, recognizing the limitations of public data alone, began supplementing their assessments with supplier self-disclosures or, where feasible, on-site audits. These efforts were not without their challenges. Suppliers, particularly those several tiers removed from final assembly, varied widely in their willingness and ability to provide the necessary data. Moreover, the very global nature of semiconductor supply chains meant that even well-intentioned suppliers sometimes struggled to trace inputs back to original processing sites with the granularity required. It was, in short, a learning process—one that demanded patience and persistence in equal measure.
The broader industry response to NSPM-33 reflected a mix of pragmatism, caution, and, occasionally, quiet resistance. While most firms acknowledged the legitimacy of national security concerns, there was also unease about the potential for overreach or unintended consequences. The semiconductor sector, after all, had long thrived on global integration. Reconfiguring supply chains to meet new security requirements risked introducing inefficiencies or, worse, stifling innovation by constraining access to best-in-class materials and technologies. This tension was rarely discussed openly, at least not in official statements or formal risk assessments, but it was palpable in private conversations among industry leaders.
As firms worked to draft their annual risk assessments, they grappled with how to reflect this complexity without triggering undue alarm or appearing noncompliant. Some opted for cautious phrasing, noting areas of incomplete visibility or emerging risks without committing to specific mitigation timelines. Others took a more assertive approach, highlighting proactive steps—new supplier vetting procedures, pilot domestic sourcing programs—even where these were still in early stages. The resulting documents, while varied in style and substance, shared a common feature: they marked a shift in how semiconductor firms conceptualized supply-chain responsibility. Where once supply-chain security had been a background consideration, it now sat alongside cost, performance, and time-to-market as a core design and operational priority.
In many respects, NSPM-33 catalyzed a cultural change within the semiconductor sector. Firms that had long prioritized technical excellence above all else began to engage more seriously with the geopolitical context in which they operated. This was not a comfortable shift for everyone. Engineers and designers found themselves navigating unfamiliar terrain, balancing innovation imperatives with security constraints that were sometimes vaguely defined or subject to rapid change. But the experience also fostered new capabilities—teams developed stronger cross-functional linkages, and supply-chain mapping, once seen as peripheral, became a recognized area of strategic expertise.
Where this would ultimately lead remained, in late 2018, an open question. The industry was only just beginning to absorb the implications of NSPM-33, and the balance between openness and security would likely continue to evolve. What was clear, though, was that the era of supply-chain invisibility had ended. In its place was a new, more demanding standard—one that required semiconductor firms not just to design cutting-edge chips, but to understand, document, and, where necessary, reconfigure the complex material networks that made those designs possible.