Table of Contents
Heightened geopolitical tensions and the AI-driven push for sovereignty is pushing mandates for end-to-end traceability in the semiconductor supply chain
The global semiconductor materials market, which includes wafer fabrication and packaging materials, is projected to grow from USD 74.85 billion in 2026 to USD 104.22 billion by 2034. Taiwan, South Korea, and China account for nearly 90% of global semiconductor foundry revenue, making Asia the world’s primary hub for sales and manufacturing (with a global market share of about 51%).
In 2025 and now in 2026, geopolitical tensions around AI and the push for AI sovereignty continue to draw attention to global sourcing and semiconductor supply chains, with provenance and “traceability” becoming a much bigger focus. For example, the National Institute of Standards and Technology (NIST) told RCRTech it recently hosted two events focused on traceability and provenance for better proving where chips come from and whether they can be trusted:
- Semiconductor Traceability and Provenance Workshop
- Sustainable Hardware Security (SUSHI) @NIST Workshop
Increasingly, government, industry, and academia are convening to establish documentary standards that help manufacturers follow materials through the supply chain. Currently, traceability is incredibly complex because modern chips can contain more than 100 billion complex nanodevices, many of which are less than 50 atoms across. With more than 800 diverse materials (also with multiple touchpoints), 70 different masks, and up to 2,000 steps in wafer fabrication, provenance, indeed, becomes a daunting task.
The intricacy makes measuring and monitoring advanced chips vastly more difficult, which has made metrology a frontline focus, and one that consumes as much as 50% of semiconductor manufacturing steps.
From production to end use, materials involve multitudes of people (organizations, teams, individuals), locations (facilities, production lines within facilities, storage, equipment, warehouses, mines, barrels, etc.), and physical contact items (additives, solvents, gases, etc.).
Recording all the interconnections and interactions as people inspect, move, store and transport materials requires new levels of certification and accreditation, and a multi-layered approach that integrates physical tracking hardware with digital verification systems, such as:
- U.S. Chips and Science Act, which mandates documentation of origin and integrity of silicon and other raw materials, with blockchain-based ledgers that help foundries establish a chain of custody;
- OECD Due Diligence and platforms like the Responsible Minerals Initiative (RMI), which help to trace minerals like cobalt and tantalum back to conformant smelters. Strict vendor procurement processes also intend to hold sub-suppliers to the same standards;
- NIST Stamp, an evolving project that will push for “immutable” physical markings for chip traceability from design to decommissioning. NIST is also drafting IR 8536, a framework to improve traceability across complex manufacturing ecosystems;
- SEMI T23, GS1, and ISO/IATF standards that ensure consistent tracking across different software platforms;
- SSCA for standardized cyber assessments to verify the security readiness of every supplier in the chain;
- AS6496 and JEP160 frameworks mitigates counterfeit products by ensurng the quality, authenticity, and long-term reliability of electronic components, particularly within the aerospace and high-reliability industries.
In addition, foundries like TSMC build “gigafab” clusters in the U.S., Europe, and Asia, and they log and digitally sign off on every handoff to establish an authenticated chain of custody. Physical identifiers are embedded directly into the silicon during the manufacturing process, as with ECID serial numbers and cryptographic keys, PUF “fingerprints” and high-res laser markings.
Increasingly, they are also using digital twins to simulate and verify production processes, as well as next-gen RFID tags for real-time, non-line-of-sight tracking to maintain quality during transit.
AI-driven workflows at the machine level are also automating the capture of manufacturing metadata, which is expected to greatly improve the integrity of documentation by reducing human error.
In addition to these many evolving safeguards, the Trump administration issued a proclamation under Section 232 of the Trade Expansion Act to reroute certain high-end chips (like the Nvidia H200) manufactured in Taiwan and bound for China to the United States for third-party testing. Also, (as of January 15, 2026), export policy for certain AI chips require a case-by-case review, which means exporters must provide extensive certifications and documentation detailing the number of units shipped to U.S. customers, confirmation that global foundry capacity for U.S. orders is not being diverted, and rigorous “Know Your Customer” (KYC) procedures to prevent unauthorized remote access to the hardware. A 25% tariff on certain advanced semiconductors was also imposed, though with broad exemptions for chips used to build out the domestic U.S. technology supply chain. The exemptions require verified U.S. capacity expansion or proof of domestic use cases like data centers or R&D.
The process has become frought and onerous, so there is copious debate about what makes sense and what does not in terms of tariffs and export restrictions. Currently, the U.S. and China have each put forth tit-for-tat measures, which then have led to more nuanced, “transaction-by-transaction” measures. Regardless of what happens, most agree that traceability and visibility into the supply chain are of paramount importance. RCRTech will continue to cover the story as it evolves.
