Semiconductor traceability and provenance, long discussed in specialized supply-chain and security circles, are rapidly moving into the mainstream of chip industry concerns. That shift was on full display at the Semiconductor Traceability and Provenance Workshop hosted on Jan. 27 by the U.S. National Institute of Standards and Technology (NIST). Among the most prominent participants were hyperscalers and PC companies Google, Microsoft, IBM, and HP; automotive giants Bosch, Stellantis, and Volkswagen; semiconductor manufacturers including AMD, Intel, Micron, and Qualcomm; and EDA vendors Siemens EDA and Synopsys, alongside representatives from U.S. government agencies and industry bodies such as SEMI and USPAE.
The event reflected a growing realization that semiconductor supply chains have become too complex, too global, and too critical to operate without stronger mechanisms for trust and verification. AI, advanced packaging, and heterogeneous integration are increasing the number of actors and components involved in a single system, making it harder to know not only where a device came from but also whether it is functioning as intended. Against that backdrop, traceability is no longer viewed merely as a logistics or compliance issue, but as a foundational element of security, reliability, and economic resilience.
www.eic.net.cn
According to Kostas Amberiadis, the NIST official who organized the workshop, the event was built directly on an earlier NIST meeting held in April 2025, focused on trust and provenance in the semiconductor supply chain. In that earlier session, attended primarily by semiconductor companies and EDA firms, traceability emerged as the top short-term and long-term priority for industry participants. The consensus was that the discussion needed to expand beyond chipmakers themselves to include the broader ecosystem that consumes and integrates semiconductors.
“The semiconductor ecosystem is a lot bigger than just the semiconductor companies,” Amberiadis told EE Times in an interview. “Last year, we had the big semiconductor and EDA players—companies like Qualcomm, Intel, AMD, IBM, and Cadence. But they looked around the room and said, ‘We don’t have hyperscalers, we don’t have personal-computing, we don’t have auto companies.’”
www.eic.net.cn
The agenda of the second workshop was deliberately structured around viewpoints from hyperscalers and personal-computing companies, automotive manufacturers, government agencies, and the semiconductor and EDA sectors. The goal was not to dictate a single solution but to surface differing interpretations, technical approaches, and business constraints that influence how traceability is defined and implemented across sectors.
Amberiadis noted that attracting those companies was not trivial. “It’s not easy to get them into the same meeting,” he said. “We had presentations from Google, Microsoft, IBM, and HP. On the automotive side, we had Bosch, Stellantis, and Volkswagen. These are huge companies, and if the big guys say something, the small ones follow.”
One of the clearest outcomes of the workshop was that “traceability” does not have a universal meaning, even among semiconductor companies. For some, traceability is closely tied to serialization—assigning a unique identifier to each device, either physical or virtual IDs. For others, it extends into digital records, system-level data architectures, or metrology-based verification methods.
Amberiadis emphasized that this diversity of definitions is not necessarily a problem but a reality of an industry that spans everything from commodity microcontrollers to high-performance AI accelerators. “Traceability means different things to different people,” he said. “There is no universal definition.”
What participants appeared to agree on, however, is that any workable approach must be bounded and practical. Efforts that attempt to cover every material, wafer, die, package, and system layer simultaneously risk becoming too complex or too costly to implement at scale. Instead, many discussions gravitated toward focusing first on finished devices and packages, where identification and verification can have the most immediate operational impact.
www.eic.net.cn
Amberiadis said that if there was one recurring theme throughout the workshop, it was cooperation—tempered by a strong desire to preserve existing business models and technical infrastructures. Participants broadly agreed that the problem of semiconductor traceability is too large for any single company to solve alone. At the same time, few expressed interest in abandoning their current internal systems in favor of a universal, top-down framework.
This dynamic was described by Amberiadis as a preference for “cooperation without tearing down what they already have.” In practice, this translates into an industry willing to align on standards, interoperability mechanisms, and shared terminology, but reluctant to expose proprietary data or redesign core processes that provide a competitive advantage. Concerns about IP and data confidentiality were implicit in many discussions about information sharing across the supply chain.
The afternoon panel discussion reinforced this point. It featured representatives from AMD, Intel, Micron, and Qualcomm, alongside EDA companies Siemens EDA and Synopsys, underscoring that both chipmakers and design-tool vendors see traceability as strategically important, but also as an area where competitive sensitivities remain high.
Cost was another factor that surfaced repeatedly. While participants acknowledged that traceability infrastructure carries a price, there was no clear consensus on how those costs should be allocated. The prevailing assumption was that, over time, they would ultimately be absorbed by end consumers, regardless of application sector.
NIST’s role in the workshop was intentionally neutral. Rather than advocating a specific technical solution, the institute positioned itself as a convener and facilitator—bringing together disparate actors to identify common needs and gaps. This stance is consistent with NIST’s broader mission to advance measurement science, standards, and technology in ways that enhance economic security and quality of life.
Where NIST may have its most immediate impact, according to Amberiadis, is in standards development and metrology. Participants repeatedly called for better visibility into existing standards, as well as clearer mappings of where additional standards might genuinely be needed versus where interoperability between current frameworks could suffice. Pilot efforts and reference architectures were also discussed, though these would likely require dedicated funding and cross-industry commitment.
www.eic.net.cn
The rise of advanced packaging and heterogeneous integration added a new layer of urgency to the conversation. Modern systems increasingly combine chiplets from multiple vendors into a single package, complicating efforts to assess reliability, origin, and functional integrity. In such environments, traceability becomes not just a matter of tracking a single supplier but of verifying the trustworthiness of an entire network of contributors.
Participants noted that as AI workloads and high-performance computing continue to drive chiplet-based designs, the need for reliable provenance information will only intensify. Without mechanisms to confirm that each component performs as intended and meets agreed-upon reliability criteria, system-level risks multiply.
www.eic.net.cn
The workshop did not produce a definitive roadmap or a single technical blueprint for semiconductor traceability. Instead, its value lies in exposing the diversity of perspectives and clarifying where alignment already exists, and where it is lacking. Panel discussions and breakout sessions provided opportunities for cross-sector dialogue, allowing automotive manufacturers such as Bosch, Stellantis, and Volkswagen to hear directly from hyperscalers like Google and Microsoft, and government agencies to better understand industry constraints.
For companies that did not attend, Amberiadis offered a succinct takeaway: Semiconductor traceability is too large and too interconnected a challenge to be addressed in isolation. Progress will depend less on imposing uniformity and more on building cooperative frameworks that respect sector-specific needs while enabling interoperability and shared assurance.
www.eic.net.cn
In that sense, the NIST workshop marked less an endpoint than a starting line. As chip supply chains continue to expand in scale and complexity, the industry’s ability to answer a simple question—“Where did this device come from, and can I trust it?”—may become one of its most consequential technical and economic challenges.
易IC库存管理软件
www.eic.net.cn
易IC库存管理软件
