Semiconductor industry sales for the year 2016 reached $339 billion worldwide, with $164 billion in sales going to U.S. companies alone. And yet, as Daniel Holcomb of the University of Massachusetts Amherst Electrical and Computer Engineering Department says, there is a burgeoning danger in how companies currently manage their semiconductor supply chains. “Supply-chain threats such as counterfeits and hardware Trojans can compromise reliability of integrated circuits and lead to unexpected or malicious functionalities embedded within them,” says Holcomb.
Holcomb adds that “These threats endanger critical infrastructure, national security, and semiconductor industry profits. While some of the biggest concerns go beyond financial considerations, in financial terms it is estimated that counterfeits alone lead to billions of dollars in annual losses to semiconductor companies.”
This growing national security threat explains why Holcomb was recently awarded a five-year, $596,160 CAREER Award from the National Science Foundation (NSF) to study supply-chain security for integrated circuits. Holcomb’s work, when effectively completed, will help to provide a framework for securing the electronic parts upon which so much of modern society relies.
As Holcomb explains about his NSF grant, “The proposed work, if successful, will positively impact national security and society in general by helping to keep possibly-malicious parts out of critical systems such as defense aircraft and medical devices.”
Holcomb also notes that “The research in this proposal can have significant impact on how the supply chain is secured five to 10 years into the future.”
As Wesley K. Clark and Peter L. Levin said in the November/December 2009 issue of Foreign Affairs: “In addition to building diverse, resilient IT infrastructure, it is crucial to secure the supply chain for hardware. This is a politically delicate issue that pits pro-trade politicians against national security hawks. Since most of the billions of chips that comprise the global information infrastructure are produced in unsecured facilities outside the United States, national security authorities are especially sensitive about the possibility of sabotage."
For those of us not well-versed in corporate jargon, Investopedia defines a “supply chain” as “a network between a company and its suppliers to produce and distribute a specific product, and the supply chain represents the steps it takes to get the product or service to the customer.” Consistent with this definition, Holcomb’s work addresses threats on both the production and distribution sides of the integrated circuit supply chain.
Related to the semiconductor industry, fabrication in advanced technology nodes is necessary to compete in the current ultra-competitive market, Holcomb explains, but bringing up a fabrication line in cutting-edge technology now costs billions of dollars, which is beyond the reach of most companies. One answer is outsourcing the fabrication, testing, and distribution of chips, thus driving a trend toward fabless design with manufacturing consolidated in a small number of large foundries that can amortize the outlay of advanced fabrication technology over massive production volumes.
Yet threats to such cost-effective production methods “are hard to mitigate,” says Holcomb, “because the global nature of the supply chain presents a vast attack surface; the process of transforming a design from an initial specification to a deployed integrated chip involves a variety of actors operating across design abstractions, companies, and political boundaries. Some of the actors are inherently trusted, but others are not. Preventing the untrusted actors from causing harm is the focus of this proposal.”
Holcomb says his research proposes a new framework for supply-chain security organized around three main thrusts. The first thrust proposes a new design style for digital chips that uses programmable logic to prevent malicious actors in the supply chain from understanding a design well enough to manipulate it in a targeted manner. The second thrust of the work addresses the design side of the supply chain, and seeks new formal abstractions for understanding hardware Trojans that could be inserted by a malicious actor in a foundry. The third and final thrust addresses the distribution side of the supply chain and proposes a low-cost authentication framework that aims to keep counterfeit integrated circuits out of products and systems.
Holcomb reveals that his long-term career goal is to continue finding new ways to improve security in the embedded systems that take on ever more significant roles in infrastructure, defense, medicine, and other critical applications.
Economic trends in the semiconductor industry have created an environment of outsourced production and distribution that makes supply-chain security a leading threat, but ultimately a secure supply chain should be a foundation and first step for building and deploying billions and billions of secure embedded systems in the field.
“If we are going to entrust so many critical functions to electronic systems,” Holcomb declares, “we must have a supply-chain framework in place for trusting the chips.”
As Holcomb concludes, “The scientiﬁc objective of this project is to develop efficient, scalable, and rigorous design frameworks for preventing malicious insertions by an untrusted foundry, and to create usable authentication technologies to keep counterfeit parts out of critical systems.” (February 2018)