SIA's Latest Assessment: Deep Risks in the Chip Supply Chain Have Been Severely Underestimated

The Semiconductor Industry Association (SIA) recently released an in-depth report revealing that the global supply chain underpinning the U.S. chip industry is far more fragile than surface-level data suggests. While 95% of equipment and materials used by the U.S. semiconductor industry come from suppliers headquartered in the U.S., Europe, or Japan, that proportion plunges to 68% when traced to actual production locations, with the remaining 32% dependent on production capacity in mainland China, Israel, Southeast Asia, South Korea, and Taiwan. The report warns that weak links hidden deep in the upstream global supply chain are becoming the most dangerous vulnerability in the U.S. chip ecosystem.

The report, titled "Understanding Dependencies in the U.S. Semiconductor Supply Chain," represents the SIA's first systematic quantification of the fragility of the upstream supply chain serving the U.S. chip industry. It notes that semiconductor manufacturing takes months and requires sourcing a wide range of materials, equipment, and components from multi-tiered suppliers across the globe. Chipmakers work directly with more than one thousand Tier 1 suppliers to procure bare wafers, specialty gases, photoresist chemicals, and manufacturing equipment such as lithography and deposition tools. Each of these Tier 1 suppliers, in turn, maintains an intricate network of sub-tier suppliers involving intermediate products and raw material providers.

To assess supply chain risk, the SIA collaborated with procurement and strategy leaders from chipmakers, fabless design companies, backend packaging firms, equipment manufacturers, and materials suppliers to develop a vulnerability index spanning five dimensions: geographic concentration of production, degree of dependence, supplier concentration for specific products, ownership structure of critical suppliers, and dependency relationships at the sub-tier supplier level. The analysis covered 200 key Tier 1 equipment and material inputs, along with nearly 100 categories of upstream raw materials, engineered materials, and sub-components.

The Stark Gap Between Headquarters Location and Production Site

One of the report's core findings is that supply chain analyses based on supplier registration locations severely obscure true dependency relationships. The U.S. semiconductor ecosystem spends approximately $27 billion annually on equipment and materials, with 95% of that supply coming from suppliers headquartered in the U.S., Europe, or Japan. However, when assessing actual production or assembly locations, the from the U.S., Europe, and Japan drops to 68%.

This "headquarters-production separation" phenomenon is pervasive across multiple critical areas. For example, U.S.-headquartered semiconductor manufacturing equipment suppliers assemble physical vapor deposition (PVD) equipment in Singapore and etch equipment in the United Kingdom. U.S. materials suppliers produce photoresists and ultra-high-purity gases needed by chipmakers at chemical plants in Japan, Taiwan, and South Korea. By contrast, bulk gases such as hydrogen and argon are typically produced by European companies, but their production often occurs through air separation units located near wafer fabrication facilities in the United States.

Applying the vulnerability index to the Tier 1 supply chain, approximately half of the analyzed categories were classified as high (11%) or medium (38%) vulnerability. High-risk Tier 1 inputs include: rare inert gases such as xenon and krypton used in lithography and plasma etching; advanced packaging equipment with roughly 65% of capacity concentrated in Singapore, Malaysia, and South Korea; Ajinomoto build-up film (ABF) substrates used in high-density packaging for AI accelerators, with Japan accounting for approximately 75% of supply; and DUV/EUV photomask blank substrates, with Japan supplying roughly 70%.

Hidden Risks in Upstream Raw Materials

The report's most striking findings come from its deep-dive analysis of the sub-tier supply chain. Chipmakers typically do not purchase raw materials directly from mineral producers; instead, they buy highly purified chemicals or components from specialized materials suppliers. After evaluating nearly 100 categories of upstream Nth-tier base materials, engineered materials, and sub-components, the SIA found that nearly one-third of upstream inputs fall into the high-risk category.

These high-risk materials include 13 of the 17 rare earth elements (scandium, yttrium, cerium, neodymium, dysprosium, terbium, lanthanum, erbium, thulium, lutetium, europium, gadolinium, and samarium), as well as critical minerals such as gallium, germanium, and graphite, along with precursor chemicals. The overwhelming majority of production capacity for these materials is concentrated in China or Russia.

The report details a timeline of China's export controls on critical materials in recent years: export restrictions on gallium, germanium, and graphite were imposed in 2023, escalating to export bans the following year; export controls on seven medium and heavy rare earth elements were implemented in April 2025; and in October of the same year, the scope was further expanded to cover additional rare earth categories, mining and smelting equipment, and rare earth-related technologies.

According to Chinese customs data, as of June 2026, China's exports to Japan of certain controlled critical minerals—including gallium, dysprosium, terbium, and yttrium—remained at zero or negligible levels. Given that Japanese suppliers play a significant role in supplying semiconductor materials, components, and manufacturing equipment to the United States, prolonged restrictions on these raw materials could create further downstream supply risks for the U.S. semiconductor industry.

The report also specifically highlights the vulnerability of polysilicon. China is the world's largest polysilicon producer, accounting for more than 90% of total capacity, almost entirely of solar grade. Although approximately 85% of the electronic-grade polysilicon market is held by a small number of producers in the U.S., Europe, and Japan, these suppliers must rely on large-scale production of solar-grade polysilicon to achieve the economies of scale needed for producing semiconductor-grade ultra-high-purity polysilicon. China's rapid expansion of solar-grade capacity, along with advances in electronic-grade production technology, is creating increasingly intense competitive pressure on existing qualified suppliers.

The most sobering finding: after accounting for sub-tier dependencies, more than three-quarters of Tier 1 supply categories depend on at least one high-risk Nth-tier input, and more than half depend on two or more high-risk upstream materials. Many Tier 1 equipment and materials originally rated as low- or medium-risk conceal critical vulnerabilities upstream. For example, core semiconductor equipment such as chemical vapor deposition (CVD), electrochemical deposition (ECD), sputtering, and ion implantation tools all depend on highly specialized custom components—turbo pumps, dry pumps, vacuum robots, and transfer valves—often available from only a single supplier; die attach film depends on gallium, BT packaging substrates depend on tungsten, and nitrogen trifluoride depends on fluorite.

A Multi-Pronged Response Strategy

Faced with such complex supply chain risks, the SIA proposes a four-part strategic framework.

First, expand domestic production and strategic stockpiles. The report acknowledges that relocating all materials and equipment needed for semiconductor production to the United States is neither practical nor economically viable, but for certain critical inputs, establishing or enhancing domestic capacity is a feasible and essential near-term strategy. For example, ABF substrates and photoresists overwhelmingly come from Japan, and investing in U.S. domestic production capacity would enhance supply chain resilience. Sputtering targets represent another area well-suited for expanding domestic capacity.

For materials that cannot be rapidly localized, building strategic inventories serves as an important buffer. Semiconductor manufacturers already use excess inventory to guard against supply disruption risks, often maintaining up to 12 months of stock where feasible—for example, rare earth magnets, neon gas, and motors and motor components. However, not all inputs can be stockpiled. Scandium metal is particularly difficult to stockpile due to its rapid oxidation rate and shelf life of only 9 to 15 months. Certain fluorinated gases are subject to U.S. and global environmental regulations, preventing companies from purchasing and building multi-year supply buffers. The report recommends that the U.S. government consider establishing exemption provisions to existing environmental regulations to allow semiconductor manufacturers to import and stockpile such gases.

Second, build and maintain allied supply chains. The supply chain supporting the U.S. semiconductor ecosystem is inherently global. Japan, Taiwan, Malaysia, Singapore, South Korea, and European partners including Germany, France, Italy, the Netherlands, and the United Kingdom are important production locations for critical engineered materials and components needed in semiconductor manufacturing equipment. Producers in these countries and regions depend on sourcing critical mineral raw materials from Australia, Canada, India, and Mexico. Cooperation with Vietnam and Indonesia, including through public-private partnership investments in local refining and processing infrastructure, could also help expand processing capacity outside high-risk regions.

Third, reduce dependence on unreliable sources of rare earths and critical minerals. The report stresses the urgent need to establish new mining and processing capacity. Beyond expanding domestic mining and refining capacity where feasible, the U.S. government needs to work in coordination with other governments to support raw material extraction and processing projects already underway in Australia, Canada, Japan, Mexico, and South Korea. Coordinated policies—such as long-term offtake agreements—are essential to ensuring sufficient demand to support new critical mineral capacity. The report cites multilateral efforts including the State Department-led Pax Silica initiative, the U.S.-EU Critical Minerals Supply Chain Resilience Action Plan, and the G7 Leaders' Declaration on securing critical minerals supply chains.

Fourth, increase R&D investment. The report calls on the U.S. and partner governments to increase and sustain funding for research and development in materials science, chemistry, and engineering. The Department of Energy-funded Minerals to Materials Supply Chain Facility (METALLIC) project, involving nine U.S. national laboratories, holds promise for developing new extraction and processing methods for mineral production. Advanced materials research grants provided by the Commerce Department's CHIPS Program Office also represent important support. Recycling technology research is equally promising—for example, emerging wafer recycling technologies for silicon carbide substrates could reduce substrate consumption per device by up to 10 times.

The report concludes that building a trusted and reliable semiconductor supply chain is a long-term endeavor requiring sustained coordination and cooperation among government, industry, and international partners. Given the strategic importance of the semiconductor industry to the U.S. economy and national security, enhancing the resilience of the global supply chain that supports the domestic chip ecosystem must remain a top priority for both industry and policymakers. If these efforts are pursued with unwavering commitment, the U.S. semiconductor ecosystem and its diverse customer base stand to become stronger, more resilient, and more competitive over the decade.

The deep risks revealed in the report also serve as an important warning to participants across the global semiconductor industry: in a highly interconnected global supply chain, surface-level data often obscures true dependency relationships, and the real vulnerabilities may lie hidden in the raw material stages at the very top of the supply chain.