Breaking The Circuit: US-China Semiconductor Controls – Analysis
By Published by the Foreign Policy Research Institute
By Catherine Tan
(FPRI) — In October 2022, the Biden administration introduced export controls to limit China’s access to advanced US semiconductors and technologies, aiming to maintain US technological superiority and address security concerns. These controls, targeting areas such as advanced chips and supercomputer components, were tightened in 2023. While the controls have disrupted China’s semiconductor industry in the short term, concerns persist about their long-term effectiveness and potential Chinese retaliation, along with significant policy gaps that will be explored further in the following brief.
“Small Yard, High Fence”: The Biden Administration’s Strategy
In the early 2000s, bringing China into the global community was widely seen as a strategic decision. The Bill Clinton and George W. Bush administrations supported integrating China’s economy into the international rules-based system, believing economic interdependence would promote stability and mutual economic gain. However, this did not lead to China’s democratization as hoped. Instead, China has used this economic partnership to implement a “Military-Civil Fusion” plan, which uses civilian technology—much of it from US partners—to strengthen China’s military capabilities. Many have now come to see technological ties with China as a potential vulnerability. Concurrently, however, the United States has recognized its own capacity to exploit this interdependence to gain strategic advantages over China.
Despite their economic interdependence, the United States and China have distinct strengths. The United States is exploiting these imbalances by controlling key “chokepoints” in the semiconductor industry, where it holds a significant advantage. With the United States and its allies dominating over 90 percent of global semiconductor equipment manufacturing, China is heavily reliant on foreign sources for chips. The Biden administration is using this leverage to slow China’s chip industry growth, particularly in artificial intelligence (AI). They’re adopting a “small yard, high fence” approach by regulating critical supply chain points like lithography equipment from the Dutch company ASML and advanced AI chips, aiming to target key areas while minimizing broader economic disruptions.
Significant Wins, but Gaps Remain
Since the rollout of the export controls, China’s semiconductor industry has taken a serious blow. Chinese semiconductor output plummeted 17 percent in early 2023. Furthermore, the US restrictions on Chinese access to Nvidia’s A100 and H100 chips and Dutch firm ASML’s high-tech lithography machines, which are essential for making advanced chips, seriously damage China’s chip-making capabilities. Experts estimate that China’s semiconductor capacities are anywhere from five to ten years behind.
However, the export controls are still in their leaky early phase, and the Biden administration is constantly scrambling to plug holes as companies find ways to avoid losing the huge China market. Immediately after the October 2022 sanctions, Nvidia quickly tweaked its H100 and A100 AI chips to keep the data transfer rate just under the new rules’ limits. But when even tighter controls dropped in October 2023 and closed that loophole, Nvidia pivoted again, rolling out three brand new AI graphics cards tailor-made for the China market. The specs on Nvidia’s new graphics processing units (GPUs) show the company is carefully toeing the line, balancing peak performance while staying in step with the latest US rules.
Another loophole in US semiconductor export controls has come to light: cloud computing services. Chinese organizations, including some with military ties, are circumventing restrictions by using cloud platforms like Amazon Web Services and Oracle Cloud. This allows them to harness the power of cutting-edge chips such as Nvidia’s A100 and H100 GPUs without physically owning them. Even companies on the US Entity List have found ways to access powerful computing resources through the cloud. For example, iFlyTek, a Chinese AI firm blacklisted due to human rights concerns, has legally rented high-performance Nvidia chips via cloud services despite sanctions. This issue may escalate as Nvidia expands its cloud supercomputing offerings in China through local partnerships.
Experts are also increasingly concerned about chip smuggling as a means to bypass export controls. While an estimated 10,000 chips are smuggled into China annually, this figure isn’t alarming on its own, considering global tech giants use millions of chips each year. What’s truly worrying, however, are reports of entire servers being traded on the black market. These servers, each housing eight top-tier Nvidia chips and valued at over $300,000, could indicate the rise of sophisticated smuggling networks that industry watchers fear could grow significantly over time.
To address these issues, some cybersecurity experts suggest fitting server racks with global positioning system trackers to monitor their locations and usage. Additionally, the US government should consider imposing stricter reporting requirements on American cloud service providers to prevent sanctioned entities from accessing restricted technology. These measures could help close the current gaps in export control enforcement.
Can China Retaliate?
The United States may lead in cutting-edge semiconductor tech, but China has sway in other key areas in the semiconductor supply chain—mainly, rare earth elements that are must-haves for making chips. China extracts 60 percent of the world’s rare earth minerals and handles over 85 percent of processing. In response to US export controls, China has restricted gallium and germanium exports to signal its displeasure. While China hasn’t fully weaponized its rare earth resources yet, it’s hinting at the possibility of doing so in the future.
Smartphones contain twenty to forty chips, mostly older “mature node” types handling essential functions like memory and charging. While cutting-edge processors grab headlines, these less glamorous chips are equally vital to device performance. China currently controls around 24 percent of global capacity for 50–180 nm chips, and this share is projected to surge to 50 percent by 2030 as China pours massive subsidies into mature node manufacturing. Many raise the concern that Chinese state-run chipmakers could flood the market with foundational chips, wreaking havoc on chipmakers’ share of the industry. For this reason, the Biden administration recently announced a plan to increase tariffs on semiconductors from 25 percent to 50 percent by 2025, and the US Department of Commerce has already taken steps to survey US semiconductor firms for reliance on mature nodes. These steps are a move in the right direction, proactively addressing potential vulnerabilities while maintaining a competitive edge in the semiconductor market.
The Long Game: Could Export Controls Backfire?
Looking ahead, the semiconductor export controls could have significant unintended consequences for both the United States and China. First, US chip companies face substantial financial risks due to the loss of the Chinese market. The US Chamber of Commerce in China projects that this could result in an annual loss of $83 billion in sales and 124,000 jobs. US semiconductor equipment firms are particularly vulnerable, given their reliance on Chinese sales, with an estimated 30–40 percent of sales going to China. This shift is evident in companies such as SMIC, a major Chinese chipmaker, which has transitioned from 60 percent foreign customer production five years ago to 80 percent domestic production currently. The potential revenue loss could initiate a detrimental cycle where reduced research and development budgets compromise US firms’ global competitiveness.
Paradoxically, these export controls may also catalyze China’s efforts to establish an independent semiconductor supply chain. By removing access to cheaper foreign chips, US policies have inadvertently achieved what years of Chinese government initiatives could not: compelling Chinese companies to develop domestic alternatives. Evidence of this trend is already emerging. In 2023, Huawei released the Mate Pro 60, featuring a 7nm chip produced by SMIC, a feat previously thought impossible without restricted extreme ultraviolet technology machines. While full replacement of US suppliers will take time, the incentives for this transition are now firmly in place.
Another factor to consider is the nonlinear nature of technological innovation. Traditionally, advances have been driven by Moore’s Law—the observation that the number of transistors on a chip doubles about every two years, leading to exponential increases in computing power. This principle has guided the industry’s focus on lithographic scaling, with lithographic machines like those produced by ASML becoming critical technological chokepoints that China cannot replicate. However, the semiconductor industry is increasingly moving away from pure lithographic scaling. As transistors approach atomic scale and fabrication costs skyrocket, industry experts anticipate that computing power gains from Moore’s Law will plateau by 2025, pushing the sector to explore alternative avenues for improving chip performance.
This shift gives China a chance to focus on areas such as memory-centric computing and system-level optimizations, which may help them catch up to the United States technologically, even without advanced lithography machines. By investing in these technologies, China could create high-performance chips that sidestep US restrictions and weaken the effectiveness of US sanctions. This broader separation could isolate the United States from the global economy—a situation some have compared to “the technology equivalent of de-dollarization.”
US export controls have disrupted China’s high-end chip production in the short term, but long-term success depends on adapting to rapid tech changes. Policymakers must monitor and adjust regulations to maintain a competitive edge, focusing on closing loopholes, tracking China’s chip capacity, and developing alternative rare earth supply chains. Balancing aggressive controls with global influence is crucial, as effectiveness hinges on responding to emerging industry challenges.
- About the author: Catherine Tan is a former research intern for the Asia Program and a rising junior at Colby College studying Anthropology and Global Studies.
- Source: This article was published by FPRI