On January 2, 2025, President Trump issued an order blocking the acquisition by HieFo Corporation, a company registered in the state of Delaware and controlled by PRC citizen Genzao Zhang, of the production facilities of Emcore, an American semiconductor manufacturer engaged in the design, fabrication, and processing of semiconductor wafers.
Donald Trump’s order is a targeted step by which Washington firmly delineates the boundaries of Chinese capital’s participation in the American chip sector, making it clear that control over production facilities has become a matter of security, and China’s access to the production of American microchips is unacceptable.
There is a shift in the logic of containment, where the focus moves from US export bans to managing ownership, personnel, and physical infrastructure that ensure critical stages of semiconductor design, fabrication, and testing.
The change in US policy is driven by the fact that China’s strategy to narrow the technological gap has long gone beyond copying or importing equipment and increasingly relies on human capital, primarily through systematic processes of headhunting engineers, developers, and R&D specialists from companies like ASML.
Such specialists transfer not only technical solutions but also complete chains of production knowledge—from design logic and equipment calibration to internal quality control standards—which allows China to replicate critical technological nodes without direct access to Western systems.
The fact that a prototype EUV machine was created in Shenzhen by a Chinese entity with the participation of former ASML specialists indicates that China has reached the stage of systematic replication of critical Western technologies, where US export restrictions serve a time-delaying function but simultaneously push Beijing toward a mobilization model of scientific and technological development.
This logic means a transition from dependence on imports to concentrated internal breakthroughs based on knowledge, personnel, and state management of the technological cycle.
The coordination of this program through state institutions and Huawei points to the formation of a semi-closed strategic cluster, where research, engineering solutions, financing, and political control are consolidated into a single vertical structure, following the principle that defined the American Manhattan Project in the 20th century, within which nuclear weapons were developed.
Its essence lay in the accelerated concentration of resources around a technology capable of radically changing the balance of power, with the complete subordination of science to the state’s strategic goal.
In the 21st century, artificial intelligence and AGI are increasingly taking over the role of nuclear weapons in shaping the global hierarchy, as control over computing power, algorithms, and the production of advanced chips determines military, economic, and political superiority for decades ahead.
That is why China’s bet on EUV, semiconductors, and AI is an attempt to create its own equivalent of strategic deterrence, where technological autonomy becomes a condition of sovereignty.
In this context, US actions are aimed not at stopping China as such, but at maintaining control over the pace and configuration of this race, to prevent a situation in which Beijing gains a full cycle of production and development of AI infrastructure outside Western influence.
Thus, the confrontation in the field of chips and semiconductors takes on the significance of a long-term competition for the architecture of global power, where the stakes far exceed individual markets or technological niches.
In parallel, Beijing employs the tool of manual demand regulation in the chip industry, using permissions and blocks from the Communist Party of China on purchases of NVIDIA chips as a mechanism to influence the global market, balancing between shortages and excessive pressure on manufacturers.
The order for two million H200 chips for 2026 creates a situation in which the American company finds itself between commercial interest and political risk, while the final decision on supplies is made outside market logic.
Thus, China is forming an asymmetric model where demand and market volume are used as leverage capable of influencing pricing, investment cycles, and production priorities of leading Western chip manufacturers.
Another element of China’s strategy in the war against the US chip industry is the closure of supply chains for materials, particularly silicon wafers, to domestic suppliers, which reduces vulnerability to sanctions and increases production autonomy.
The Nexperia China case demonstrates how corporate conflicts and European state interventions (when the Dutch government took control of the local facilities of the Chinese company Nexperia) can accelerate operational separation, essentially contributing to the formation of a parallel Chinese ecosystem of power chips.
The transition to local wafer suppliers for chips means that China is building its own critical infrastructure for electric vehicles and industry, even at the cost of short-term shortages and disruptions.
In this context, Trump’s decision to block the Chinese acquisition of Emcore becomes a response to the totality of these processes, aimed at closing off channels for China’s penetration into Western chip chains through control over strategic assets.
Washington demonstrates a willingness to act preemptively, blocking potential nodes of technological leakage before they turn into structural dependence or a tool of Chinese influence.
This approach indicates the US transition from a policy of containing chip exports to a policy of “containing integration,” where the participation of Chinese capital in critical segments is viewed as a risk regardless of the companies’ formal jurisdiction.
For China, this means accelerating the course toward full semiconductor self-sufficiency, even under conditions of lower efficiency and higher costs in the initial stages.
For the US, Trump’s order fixes the intention to preserve technological asymmetry as an element of global containment, recognizing that completely blocking Chinese progress is already beyond reach.
In summary, the struggle for chips is transforming into a struggle for the architecture of the future economy, where every government decision becomes an element of a grand geopolitical game with a long-term horizon.
Moreover, the Trump administration formalizes its position toward China as a systemic violator of market rules in the semiconductor sector, while deliberately postponing tariff escalation until mid-2027, turning the tariff tool on chips into a deferred method rather than an immediate economic blow.
This pause in tariffs is an attempt to buy time for structural reconfiguration of Western supply chains, as Washington recognizes that a sharp cutoff of Chinese microchip imports would create inflationary shocks for American industry, which is dependent on cheap and mass-produced components.
The published report from the US Trade Representative’s Office fixes the main threat in Beijing’s ability to use control over segments of the value chain as a tool of political and economic pressure, which undermines the predictability of the global chip market.
Accusations of forced technology transfer, systemic theft of intellectual property, and artificial suppression of labor costs indicate that the US views the Chinese model as non-market by nature, and thus incompatible with long-term coexistence in a single technological ecosystem.
At the same time, tying potential tariff increases to the political calendar, including the expected Trump-Xi meeting in spring 2026, indicates that semiconductors are finally integrated into the highest level of negotiations, where trade decisions become derivatives of geopolitical agreements.
Against this backdrop, China’s escalation of restrictions on exports of rare earth minerals demonstrates symmetric logic of action, whereby Beijing signals readiness to respond asymmetrically, striking at bottlenecks in Western technological infrastructure.
It is in this configuration that the emergence of platforms like NVIDIA Vera Rubin takes on significance that goes far beyond commercial breakthrough, as the sharp increase in computational efficiency changes the very mathematics of access to advanced AI models.
Rubin compresses the gap between the number of chips and the volume of achievable computations, allowing the training of trillion-parameter models with exponentially less hardware base, which increases the strategic value of each unit of such equipment for states and corporations.
The multiple-fold increase in training productivity transforms new GPUs from a mass-market commodity into a critical resource, control over which will determine who can deploy physical AI, autonomous systems, and complex simulation environments.
Early interest in this product from AWS, OpenAI, and Anthropic demonstrates that key market players view Rubin as the infrastructural foundation for the next cycle of AI development, which further sharpens China’s interest in access even to limited volumes of such systems.
In this context, the practice of manual approval of chip purchases by Chinese authorities will extend from H200 to new generations of NVIDIA, as demand control becomes a tool for testing the resilience of American export regimes.
Thus, the combination of deferred tariffs, strict investment bans, and technological outpacing through new platforms forms a multi-level containment model for the US, where each element operates in its own time horizon.
China, in turn, responds by accelerating internal self-sufficiency programs and closing supply chains to local manufacturers, which gradually reduces the effectiveness of classic trade levers.
In the medium-term perspective, this means that the global semiconductor market will move in a state of controlled fragmentation, where access to advanced AI chips will be determined not by price or demand, but by political permissions and loyalty.
For Washington, orders like the blocking of the HieFo-Emcore deal become a way to buy the time needed to consolidate technological advantage, while Beijing uses the same time to reduce dependence on Western bottlenecks.
