In 2024, China represented approximately 28% of global capacity measured in 200 mm equivalent wafers according to SEMI. Chinese public funding devoted to the sector is estimated at approximately $150 billion since 2014, according to the CSIS. Maintaining a coalition-based blockade over the long term requires political consistency that spans multiple electoral cycles and resists the economic pressures of European and Asian allies—a constraint that neither Soviet rivalry nor the trade war with Japan imposed at this scale.
The essentials
- According to SEMI’s 300 mm fab scope, China’s share rose from 8% in 2015 to a projected 20% in 2024. Cumulative public funding is estimated at approximately $150 billion since 2014 according to CSIS.
- This catch-up concerns mature nodes (28 nm and above): the most advanced chips (3-5 nm) remain out of reach due to lack of access to ASML equipment and Western EDA tools.
- Mathieu Duchâtel (MERICS) identifies semiconductors as a long-distance industrial marathon, whose outcome depends as much on the cohesion of the Western coalition as on Chinese capabilities.
- Chinese catch-up in advanced nodes would require an extended timeframe under current conditions, assuming restrictions remain in place and an endogenous breakthrough is possible—a duration spanning multiple political cycles in the West.
- Technological bipolarity is underway, but reversible under certain conditions: the open question is one of political durability of the blockade, not its technical feasibility.
The rise in power that deserves to be measured correctly
Tripling a market share in nine years deserves some perspective before drawing conclusions.
In 2015, China held 8% of global 300 mm fab capacity according to SEMI. In 2024, China represented approximately 28% of global capacity measured in 200 mm equivalent wafers according to SEMI. The progression is real. But it concerns a specific segment: mature nodes, meaning chips at 28 nanometers and above. These components power consumer electronics, automobiles, industrial equipment, conventional defense systems.
They are useful, profitable, and strategic at their level. They do not constitute the summit of the chain.
The summit still belongs to TSMC, Samsung, and Intel: chips at 3 to 5 nanometers that neither SMIC nor any other Chinese foundry produces to date. This gap is a technical and industrial frontier that restrictions on ASML’s EUV exposure equipment and on EDA design software keep closed.
Understanding Chinese rise in power thus requires two simultaneous readings: real progress in mature nodes, relative stagnation in advanced nodes. The two coexist, and both matter.
The limits of $200 billion in public investment
State investment is massive by any reasonable comparison. Since 2014, the “Big Fund” has deployed three financing cycles: funds I, II, and III, the latter launched in 2024. Cumulative public funding is estimated at approximately $150 billion since 2014 and covers fab construction, engineer training, and support for domestic demand, without verified breakdown of spending categories.
These resources accelerated and broadened an already existing Chinese industrial base. According to Rhodium, SMIC can produce certain 7 nm SoCs using alternative DUV techniques, subject to access to equipment parts and services. CXMT is ramping up in DRAM memory. The packaging and testing (packaging) supply chain has considerably strengthened.
ASML’s EUV exports to China have been practically blocked since the non-renewed license expired in 2019; Dutch rules were strengthened in September 2023. Certain EDA software and technologies from Synopsys and Cadence are subject to targeted U.S. controls and licensing requirements, notably for certain technologies or Chinese entities. These two bottlenecks can be circumvented through access to foreign technologies or endogenous breakthroughs, and cannot be resolved by funding alone.
Mathieu Duchâtel identifies the semiconductor race as a marathon rather than a sprint, whose outcome depends on multiple political and technological factors.
The blockade as a political challenge as much as a technological one
Such a duration exceeds the natural horizon of any democratic coalition. The 2018 ECRA constitutes a legal framework; U.S.-specific controls on advanced Chinese semiconductors were substantially expanded in October 2022, with successive updates. The Netherlands instituted its own strengthened controls on certain advanced equipment in September 2023, in a context of coordination with allies. Japan joined the coalition on etching equipment. This trilateral coordination represents a significant convergence in technological controls.
But maintaining this convergence across five to six American electoral cycles, two to three Dutch governments, and several Japanese alternations represents a constraint of a different nature. Soviet rivalry had CoCom, but concerned technologies less integrated into the global economy. The pressure on Japan in the 1980s involved an ally under a security treaty, not a systemic competitor that simultaneously represents the European Union’s leading commercial partner.
This point deserves to be set against a competing interpretation. Economist Thomas Philippon, in his work on market concentration and structures of economic power, recalls that coalitions of control tend to fracture where commercial incentives are strongest. European semiconductor companies, ASML foremost, but also German equipment makers and Japanese chemists, face real pressure from their shareholders and industrial clients who would like to regain access to the Chinese market. Temporary exemptions and licensing regimes have existed to limit supply chain disruptions; their exact causality cannot be reduced to commercial pressure alone.
The scenario of a blockade maintained for thirty years assumes a political coherence that democracies have rarely demonstrated on economically costly issues for their own actors. The United States has maintained the Cuban embargo since 1962, which shows that such maintenance remains possible; this structuring variable, however, escapes purely technological analysis.
To this is added a dynamic already observed in other geographies: countries that do not design their technologies seek to host or finance them, hoping to shorten their dependence. China is doing more than that; it is attempting to rebuild the entire chain, but the timeline remains governed by physical and organizational realities that cannot be bought.
China’s rise in power in mature nodes
While debate concentrates on the gap in advanced nodes, China is strengthening its capacity in mature nodes. And this has concrete consequences for global industry.
Chips at 28 nanometers and above represent a significant share of the global semiconductor market by value and equip critical applications. They power electric vehicles, industrial systems, telecommunications equipment, conventional weapons. Chinese capacity in certain mature segments creates sectoral dependencies that must be evaluated by product, supplier, and supply chain.
It is a coherent strategy. China does not need to manufacture 3 nanometer chips to equip its military drones or electric vehicles. It needs to secure its supply in categories where demand is broadest and substitutability is lowest. The rise in power in mature nodes fulfills that objective.
Direct corollary: Taiwanese, Korean, and European foundries face increased pressure on these market segments. TSMC remains present in mature nodes with a strategy centered on specialized technologies, while Chinese competition there is strengthened by public support. This logic of supply chain capture through subsidization is well documented in other Asian industries, and it repeats itself here with higher strategic stakes.
Two generations to settle: long-term scenarios
Duchâtel identifies the question without settling it, and available data do not allow it to be done with certainty: the trajectory that technological bipolarity will follow by 2050 remains undetermined.
Two scenarios structure the debate, though neither is predictable with certainty.
In the first scenario, the restrictive framework holds. The Netherlands would maintain ASML restrictions despite commercial pressures. Japan and South Korea remain in the coalition. China progresses in mature nodes, but the gap in advanced nodes narrows over the long term. Two technological trajectories emerge: one centered on advanced Western chips, the other on China’s rise in mature nodes.
Third countries—India, ASEAN, Africa—choose their infrastructure standards according to their political alliances. This bifurcation would affect the structure of global component trade and have consequences for non-aligned players.
In the second scenario, the restrictive framework could loosen under commercial pressure. European pressures could result in expanded exception regimes. China would access intermediate-generation equipment.
In this scenario, the question is no longer one of blockade but of negotiation: which technologies remain to be shared, and on what terms.
The signals to watch are precise. China’s share in the 14 nanometer and above nodes will give a quarterly reading of progress in mature segments. Licenses granted or refused by The Hague for intermediate ASML equipment will indicate the robustness of the Dutch coalition. TSMC’s investments outside Taiwan, in Arizona, Germany, and Japan, will signal how fast the critical chain is decentralized and insulated from Taiwan contingency.
A third factor, less discussed, could prove decisive: the speed at which AI itself accelerates the design of new nodes. If automated design tools allow compression of development cycles, the twenty to thirty year delay estimated today could prove too long, or too short depending on who has these tools first. This is a variable that current projections struggle to model.
The cost of a fragmented global economy
A dimension often underestimated in this analysis: what bipolarity costs to economies that are neither Chinese nor American.
Europe, particularly, finds itself in an uncomfortable position. It hosts ASML, the first link in the global equipment chain, and its automotive sector depends both on American-Taiwanese advanced chips and the Chinese market. The Chips Act provides for over 43 billion euros in public investment; total public and private investment oriented by policy exceeds 100 billion euros and aims to bring Europe’s global semiconductor market share to 20% by 2030. Geopolitical fragmentation makes this goal simultaneously more urgent and more difficult.
Emerging economies bear the consequences without weighing in on decisions. Their industries digitize with chips whose origin conditions their suppliers’ diplomatic relations. As AI deployed in Gulf countries reveals dependence on infrastructure designed elsewhere, fragmentation of semiconductor chains creates forms of technological dependence that third countries neither have the means to resolve nor the weight to influence.
The cost is thus borne not only by China, which invests massively to reduce its vulnerability, nor by the United States, which finances the CHIPS Act to strengthen its own industrial base. It is also borne by allies applying restrictions on their own exporters, and by emerging markets caught between two incompatible technological systems.
The situation twenty years from now remains uncertain. Decisions made in the next three years at ASML, in Washington, and in Beijing will probably weigh more in the balance than all the five-year plans already announced.
Sources
- Mathieu Duchâtel, “Semi-conducteurs : le marathon chinois,” Revue Défense Nationale, 2026, https://shs.cairn.info/revue-defense-nationale-2026-7-page-46?lang=fr
- CSIS China Power, China’s Progress in Semiconductor Self-Sufficiency (csis.org/programs/china-power-project)
- Rhodium Group, China’s Chip Industry Under Export Controls, reports 2023-2024
- European Commission, European Regulation on Semiconductors (European Chips Act), 2023
- Synopsys / Cadence, EDA restrictions, Bureau of Industry and Security, U.S. Department of Commerce (October 2022 rules and 2023 updates)
