A modern chip may cross a dozen borders before it reaches a phone, a car, or a missile. Design in America, lithography in the Netherlands, fabrication in Taiwan and Korea, packaging in Malaysia. The chain is a miracle of specialization, and a single point of failure in a dozen places.
Published 14 October 2026 · 06:00 GMT
There is no modern machine without a chip, and there is no chip made by one country. The device in your pocket contains silicon designed in one place, printed by a machine built in another, fabricated in a cleanroom in a third, and packaged in a fourth. The supply chain is the most complex manufacturing system humans have ever built, and it works so well that nobody thought about it until it stopped working.
That changed in 2021, when a shortage of chips shut down car plants, delayed medical devices, and forced companies to ration products they had previously taken for granted. The shortage was not caused by one event. Pandemic demand swung from cars to laptops and back, factories ran at capacity, and lead times stretched from weeks to months. The lesson was blunt: a chain this specialised is also this fragile, and the fragility is now a matter of national security.
A chip begins as an idea: a design, drawn in software, by engineers who may never touch silicon. The leading design firms are concentrated in the United States, with important players in Europe and Asia. The design is then printed onto silicon wafers in a fabrication plant, or fab, using lithography machines that project patterns with light. The finest patterns require extreme ultraviolet lithography, and the only company that builds those machines at scale is ASML in the Netherlands. Each machine costs more than a passenger jet, contains hundreds of thousands of parts, and requires a global supply chain of its own.
Fabrication is where the concentration is sharpest. The most advanced logic chips, the processors in phones and AI accelerators, are made by a handful of companies, above all TSMC in Taiwan and Samsung in South Korea. These fabs run twenty-four hours a day, in rooms cleaner than operating theatres, with air filtered to remove particles invisible to the eye. A single earthquake, a drought that cuts the ultrapure water supply, or a power failure can idle a plant whose output the whole world is waiting for. The industry has managed these risks for decades. Governments have only recently started to price them.
After fabrication comes packaging and testing, often in Malaysia, Vietnam, the Philippines, or China, where chips are cut, encased, and checked. Packaging was once treated as the low-value tail of the chain. Advanced packaging, which stacks and connects chiplets into a single system, is now a frontier of its own, and it is where some of the next performance gains will come from. The chain ends in assembly: phones, cars, servers, weapons. At every step, the value is in the knowledge, and the knowledge is in very few hands.
Governments have responded to the fragility with money. The United States passed the CHIPS and Science Act in 2022, committing tens of billions of dollars to domestic fabrication, research, and workforce development. The European Union adopted its own Chips Act with similar aims. Japan, South Korea, India, and China have all announced major support for domestic capacity. The result is a global subsidy competition, with each capital promising that its fabs will make it safer.
The economics of the subsidy war are uncomfortable. A leading-edge fab costs tens of billions of dollars to build and billions more to run, and it needs a workforce of thousands of specialised engineers. Subsidies can move a fab. They cannot move the ecosystem: the suppliers, the chemical plants, the maintenance crews, the universities that train the engineers. Taiwan and Korea built those ecosystems over decades. A government that funds a fab without funding the ecosystem gets a building. A government that funds the ecosystem gets a chance.
There is also a question of what self-sufficiency means. No country can build the whole chain alone, and the attempt would be ruinously expensive. The realistic goal, for most capitals, is resilience: enough domestic capacity in the most critical nodes, enough allied capacity elsewhere, and enough inventory to survive a disruption. The difference between self-sufficiency and resilience is the difference between a slogan and a strategy, and the debate in every capital is which one its government is actually pursuing.
The sharpest edge of chip policy is export control. The United States has restricted exports of advanced chips, chipmaking equipment, and related technology to China, arguing that the most capable chips have military applications and that China must not gain access to them. The Netherlands and Japan have aligned parts of their own controls with Washington. China has responded with its own restrictions on critical minerals and with accelerated investment in domestic chipmaking.
The controls have worked in the narrow sense: Chinese firms face real difficulty obtaining the most advanced lithography machines and the most capable AI accelerators. They have also produced second-order effects. Chinese firms have invested heavily in mature-node chips, the older but still essential chips used in cars, appliances, and industrial equipment, where they are becoming formidable competitors. Allied companies have lost revenue in the Chinese market. And the controls have given Beijing a powerful argument for self-reliance, the same argument Washington used to justify its own subsidies.
The deeper question is whether the chip world splits into two ecosystems, one American-led and one Chinese-led, or whether the commercial logic of a single global chain reasserts itself. The industry, which profits from selling to everyone, prefers one ecosystem. The security establishments, which fear dependence, prefer two. The companies are caught in between, building fabs in Arizona and Arizona-style fabs elsewhere, while trying to keep selling to the market that buys the most chips. The tension is the defining fact of the industry.
Western coverage reads chips as the central technology of the century. In the American telling, the CHIPS Act is a correction: the country that invented the industry allowed its manufacturing to concentrate abroad, and it is now buying its way back. In the European telling, the Chips Act is about a share of a market Europe helped create and then neglected. Both tellings share an assumption: the chain as it stands is too concentrated to be safe, and the concentration must be diluted.
The Western lens is also the most attentive to the Taiwan question, and the most likely to frame it in security terms. A disruption in the Taiwan Strait, whether from blockade, conflict, or natural disaster, would be an economic shock without modern precedent. The Western answer is diversification: fabs in the United States, Europe, Japan, and India, and stockpiles to bridge the gap. The question the Western lens asks least often is whether diversification at this scale can ever be fast enough.
Eastern coverage reads the same chain as a story of dependence and its reversal. From Beijing, the export controls are proof that the old division of labour, in which China assembled and others supplied the critical parts, was a vulnerability by design. The response is a national drive for self-sufficiency, not as ideology but as insurance. In this telling, the controls accelerated what China was already trying to do, and the subsidies in Washington and Brussels are the same industrial policy, practised with different rhetoric.
The Eastern lens also stresses that China is not only a consumer of chips. It is the largest buyer, the largest assembler, and increasingly a producer of mature-node chips at scale. The reading from Taipei and Seoul is more complicated: both depend on the Chinese market and on American security guarantees, and both have built their prosperity on a chain that the great-power contest now threatens. The Eastern lens sees the chain as a hostage of politics, and politics as the variable that will decide it.
Global South coverage reads chips as a development story. The countries that package and test chips, Malaysia, Vietnam, the Philippines, and others, have built real industrial capability, but they capture a small share of the value. The Southern lens asks whether the subsidy war will lift them up the chain, into design and fabrication, or push them down it, as the rich countries re-shore the high-value steps and leave the rest.
The Southern lens also notes that the chip shortage hit the South hardest. Car plants in Africa and Latin America waited longer for chips than plants in Germany or Japan, because allocation followed purchasing power. In this telling, the lesson of 2021 was not only that the chain is fragile. It was that fragility is distributed by power, and that the countries with the least say in the chain bear the most risk when it breaks. Resilience, for the South, means a seat at the table, not just a shipment of chips.
