Taiwan is a small island. That sentence is so often used as a preface to a threat, or to a miracle, that it has almost stopped meaning anything. It is still the right place to start. Twenty-three million people, no oil, mountains that leave only a strip of plain along the west, a political life that has never been allowed to be boring. Out of that, in two generations, came a way of making things that the rest of the world now depends on, and a list of people whose names sit on the machines that run the present tense: Morris Chang, Terry Gou, Jensen Huang, Lisa Su, and a longer bench behind them.
They are not a coincidence, and they are not proof that Taiwanese people are magically better at maths. They are what happens when a frightened, well-schooled society decides that the only safe ground it has is human skill, then builds the institutions to match. This is a human story. The companies are the visible part.
An island that had to become a brain
After 1949 the Republic of China government on Taiwan faced a plain problem. It had lost a continent. It had a lot of people, including mainland professionals and soldiers, stacked onto an island that already had its own society, language, and memory of Japanese colonial rule. Land reform in the 1950s broke up large holdings and created a countryside of small farmers. American aid helped. Nine-year compulsory education, rolled out in 1968, did something more lasting than either: it made a generation that could be trained.
The economic planners of that era, especially K. T. Li (Li Kwoh-ting) and Premier Sun Yun-suan, are less famous than the men they later recruited, which is a kind of justice. They were civil servants. They looked at what a small place could not do — heavy steel on Japanese or Korean terms, oil, a captive continental market — and at what it might do if it stayed close to American technology and obsessed over manufacturing yield. The Industrial Technology Research Institute, ITRI, was set up in 1973 in Hsinchu. The Hsinchu Science Park opened in 1980. A technology transfer from RCA, through ITRI’s electronics lab, became United Microelectronics Corporation in 1980. None of this was romantic. It was a bet that engineers could be a national resource.
Hsinchu is still the physical fact of that bet. The park sits where the hills come down toward the coast. You can see university campuses, fabrication plants, and scooter traffic in one glance. The air is humid. The mountains are close. It does not look like Silicon Valley, and it was not trying to. It was trying to give Taiwan a place where a graduate could go to work without leaving the island, and where a government that could not print a reserve currency could still print competence.
Morris Chang, who was asked to come
Morris Chang was not born in Taiwan. He was born in Ningbo in 1931, got out through Hong Kong, studied at Harvard and then MIT, took a Stanford doctorate, and spent the meat of his career at Texas Instruments. He was in his fifties, passed over for the top job, and briefly at General Instrument when Taiwan came calling. Li Kwoh-ting wanted him to run ITRI. Chang has said there would be no TSMC without Li. That is not false modesty. The island imported a man who already knew how American chip companies thought, and then it gave him a state.
In 1987 Chang founded Taiwan Semiconductor Manufacturing Company as a foundry: it would make other people’s chips and not compete with its customers by designing its own. Intel was not interested in the idea. Philips, the Dutch electronics firm, put in money and process technology for a stake of about 28 percent. The Taiwan government, through the National Development Fund, put in about 48 percent. The first plant was on the ITRI campus. The model looked, to people used to Intel or Texas Instruments, like a factory without a soul. It was the opposite. It was a factory that refused to pick a favourite child among the world’s designers.
That refusal is the hinge of modern electronics. Once a company could send a design to Hsinchu and get back silicon, it no longer needed to own a billion-dollar plant in order to exist. Nvidia could be a design firm in San Jose. MediaTek could be a design firm in Hsinchu. Apple could be a design firm in Cupertino. TSMC would take the yield risk. Chang ran the company as chief executive into the 2000s and as chairman until 2018. TrendForce, in September 2026, put TSMC’s share of the global foundry market at 72.5 percent in the second quarter. Samsung was a distant second. That number is not a personality cult. It is what a 1987 organisational choice looks like after forty years of not blinking.
Chang is the adoptive father in this family. Taiwan did not give birth to him. It recognised him.
Terry Gou, who stayed and assembled the world
Terry Gou was born in 1950 in Banqiao, then Taipei County, to parents who had come from Shanxi. In 1974, with about NT$200,000, he started Hon Hai Precision Industry making plastic parts — the humble connectors and knobs that sit at the edge of a television or a computer. The international name, Foxconn, came later. The company learned to take a design, a bill of materials, and a launch date, and turn them into millions of finished devices. That sounds like a simple sentence. It is not a simple skill.
Gou’s great geographical decision was to put the labour in China and keep the brains of the operation — the process engineering, the customer relationship, the ability to stand up a new line in months — in a Taiwanese firm. Zhengzhou, the iPhone city, is a Chinese demographic fact. The company that knew how to run it was Gou’s. This site has already written about that bargain at length: American brands kept the rents, Taiwan invented the intermediary, China supplied the hands. The point here is narrower. Gou is what the island’s manufacturing nerve looks like when it is allowed to become a civilisation of connectors, jigs, night shifts, and customer panic. He later went into politics. The company remains larger than the candidate.
He is not alone in that layer. Barry Lam’s Quanta became a giant of notebook computers made for other people’s logos. Compal, Wistron, Pegatron — the alphabet of electronics manufacturing services — are Taiwanese inventions of the same family. Stan Shih founded Multitech in 1976, later renamed Acer, and drew his “smiling curve”: the money, he said, sits at the two ends, brand and components, while the middle, pure assembly, is the thin part of the grin. Taiwan spent decades in the middle and then climbed both slopes. Shih’s curve is still the honest diagram of why Gou and Chang are different children of the same house. One mastered the middle. One made the right-hand end of the smile, the component that everyone else had to buy.
Jensen Huang, who left as a child and took the island with him
Jensen Huang (Huang Jen-hsun) was born in Taipei on 17 February 1963 and grew up in Tainan, the old southern city. His father was a chemical engineer at a refinery. His mother taught school. The family spoke Taiwanese at home. When he was five they moved to Thailand for his father’s work. When he was nine, with Southeast Asia unstable, the boys were sent to an uncle in Tacoma, Washington. The uncle, meaning well and knowing little, enrolled them at Oneida Baptist Institute in rural Kentucky, which was not the elite boarding school he thought it was. It was a religious school for difficult children. Huang has said he cleaned bathrooms. His brother worked tobacco. The parents arrived, saw the place, and pulled them out. The family settled near Portland. Huang went to Aloha High School, Oregon State, then Stanford.
In 1993, in a Denny’s in San Jose, he and two colleagues founded Nvidia. The company made graphics chips for games, then for parallel computing, then for the training of large models. TSMC made the silicon. Huang has been unusually open about that dependence. Morris Chang has a cartoon of the two of them on his office wall. The foundry Chang built is what made a design shop in California able to become, for a time, the most valuable company in the world without owning the plant that prints its ideas.
Huang is a Taiwanese-American, not a Hsinchu plant manager. That is the point of including him among Taiwan’s children. The island’s brilliance is not only the people who stayed. It is also the people who left with a language, a way of sitting an exam, and a later ability to plug back into Hsinchu when they needed a wafer. In 2024 ITRI named him a laureate, alongside Lisa Su. He has put an Nvidia research centre in Taiwan. The boy who left Tainan did not, in the end, leave the industrial family.
Lisa Su, and the others who do not fit on a magazine cover
Lisa Su was born in Tainan in 1969 and was in New York by the age of three. She took three degrees at MIT, worked at Texas Instruments and IBM, and in 2014 took over Advanced Micro Devices when the company was a punchline. She is, depending on how you count the family tree, Huang’s first cousin once removed. She has called the link distant. Genealogists have made it a parlour game. The interesting fact is not the blood. It is that two children born in the same southern Taiwanese city, six years apart, both raised in the United States, now run the two firms whose processors sit at the centre of the artificial-intelligence boom, and both send their hardest work to the same foundry Chang built.
If you stop at cousins you miss the bench. Tsai Ming-kai’s MediaTek, spun out of UMC, is one of the world’s large fabless chip designers and is headquartered in Hsinchu. C. C. Wei now runs TSMC. Doris Hsu’s GlobalWafers sits in the less glamorous but essential business of the silicon disc itself. Jerry Yang (Yahoo) and Steve Chen (YouTube) were born in Taiwan and did their famous work in American software. Min Kao co-founded Garmin. King Liu built Giant into a bicycle company that treated a humble object with the same process seriousness the chip people brought to wafers. The list is not a hall of fame. It is a pattern: a small place that exported engineers the way other places exported grain.
What the brilliance actually is
It is easy, and wrong, to turn this into a story about lonely genius. Chang was a professional in late middle age with a government behind him. Gou was a connector salesman who learned to run cities of workers. Huang was a child of the Taiwanese middle class who got a second education in Oregon and a third in a bad Kentucky mistake. Su was a MIT doctor of electrical engineering who walked into a wounded company. The common pieces are more ordinary than the magazine profiles suggest.
Schools that took exams seriously. Families that treated engineering as a respectable life, not a fallback. A state that built ITRI and a science park instead of pretending the market would invent a semiconductor industry on a rice island. A foundry model that let the restless children design in California while the careful children made the wafers at home. A habit, learned under Japanese rule and under the KMT and then under democracy, of being the smaller party in the room and still finishing the job.
There is also a darkness that belongs in the same paragraph. Foxconn’s China plants were not a pastoral. TSMC’s concentration is a strategic vulnerability as well as a shield; people now talk about a “silicon shield” as if a factory could be a treaty. Young Taiwanese still leave. Fertility is among the lowest in the world. The property of being indispensable is not the same as the property of being safe. None of that cancels the achievement. It is the weather the achievement has to live in.
The people named above did not invent Taiwan. Taiwan, in the form it took after the 1950s, invented a way of producing people who could sit with a hard problem until it yielded. Some of them assembled phones. Some of them refused to design chips so that everyone else could. Some of them left as children and came back as customers. The island is still small. The work is not.