InProfile:HowMorrisChangBuiltTSMCIntoaChipmakingColossus

On Jan. 29, Nvidia Corp. CEO Jensen Huang landed in Taiwan and headed straight to the residence of Taiwan Semiconductor Manufacturing Co. Ltd. (TSMC) founder Morris Chang in Taipei’s Dazhi area. Later that evening, the two were spotted dining at a well-known local restaurant. In photos circulating online, Chang, seated in a wheelchair, appeared with a head of white hair and his characteristic high cheekbones, looking spirited despite his advanced age.

This was not merely a social call, but was a summit between the two most powerful figures in the global semiconductor industry. Huang’s Nvidia has become the computational bedrock of the new generation of artificial intelligence (AI), driving an ecosystem that has seen the company’s market capitalization soar to touch $5 trillion. Chang, meanwhile, is the architect of TSMC, the world’s largest contract chipmaker. His company was the first to mass-produce 2-nanometer semiconductors, maintaining a stranglehold on the world’s most advanced chip manufacturing technology.

The relationship between the two spans more than two decades. In 1997, when Nvidia was teetering on the brink of bankruptcy, Huang received a pivotal telephone call from Chang. That conversation initiated an exclusive manufacturing partnership that saved the American chip designer. Huang has frequently expressed his gratitude and admiration for the elder statesman of the industry, often saying bluntly that “without TSMC, there is no Nvidia today.”

Chang, who was born in 1931 and will turn 95 this year, has lived a life that mirrors the tumultuous history of modern industry and geopolitics. Born in Ningbo, Zhejiang province, he spent his youth as a refugee of war, moving with his family between Ningbo, Nanjing, Guangzhou, Chongqing, Shanghai and Hong Kong before the age of 18. He eventually traveled to the U.S. for his education, became an American citizen and worked in the American tech sector for 30 years. In 1985, at the age of 54, Chang accepted an invitation to move to Taiwan. There, he spent two years analyzing global industry trends and local labor resources before founding TSMC. Over the subsequent three decades, he guided the company to an irreplaceable position at the pinnacle of the global economy.

From an art aficionado to an engineer

According to his autobiography, following his father’s job transfer, Chang’s family moved across China, eventually crossing the front lines of the War of Resistance Against Japanese Aggression to reach Chongqing. There, Chang attended Nankai Middle School, an institution he credited with sparking his dreams of becoming a writer and his love for classical music.

Following the end of the war in 1945, the family returned to Shanghai. While Chang harbored literary ambitions — admiring works by Lu Xun, Hu Shih and Xu Zhimo — pragmatic concerns about survival drove him toward commercial studies. In 1948, amidst the Chinese Civil War, Chang, then a student at the University of Shanghai’s banking department, left for the U.S. for education.

In 1949, at 18, Chang entered Harvard University as the only Chinese student among more than 1,100 freshmen. Although his major was engineering, he immersed himself in Western culture, studying literature, listening to opera and visiting museums.

However, recognizing the better employment prospects in engineering, Chang transferred to the Massachusetts Institute of Technology (MIT) in 1950. He earned his bachelor’s degree in mechanical engineering three years later and his master’s degree a year after that. After failing his doctoral qualifying exams twice, he entered the workforce, landing at a semiconductor company.

Chasing the American dream

Chang’s professional career began in 1955 at Sylvania Electric Products in Boston. The semiconductor industry was in its infancy, with only two or three dozen companies operating in the U.S. Chang acquired knowledge about semiconductors by reading Electrons and Holes in Semiconductors by Nobel laureate William Shockley and by consulting local chip experts. In 1956, he led a team that developed various germanium transistors and presented his first academic paper in December of that year. Despite these technical successes, Sylvania struggled to coordinate its research, production and sales, leading to persistent losses in its semiconductor division. Three years later, Chang joined Texas Instruments Inc. (TI), a rising star in the sector.

At TI, Chang established his reputation by continuously improving manufacturing quality for transistors produced for IBM Corp. Recognizing his potential, the company sent him to Stanford University in 1961 to pursue a doctorate. Upon his return, he was appointed general manager of TI’s germanium transistor department. He climbed up the corporate ladder steadily, rising to become a departmental vice president and general manager of the integrated circuit department in 1967. During that period, he evolved from an engineer into a comprehensive manager with expertise in marketing, sales, finance and strategy.

In 1972, Chang became a group vice president responsible for TI’s global semiconductor business, effectively the number four executive in the company behind the chairman, president and executive vice president. He achieved his American dream and set his sights on the presidency of the company.

However, the company’s strategic shift brought challenges. That same year, TI decided to enter the consumer products market, manufacturing calculators, electronic watches and home computers. Chang recalled that he was required to fully support the consumer business, which placed immense pressure on the semiconductor division. After five years of difficult conditions, he resigned his position in 1978. He was subsequently appointed general manager of the consumer products group, a role that was followed by a demotion to head of quality and productivity. Chang left TI in 1983. Following a brief one-year stint at General Instrument Inc., he decided to accept an invitation from Taiwan’s Industrial Technology Research Institute (ITRI) to continue his career in Asia.

The struggle to build TSMC

In August 1985, Chang officially became the president of ITRI, a government-sponsored research organization established in 1973 to drive industrial development. ITRI had been researching semiconductors since 1975 and built Taiwan’s first 4-inch wafer factory in 1977, mastering 7-micron CMOS process technology.

Upon taking office, Chang faced pressure from Li Kwoh-ting, a senior government official known as the godfather of Taiwan’s economic miracle, to solve a specific problem: three local chip design companies were demanding that the government fund the construction of wafer fabrication plants. Chang analyzed the situation and proposed the concept of a dedicated foundry — a company that manufactures chips for others without designing its own.

Chang reasoned that ITRI’s chip design capabilities were weak and its manufacturing technology lagged the world’s most advanced by about three years. It also lacked sales capabilities. However, its one strength was the excellent yield rate of its production line. Therefore, the viable path was to manufacture wafers using mature processes for clients while continuously researching new technologies.

The model faced immediate internal resistance from the government, but with the backing of Li and then-Premier Yu Kuo-hwa, Chang proceeded to form a contract chipmaking firm, namely TSMC, jointly owned by ITRI and a multinational firm. Finding investors, however, proved difficult. In 1985, the global semiconductor industry was in a downturn. Major players including Intel Corp., Texas Instruments and Motorola all rejected Chang’s proposal, showing no interest in investing in a pure-play foundry that only makes chips for other firms. Only the Dutch electronics giant Philips expressed willingness to invest.

TSMC completed its fundraising in August 1986 with a total capital of $145 million. The Taiwan government’s development fund offered 48.3% of the capital, Philips 27.6% and the remainder was provided by firms including Formosa Plastics. TSMC was officially registered in 1987 with Chang as chairman. To build the team, Chang persuaded James Dykes, a former General Electric Co. executive, to serve as general manager and recruited the core team from ITRI’s wafer factory.

Path to global leadership

In TSMC’s first year, Chang established a strategy to penetrate European and American markets, consolidate yield rate and win client trust. Initially, customers were scarce. The monthly capacity of its first 6-inch fab was only in the thousands, and utilization was low. The turning point came in 1988 when Intel, needing capacity for its advanced products, outsourced the manufacturing of 1.5-micron microcontrollers to TSMC. This certification from a world-class client validated TSMC’s capabilities.

At Intel’s insistence, TSMC implemented a comprehensive world-class production system, including statistical process control. Subsequently, Intel outsourced 1.2-micron DRAM production to the company. With Intel on board, other giants such as Philips, Texas Instruments and Motorola followed.

In 1990, TSMC’s revenue hit NT$2.2 billion ($70 million) and in 1995, it reached $1.1 billion. Under Chang’s leadership, TSMC listed on the Taiwan Stock Exchange in 1994 and the New York Stock Exchange in 1997.

Even after listing, the industry was still dominated by Integrated Device Manufacturers (IDMs) like Intel and Samsung, which designed and manufactured their own chips. Top talent preferred those companies. To change this dynamic, Chang set a “technology leadership” strategy in 1997, significantly boosting R&D spending. TSMC notably rejected an invitation from IBM to co-develop 0.13-micron process technology, choosing instead to develop it independently.

Between 2001 and 2009, TSMC’s annual capital expenditure hovered between $2 billion and $2.5 billion. Seeing an opportunity in 40-nanometer and 28-nanometer technologies, Chang raised capital expenditure to $5.9 billion in 2010. The 28-nanometer technology became TSMC’s entry point into the smartphone market and remained a major revenue driver for years.

The ascent was not without setbacks. TSMC encountered quality issues with its 40-nanometer process, causing significant losses for Nvidia, its key partner. Chang ultimately resolved the matter and retained Nvidia’s loyalty by paying hundreds of millions of dollars in compensation.

A significant gamble occurred during its development of 28-nanometer technology. Apple Inc. approached TSMC requesting 20-nanometer technology. At the time, TSMC was preparing to develop 16-nanometer tech, rather than 20-nanometer tech. To win Apple’s order, Chang decided to prioritize the development of the 20-nanometer process.

Samsung developed the 16-nanometer technology ahead of TSMC. In the second half of 2013, Apple adjusted its roadmap and chose Samsung as the contract maker of its mobile processors. Nevertheless, Chang’s investment in the 20-nanometer technology won Apple’s trust.

Chang’s relentless investment in process technology eventually allowed TSMC to surpass Intel at the 7-nanometer node. Since then, the company has successfully advanced to 5-nanometer, 3-nanometer and 2-nanometer processes. Its latest roadmap targets trial production of 1.4-nanometer chips by the end of 2027, leaving rivals like Intel and Samsung trailing further behind.

According to TrendForce, a market intelligence provider, TSMC controlled 70.2% of the global foundry market in the second quarter of 2025. With its dominance in both scale and technology, the company has become a focal point of geopolitical maneuvering, expanding its manufacturing footprint beyond Taiwan to the Chinese mainland, the U.S., Japan and Germany.

Contact editor Ding Yi (yiding@caixin.com)