Taiwan Chip Odyssey : How a Small Island Became the Heart of the Global Semiconductor Industry
I recently attended a screening of the documentary A Chip Odyssey, followed by a roundtable on cooperation between Grenoble and Taiwan in the field of semiconductors. One question stayed with me throughout the evening : how did an island with a limited domestic market become an essential hub of the global digital industry?
It all began in 1974, in a small breakfast shop in Taipei, where a handful of Taiwanese decision-makers committed to building a semiconductor industry on the island and the documentary opens with the director’s quest to locate the site of that formative meeting.
In the 1970s, Taiwan was still an emerging economy, rooted in agriculture and textiles. The government chose to send a first generation of young engineers to train in the United States, as part of a technology transfer program with Regional Cooperative Agreement. This is the story told by Hsiao Chu-Chen’s film, a five-year project built on the testimonies of more than 80 people who helped shape this history, from pioneering engineers to the operators of the first production lines.
The second key figure in this story has a name : Morris Chang. Trained in the United States, he spent twenty-five years at Texas Instruments, rising to become group vice president in charge of all semiconductor activities.
In 1985, the Taiwanese government recruited him to head ITRI, its industrial research institute, with a mission to strengthen Taiwan’s semiconductor industry. Chang’s assessment was clear-eyed : at the time, the island had relative weaknesses in designing and marketing its own chips compared with Intel or the Japanese. He therefore departed from the integrated model, in which companies design and manufacture their own chips, and proposed a factory that would manufacture exclusively for others. In 1987, he founded TSMC in Hsinchu, backed financially by the Taiwanese state (48.3% of the capital) and Philips (27.6%), which brought its manufacturing processes.
This is the pure-play foundry model: TSMC sells no chips that compete with those of its customers. This golden rule would trigger the fabless revolution: companies could specialize in design and outsource manufacturing. Among the companies entrusting TSMC with the fabrication of their chips are Apple, Nvidia, and AMD. The model also offered a way beyond competing on low manufacturing costs: value shifted toward expertise in increasingly advanced manufacturing processes, down to the nanometer.
Nearly forty years later, Taiwan occupies a central place in the global fabrication of advanced chips, to the point that its industry has become a focal point of technological competition among major powers. TSMC has become its emblem, without being the whole story: around the company, an entire ecosystem of public research, universities, suppliers, and skills has been built, making the island far more than a simple production site.
This extreme concentration has turned the island into a major geopolitical focal point. In Chip War, published in 2022, historian Chris Miller shows how semiconductors became a central issue of economic and military power, notably in the rivalry between the United States and China.
TSMC is now developing capacity outside Taiwan: its first fabs in Arizona and Kumamoto, Japan, entered production at the end of 2024, while the Dresden project in Germany was launched the same year. In the United States, this expansion also reflects political pressure to bring chip manufacturing back onshore: the Trump administration pushed TSMC to accelerate the timeline of its American fabs. But this interdependence also underpins the idea of a “silicon shield”: the strategic importance of Taiwan’s semiconductor industry may help protect the island, an idea increasingly debated as geopolitical tensions evolve.
What the Grenoble-Taiwan Roundtable Taught Me
The evening continued with a roundtable on technological cooperation between Grenoble and Taiwan, and several contributions stood out to me.
The discussions first addressed leadership in the most advanced process nodes, where TSMC maintains a clear lead. They also covered silicon-based quantum computing, through the case of Quobly, a startup spun off from CEA-Leti. One contribution made me think that technological sovereignty, like landing on the Moon, is a goal that mobilizes resources over decades.
Finally, there is the question of talent. Taiwan’s semiconductor industry faces a shortage of engineers against a backdrop of low birth rates. Academic partnerships like joint PhD supervision, internships, and exchanges between CEA-Leti, UGA, and Taiwanese universities are perhaps the most tangible way to strengthen these ties. One central question emerged from these discussions: our societies depend on digital technology, and the real strategic question is now that of our technological dependencies whom do we depend on, and how do we build our resilience?
Taiwan’s journey is more than an industrial success story, it reflects a society-wide choice to invest in knowledge rather than rely on resources. For Europe, and for an ecosystem like Grenoble’s, the lesson is clear: sovereignty cannot simply be declared. It is built through expertise, cooperation, and patience.
