How Was the Spark Ignited? Seamless Progression Across Every Stage

Taiwan’s semiconductor industry is now world-renowned, but its success did not come easily.

From scratch, the industry was built through bold national industrial policy: negotiating directly with U.S. firms to obtain formal technical licenses, sending personnel to the U.S. for training, bringing technology back to Taiwan for pilot production at ITRI, and then transferring it to private enterprises—ultimately establishing an entire new industrial chain.

Today, Taiwan’s semiconductor sector stands like the Central Mountain Range, quietly protecting the island while shining on the global stage.

The key to this success lies in how Taiwan laid deep roots from the very beginning, forming a foundation that would one day support a world-shaking industry.

At every critical juncture, the right decisions were made by key individuals whose stories deserve close attention.

Why did Taiwan’s semiconductor industry succeed so remarkably?

I believe the answer lies in three distinct phases. In each phase, Taiwan made the right choices and ensured smooth transitions, enabling the industry to develop continuously, one baton passed to the next.

Phase One: RCA Technology Transfer

The first stage began with a technology transfer from RCA—Taiwan’s leap from zero to one in semiconductor know-how.

In March 1976, under the coordination of overseas scholar Pan Wen-Yuan（潘文淵）, a team of experts selected RCA (Radio Corporation of America) as Taiwan’s technology transfer partner.

RCA agreed to a 10-year technology licensing agreement with ITRI, at a cost of $2.5 million for the transfer and $1 million in licensing fees.

Beyond transferring technology, RCA also committed to training Taiwan’s personnel.

At this stage, the Taiwanese government showed foresight and trust, allowing ITRI’s Technical Advisory Committee (TAC) to conduct evaluations and make decisions.

TAC went on to make three crucial decisions that laid the foundation for Taiwan’s integrated circuit (IC) industry.

The first decision was to adopt the CMOS manufacturing process.

In the 1970s, ICs had only been invented two decades earlier, and competing processes like PMOS, NMOS, and Bipolar were still evolving.

TAC predicted that future electronic products would require lightweight, portable, and power-efficient solutions—and thus chose CMOS.

To hedge risk, ITRI still sent engineers for training in NMOS and Bipolar processes as well.

The second decision was to negotiate a comprehensive IC technology transfer.

RCA’s license covered circuit design, photomask making, fabrication, equipment maintenance, testing, packaging, quality control, facility operations, cost accounting, production scheduling, and materials management.

Among CMOS leaders globally, Hughes and RCA were top contenders.

After multiple rounds of negotiation, RCA ultimately agreed to transfer the full suite of technologies.

In April 1976, Taiwan sent its first cohort of 13 engineers to train at RCA, including notable names such as Shih Chin-Tai（史欽泰）, Tseng Fan-Cheng（曾繁城）, Robert Tsao（曹興誠）, Ming-Kai Tsai（蔡明介）, Liu Ying-Ta（劉英達）, Yang Ting-Yuan（楊丁元）, and Chang Ching-Chu（章青駒）.

Eventually, over 40 individuals received training. These “pilgrims of knowledge” would later become luminaries in Taiwan’s IC industry.

RCA, having received licensing fees, earnestly fulfilled its transfer obligations.

Trainees lived together, studied rigorously during the day, and discussed problems at night.

RCA executives quickly realized the Taiwanese had extraordinary learning capabilities and strong motivation.

Trainees discovered that RCA’s technical documentation and archives were well-organized and freely accessible—many returned with suitcases full of copied materials.

TAC’s third major decision was to build a demonstration plant at ITRI with commercial-scale capacity.

Rather than a small experimental lab to save costs, they designed the pilot fab to process 10,000 three-inch wafers per month.

This ambitious scale was intended to prepare for eventual commercialization—mass production would inevitably expose technical problems, prompting faster feedback and corrections.

Astonishingly, within just three months of mass production, ITRI’s yield surpassed that of RCA, exceeding all expectations.

These three decisions enabled Taiwan to fully absorb RCA’s technology—not merely at the surface level, but in depth.

Taiwan replicated the U.S.’s most advanced IC technology solidly and pragmatically.

Given today’s intense global competition, such a cross-border technology transfer would be unimaginable.

That $3.5 million investment by the Taiwanese government—equivalent to NT$140 million—would eventually lead to a semiconductor output worth nearly NT$5 trillion.

It was an investment well worth making.

Phase Two: Technology Transfer and the Birth of UMC

After several years of experimentation at ITRI, RCA’s technology was transferred to private enterprise, launching the second phase.

The first recipient was United Microelectronics Corporation (UMC), founded in 1980.

To ensure success, ITRI required UMC to replicate its setup exactly, following the same “copy exactly” principle RCA had imposed on ITRI.

However, by mid-1980, the U.S. and Japan had begun mass-producing four-inch wafers.

UMC insisted on upgrading from ITRI’s three-inch wafers to four-inch ones, while maintaining brand consistency in equipment to avoid unforeseen issues.

This decision significantly boosted UMC’s production capacity and enabled the company to progress from 7-micron to 3.5- and then 2-micron processes, while reducing manufacturing costs.

In 1983, UMC began mass production; by the next year, it turned a profit thanks to surging demand for home telephone ICs in the U.S.

UMC also inherited a critical department from ITRI—product design.

This function would later allow UMC to spin off numerous IC design firms, establishing Taiwan as a global center for IC design.

Initially, ITRI only intended to transfer manufacturing technologies to UMC.

Product development still required licensing from ITRI and RCA.

UMC’s management believed that without design capabilities, the company would remain dependent.

CEO Robert Tsao convinced Ming-Kai Tsai to leave ITRI and join UMC—though this move temporarily strained relations between the two organizations.

In hindsight, it was a wise choice.

UMC went on to nurture many IC design talents.

Had design remained confined within ITRI, the ecosystem might not have flourished.

Today, Taiwan ranks second globally in IC design, with four of the top ten global IC design firms—MediaTek, Novatek, Realtek, and Himax—originating from Taiwan.

The first two, MediaTek and Novatek, were direct spin-offs from UMC.

Others such as Realtek and Promos were founded by former UMC staff, and the trend extends to companies like Faraday, PixArt, Holtek, ITE, Lien Chang, and SiS—all built on UMC alumni talent.

Phase Three: TSMC’s Founding and a New Business Model

In 1984, ITRI launched a new initiative focused on Very Large-Scale Integration (VLSI), culminating in the 1987 founding of Taiwan Semiconductor Manufacturing Company (TSMC).

Taiwan’s semiconductor industry was now firmly established.

TSMC was launched at a significantly larger scale and higher level than UMC, a reflection of Taiwan’s increasingly mature industrial foundation at the time.

The initiative was led by Morris Chang（張忠謀）, then-president of ITRI, who assumed the role of TSMC’s founding chairman.

ERSO’s former pilot fab director, F.C. Tseng（曾繁城）, also led the transfer of more than 100 engineers from ITRI to the new venture.

In addition, ITRI’s 6-inch VLSI fab was appraised and transferred to TSMC as part of its initial asset base.

TSMC also secured a 27.5% equity investment from Philips and benefited from the company’s global patent umbrella.

TSMC’s founding was built on a pure-play foundry model, focusing solely on wafer fabrication without competing with customers in chip design.

This clarified the vertical division of labor within Taiwan’s semiconductor ecosystem.

The overall IC industry evolved into a fully integrated value chain—from design and manufacturing to packaging and testing—allowing each segment to flourish independently in the global market.

As of 2021, Taiwan accounted for 64% of the global foundry market, with TSMC alone commanding 53% of that share.

In advanced process nodes of 7 nanometers and below, TSMC held an overwhelming lead, dominating nearly the entire market.

The year TSMC was founded, the Walsin Lihwa Group recruited D.Y. Yang（楊丁元）and Ching-Chu Chang（章青駒）to establish Winbond Electronics（華邦電子）.

The following year, in 1988, Min-Chau Wu（吳敏求）returned from the United States with dozens of families to found Macronix International（旺宏電子）.

A vibrant startup culture began to take root in Taiwan’s semiconductor industry.

Domestic universities responded by launching new academic departments focused on electronics, electrical engineering, and materials science, building a steady pipeline of talent.

Wafer foundries and IC design houses also attracted numerous overseas experts, propelling the industry forward by a significant margin.

These three pivotal stages took Taiwan’s semiconductor industry from zero to one, establishing a solid foundation.

From there, it scaled from one to one hundred—creating what is now widely regarded as the “Taiwan semiconductor miracle.”

Throughout this journey, countless individuals—from government officials to private sector pioneers—devoted themselves selflessly.

These unsung heroes are the true source of brilliance behind the island of chips.