The Management Philosophy of TSMC Founder Morris Chang - The 'Game-Changing' Move That Moves the World
TSMC (Taiwan Semiconductor Manufacturing Company) established the world's first large-scale business model for 'semiconductor manufacturing services (foundry)' and has grown into a company boasting one of the largest market capitalizations in the world. Its founder, Morris Chang, launched the company at the age of 56 and is known as the person who reshaped the entire structure of the semiconductor industry using the 'learning curve' theory as his weapon.TSMC.
In this article, we will explain in detail the growth story of TSMC, including specific strategies for fundraising, the secret stories behind partnerships with Apple and Nvidia, and the impact brought about by its unique 'Pure Play Foundry' strategy. Furthermore, we will touch upon the latest nanometer processes and global expansion, and look ahead to where TSMC and the semiconductor industry are headed.
1. The Victory of the 'Pure Play Foundry' Contrarian Strategy
1-1. The Decisive Difference from Intel and Existing IDMs
In the early days of the semiconductor industry, IDMs (Integrated Device Manufacturers) such as Intel held a near-monopoly with their vertically integrated business models. Since they handled both product design and manufacturing in-house, the idea of outsourcing manufacturing to another company was extremely rare.
Drawing on the 'learning curve' theory he learned during his time at Texas Instruments (TI) and anticipating the globalization of the entire industry, Morris Chang presented a vision that 'because we do not design products ourselves, we can become a manufacturing partner for every company.' This was the 'Pure Play Foundry' strategy—a contrarian approach of being the first in the world to develop a large-scale, manufacturing-specialized service without ever touching design.
1-2. Symbiosis with Fabless Companies
In the late 1970s and 1980s, the rise of fabless (design-specialized) companies was gradually beginning. However, at the time, the focus was on 'small-scale outsourcing of surplus capacity,' where many device manufacturers offloaded excess production. Morris Chang was convinced that 'fabless companies will expand in the future' and built a large-scale 'full-service' foundry. In the early days, when major customers were few, he accepted work as a secondary line for other IDMs while gradually building production capacity to match the increasing demand from fabless companies.
2. Apple, Nvidia, Qualcomm—Battles with Global Customers
2-1. Meeting Nvidia: From a Small GPU Company to a Major Customer
In the late 1990s, a letter arrived at TSMC headquarters from a small GPU manufacturer. It was a direct appeal from Nvidia CEO Jensen Huang. At the time, Nvidia was on the verge of a management crisis. Although some within TSMC dismissed it as 'too small-scale,' Morris Chang trusted his intuition and decided to engage in a direct dialogue with Jensen. As a result, they won the contract for full GPU manufacturing and, within just a few years, grew Nvidia into one of their major customers.
However, in 2009, problems with yields (the percentage of chips usable as products) in the 40-nanometer process became serious, causing significant friction with Nvidia. In the end, Morris Chang returned as CEO and ultimately repaired the relationship with Nvidia by offering a settlement of over $100 million as a lump-sum compensation, and they still maintain their position as the most important customer in the GPU field today.
2-2. Capturing Qualcomm's Market Share and IBM's Struggle
Qualcomm initially used IBM's semiconductor division as its primary foundry, but eventually shifted production significantly to TSMC. This dealt a major blow to IBM's semiconductor business and further increased TSMC's presence. As production scale expanded, costs decreased, and the 'learning curve' advantage of being able to mass-produce Qualcomm's high-performance chips at lower prices became the deciding factor.
2-3. The Big Gamble with Apple: 20nm Development and Fundraising
However, TSMC's biggest turning point was the arrival of Apple. Around 2010, as the iPhone began to sweep the world, Apple was looking for a partner to handle large-scale chip manufacturing. Although TSMC had reached world-class levels in the 28-nanometer process, what Apple demanded was an intermediate process called '20nm'. Developing it required massive additional investment and research.
The Barrier of Fundraising: Corporate Bond Issuance, Capital Increase, or Maintaining Dividends?
TSMC had already invested a significant portion of its cash on hand into the 28nm process, and further capital investment posed a major financial risk. The dilemma was 'where to raise the funds.' Specifically, they considered the following options.
Reducing or Suspending Dividends
However, many shareholders expected stable dividends, and doing so risked a sharp drop in the stock price.Large-scale Capital Increase
There was a risk of backlash from existing shareholders who dislike stock dilution, and the speed of execution would also be constrained.Issuance of Corporate Bonds
Given TSMC's high credit rating at the time, there was a high probability of securing favorable interest rate conditions.
Ultimately, after receiving advice from investment banks like Goldman Sachs, they decided on fundraising through large-scale corporate bonds. Furthermore, regarding the massive production volume proposed by Apple, Morris Chang personally proposed a compromise to Apple COO Jeff Williams: 'We will accept half the order initially to prioritize the company's financial strength and yield improvement,' and tenaciously secured an agreement.
The Road to 16nm and Samsung's Lead
However, because they went through the 20nm process, the development of 16nm was slightly delayed. Taking advantage of that gap, Apple initially utilized Samsung's production lines and placed some orders with Samsung. However, when TSMC finally mass-produced the 16nm process, its high yield and supply stability were recognized, and most production was consolidated with TSMC. This dramatically strengthened the relationship between TSMC and Apple, leading the company to handle the manufacturing of high-value-added products such as iPhone SoCs and M-series chips entirely.
3. The 'Learning Curve' as a Growth Engine
3-1. The Multiplication of Cost and Mass Production Experience
The 'learning curve' theory that Morris Chang learned from the Boston Consulting Group (BCG) during his time at TI states that 'as production volume increases, experience grows and unit costs decrease.' This creates a winning pattern where the company that secures the largest production volume in the industry ultimately has strong price competitiveness and can reinvest abundant profits into research and development.
3-2. The 8% Rule—Continuous Investment in R&D
From an early stage, TSMC established a policy of 'allocating 8% of revenue to R&D.' By continuing to invest a fixed percentage regardless of economic conditions, they have steadily developed the world's most advanced manufacturing processes, such as CoWoS (advanced packaging technology) and EUV lithography. This has also been a factor in accelerating the learning curve.
4. New Frontiers: 2nm Process and Global Expansion
4-1. The Overwhelming Ecosystem of Taiwan's Hsinchu Science Park
Around the Hsinchu Science Park in Taiwan, where TSMC's headquarters is located, EDA tool vendors (such as Cadence and Synopsys), design firms like ARM and MediaTek, and university research institutions are all situated in close proximity. Expansion of factories (fabs) is underway for the development of the latest 2-nanometer generation. Such a dense ecosystem serves as a powerful foundation for accelerating yield improvements and technical collaboration.
4-2. The Arizona Fab and Supply Chain Restructuring
Due to geopolitical risks and the demand for supply chain resilience, TSMC is also building large-scale fabs in Arizona, USA. However, an ecosystem comparable to Hsinchu is still in the process of formation, and the extent to which mass production of cutting-edge nodes can be allocated to overseas bases remains a challenge. On the other hand, with the backing of the U.S. government and major clients, further strengthening of the global production system through massive capital investment is expected in the future.
This concludes our review of the journey of TSMC and Morris Chang, who pioneered the new frontier of semiconductor manufacturing known as the foundry, from a concrete perspective including capital procurement. We have examined how enormous capital investment, learning curve strategies, and the interplay with companies like Apple and NVIDIA have come to fruition to build TSMC's current 'unrivaled position.' By maintaining its unique stance as a 'manufacturing partner that competes with no one' while these elements are intricately intertwined, TSMC is stepping into the unprecedented nanometer realm. The entire global technology industry will be watching with bated breath for the next big move toward further evolution.
