From: acquiredfm
TSMC has risen to become an industry leader, largely due to its unique business strategy focused on being a pure-play foundry and its deep understanding and application of the learning curve theory [02:22:03].
The Pure-Play Foundry Model
From its inception, TSMC adopted a strategy of being a “dedicated Pure Play Foundry,” meaning it would not compete with its customers by designing its own chips or end products [01:45:23], [02:22:03]. This was a strategic choice to avoid the conflicts of interest faced by Integrated Device Manufacturers (IDMs) like Intel, which both designed and manufactured chips [01:45:07], [02:33:45].
Morris Chang foresaw the emergence of fabless companies, even before they became prevalent. His early encounters, such as with Gordon Campbell who sought funding to start a company without building a fab, and with ATM, which also had no fabs, solidified his belief in the fabless model’s future [02:16:30], [02:18:37]. This foresight positioned TSMC to cater specifically to companies that wanted to focus on design while outsourcing manufacturing.
The Learning Curve Theory
A cornerstone of TSMC’s business strategies and pricing is the learning curve theory, which Morris Chang played a significant role in refining during his time at Texas Instruments [02:02:33]. The core principle of this theory is that as a company produces more of a certain product, its unit cost goes down due to gained experience and increased efficiency [02:03:14], [02:38:45].
The goal, as per the learning curve, is to become the largest volume player in the industry to achieve the lowest possible costs [02:38:05], [02:38:20]. This means a company might strategically price its products unprofitably in the early stages of a new node generation to out-compete rivals, gain market share, and then benefit from economies of scale once demand is aggregated [02:38:40], [02:39:00].
Strategic Application of Business Strategy & Learning Curve
TSMC’s application of these principles is evident in several key decisions:
Research & Development (R&D) and Capital Expenditures (CapEx)
Morris Chang established a fixed R&D budget of 8% of revenue, regardless of economic downturns, to ensure continuous innovation [00:41:30], [00:42:04]. This provided stability and encouraged bold technical pursuits, such as the focus on the 28 nanometer node, which was considered a “sweet spot” by TSMC’s R&D team [00:43:47], [00:44:05].
The commitment to the 28nm node involved tripling capital expenditures in 2010 to almost $6 billion, a decision that faced questions from the board but was pushed through by Morris Chang, betting on the future demand in the smartphone era [00:43:01], [00:47:33]. He considered it a “bet the company” move, though he believed it would succeed [02:16:14].
Customer Relationships and Pricing Decisions
TSMC’s strategy often involves balancing long-term relationships with short-term financial gains.
-
Nvidia Dispute (2009): After Morris Chang returned as CEO, he resolved a dispute with Nvidia over 40nm node yield and quality issues by offering over $100 million in compensation, cementing a strong partnership [00:34:06], [00:35:19]. This demonstrated a willingness to make a significant financial concession to maintain a crucial customer relationship.
-
Apple Engagement (2010-2011): When Apple approached TSMC for 20nm chips, TSMC’s initial margin target was 50%, while Apple proposed 40% [01:14:37]. Despite the 20nm node being a detour from TSMC’s planned 16nm progression, Morris Chang decided to take the business. However, to manage the massive capital investment required (tens of billions of dollars), TSMC informed Apple they would only fulfill half of their requested demand for 20nm wafers [01:25:27], [01:31:32]. This decision, made in consultation with Goldman Sachs, involved borrowing money rather than cutting dividends or issuing new stock, protecting shareholder interests [01:24:51], [01:31:48].
-
Handling Competition: When Apple briefly placed initial 16nm orders with Samsung, Morris Chang expressed shock but was reassured by Apple’s commitment to buy from TSMC once their 16nm process was ready [01:51:59], [01:54:15]. This highlighted the advantage of TSMC’s pure-play model, as Samsung was a direct competitor to Apple’s products [01:55:13].
Long-term Vision
Morris Chang prioritizes long-range success over short-term gains. This was evident in his decision not to layoff employees by performance reviews in 2008-2009, fearing a loss of respect and the future need for those skilled personnel [01:15:21], [01:20:01].
Impact and Outcomes
TSMC’s strategic choices, especially in combining its pure-play model with aggressive learning curve pricing, have led to significant outcomes:
- Market Dominance: TSMC became the undisputed leader in advanced semiconductor manufacturing [01:16:13]. Its dedication to R&D and massive capital investments allowed it to surpass competitors like Intel in foundry technology [01:40:28], [02:31:01].
- Customer Trust: The pure-play model built deep trust with customers, as TSMC posed no competitive threat [01:44:01]. Intel, conversely, was not trusted by its customers because of its integrated model [01:45:07], [01:55:50].
- Financial Success: Despite initial lower margins, TSMC’s gross margins increased over time, reaching over 50% [01:15:29]. The company’s size and importance have exceeded Morris Chang’s initial expectations [02:15:06].
- Ecosystem Advantage: The concentration of partners, customers, and universities in Hsinchu Science Park in Taiwan created a powerful, physically integrated ecosystem that is difficult to replicate [02:27:45], [02:30:50].
The success of TSMC illustrates a clear example of how a well-executed business strategy centered on a unique value proposition and a deep understanding of industry dynamics can lead to unparalleled market leadership.