AI summary by TwinMind
How Taiwan Created TSMC
Asianometry
12m 14s
Background and Challenge (Early 1970s)
- Taiwan faced economic transition challenge - needed to upscale from agricultural/simple electronics assembly
- Economy consisted mostly of agricultural products and low-value electronic assembly exports
- Previous attempts to build world-competitive chemical industry failed
- Semiconductor industry identified as next logical step for electronics assembly industry
- US dominated semiconductor manufacturing, Japan secondary player
- Taiwan hosted few foreign factories (General Instruments since 1966) for low-value assembly only
- Taiwanese companies lagged far behind cutting edge, no market credibility
Government Strategy Debate
- Ministry of Economic Affairs debated best approach
- Sun Yun Shen (Minister of Economic Affairs, born in Shandong): backed entrance to development of companies and industry
- Wai Shan Ran (Director of National Science Council): argued for funding basic electric engineering sciences for semiconductors to "root" industry in Taiwan
- Jianjin Guo (Premier, Chiang Kai-shek's son): opposed basic research, favored applied research for industrial purposes
- Decision: Focus on applied research and practical skills became consistent theme
ITRI Formation and Technology Transfer
- Founded Industrial Technology Research Institute (ITRI) as publicly funded laboratory
- Purpose: Applied science research with intention to spin off projects as private companies
- ITRI created subcommittee ERSO to tackle semiconductor industry entry
- Strategy: Invite foreign companies from US/Europe to teach Taiwanese specific semiconductor technologies
RCA Technology Transfer Agreement (1976)
- 7 companies invited to bid for contract
- RCA won over General Instruments and Hughes
- Chosen not for most advanced technology but for willingness/capability to teach
- Agreement: RCA invited 30-40 Taiwanese engineers to campus for 1 year
- Training included: raw science of semiconductors, management techniques, industrial knowledge
- RCA provided licensing for technology few years behind cutting edge
- Lucky break: RCA later exited semiconductor industry, left Taiwan with license for all technologies
CMOS Technology Choice
- Taiwan chose to learn Complementary Metal Oxide Semiconductors (CMOS)
- CMOS had not fully matured, small market at time
- Over time became dominant semiconductor manufacturing process technology
- Government acted as venture capitalist, backed young technology that paid off significantly
UMC Formation and Early Results (1980)
- ERSO spun off United Microelectronics Corporation (UMC) in 1980
- UMC headquarters across street from TSMC in Xinjiu
- Used RCA knowledge, acted as traditional vertically integrated semiconductor maker
- Government provided technology, helped recruit talent, closed first deal
- UMC became Taiwan's most profitable industrial manufacturer
- Failed to be flagship success - could not compete in industry
- Produced low-level chips for toys and watches
- Field too competitive for breakthrough success
- ERSO spun off 2 other chip design companies - both failed, acquired by competitors
TSMC Formation and Success
- Represented innovative foundry business model disrupting existing semiconductor market
- Entered with much more advanced technology than UMC
- Technology from Dutch corporation Philips (continuing Dutch-Taiwanese ties)
- Philips provided 1.5 micron process technology input for TSMC's core service
- ERSO sourced high-skilled Taiwanese workers
- Taiwan government invested money but wanted private investors for market oversight
- Philips received substantial chunk of TSMC shares
- Intel and Texas Instruments offered investment chance but passed
- Taiwanese businessmen offered/ordered to invest
- Wang Ching (Formosa Plastics titan) bought 5% because government required, sold after IPO year 2, missed $20 billion fortune
TSMC Strategy Differences
- Focused on international export market from beginning (government request + Morris Chang's American background)
- Helped fuel fabless semiconductor trend in America
- UMC focused on domestic market initially - made company too comfortable, didn't become world-class
- Key lesson: Need to compete across entire world to create world-class firm
Government Development Strategy
- Small companies clustered together, guided by government
- Contrast with South Korea: gave lots of money to special companies (Samsung, LG, etc.) to dominate sectors
- Silicon Valley used as guide
- Founded many small companies in different parts of semiconductor value chain
- Provided starting boost, arranged resources including housing (Hsinchu Science Park)
- Tax incentives: semiconductor companies taxed at just 1.5% in early 1990s
- No subsidies pumping or competition shielding once companies started
- No protective tariffs for captive domestic market
- Did not prop up failing companies (contrast with Latin America/Europe domestic industry strategies)
- Government helped industry as whole during existential threats (earthquake example: state utilities provided water/power to Hsinchu Science Park first over residential areas)
Success Factors Beyond Government Control
- Reverse brain drain from Western countries to Taiwan (1980s-90s)
- Booming Taiwanese economy
- Surge in Chinese nationalism over Diaoyu Islands sovereignty
- Taiwan as "free China" benefited from overseas Chinese returning to help motherland
- Timing and luck factors crucial
- Singapore tried similar strategy few years later with mixed results
Key Lessons for Developing Countries
- Partner with more advanced foreign technologies
- Stress test products in global competitive landscape
- Success never preordained - hard work plus right place/right time
- Applicable to Chinese government's semiconductor strategy development
