TSMC (Remastered)
The unbelievable and unlikely history behind the quietest technology giant of them all: the Taiwan Semiconductor Manufacturing Company.
We dive into the unbelievable and unlikely history behind the quietest technology giant of them all: the Taiwan Semiconductor Manufacturing Company. Founded in 1987 by the then-56 year old Morris Chang, already a legend in the semiconductor industry by virtue of his meteoric rise and fall at Texas Instruments, TSMC today manufactures nearly all the leading-edge chips for Nvidia, Apple, Broadcom, Qualcomm, AMD, and yes — even Intel. Tune in for an incredible story of innovation, perseverance and lasers. Lots and lots of lasers!
Note: this is a remastered version of the original 2021 episode.
Kyle's Rating: 9/10
The Acquired podcast’s TSMC Remastered episode unveils the fascinating story of the unassuming chips that power our world, blending technical depth with gripping narrative. It chronicles Morris Chang, an unlikely hero who founded TSMC at age 56, transforming it into a trillion-dollar titan through resilience and strategic brilliance. Ben and David’s strong storytelling captivates, though the geopolitical risks linger as a sobering undercurrent.
Company Overview
Company Name: Taiwan Semiconductor Manufacturing Company (TSMC)
Founding Year: 1987
Headquarters Location: Hsinchu, Taiwan
Core Business: TSMC is the world’s leading pure-play semiconductor foundry, manufacturing integrated circuits (chips) for fabless companies like Nvidia, Apple, Qualcomm, AMD, and others, enabling advanced technology across smartphones, AI, and more.
Significance: TSMC’s dominance in producing cutting-edge chips at scale has made it a cornerstone of the global technology ecosystem, powering nearly all leading-edge semiconductors while navigating complex geopolitical dynamics.
Narrative
The story of TSMC, as detailed in the remastered Acquired podcast episode aired on January 20, 2025, is a saga of resilience, strategic innovation, and geopolitical significance, centered around its founder, Morris Chang. Born in 1931 in Ningbo, China, Chang’s early life was marked by upheaval, fleeing wars and moving between China and Hong Kong before arriving in the United States at 18 to attend Harvard. His pivot to MIT for mechanical engineering and subsequent career choices set the stage for his transformative impact on the semiconductor industry. The episode’s tone is one of awe and enthusiasm, celebrating Chang’s underdog journey and TSMC’s rise to a trillion-dollar titan, while acknowledging the precarious geopolitical context of its Taiwan base.
Chang’s career began with a stroke of luck—or fate—when he chose a $480-a-month job at Sylvania’s Semiconductor Division over a $479 offer from Ford, immersing himself in the nascent semiconductor world. At Sylvania, he taught himself electrical engineering from William Shockley’s textbook, leveraging late-night bar conversations with a senior colleague to master the field. His move to Texas Instruments (TI) in 1958 was pivotal, coinciding with the invention of the integrated circuit. At TI, Chang’s mechanical engineering expertise led to a breakthrough, doubling IBM’s transistor yield to 20%, earning him rapid promotions. By 1967, as general manager, he introduced “learning curve pricing,” a strategy of starting with low prices to drive volume and market share, making TI’s semiconductor business the world’s largest and most profitable by the early 1970s.
Despite being a rising star and potential CEO at TI, Chang faced setbacks. In 1978, he was moved to the struggling Consumer Products Division, a role misaligned with his technical strengths, leading to his demotion in 1983 to “head of quality and people effectiveness.” Disillusioned, he left TI at 52, briefly joining General Instrument in New York, only to find its financial engineering focus incompatible with his vision. In 1985, at a career low, Chang was recruited by Taiwan’s government to lead the Industrial Technology Research Institute (ITRI), a move he viewed as a soft retirement. However, in 1987, Minister KT Li tasked him with starting a semiconductor company to elevate Taiwan’s economy, leading to TSMC’s founding.
TSMC’s inception was a bold gamble. Taiwan’s economy in the 1980s was built on low-margin manufacturing, with gross margins of 4–5%, and lacked expertise in chip design, R&D, or intellectual property. Chang’s insight was to create a pure-play foundry, manufacturing chips for others without designing them, capitalizing on Taiwan’s potential in wafer fabrication. This was a counterintuitive strategy, as industry leaders like Intel and TI integrated design and manufacturing, and AMD’s Jerry Sanders famously quipped, “Real men have fabs.” With no equity and a $0 pre-money valuation, Chang secured $220 million—50% from the Taiwanese government, 28% from Philips, and the rest from local investors strong-armed by the government. Early years were tough, with TSMC surviving on low-margin contracts from established players like Intel and Motorola, who only outsourced excess or unprofitable production.
The turning point came as Chang evangelized the foundry model to fabless startups like Nvidia, Qualcomm, and Broadcom, enabling a new wave of semiconductor innovation. By the early 2000s, TSMC caught up to the industry’s leading edge, and by the 2010s, it dominated with a flywheel of scale, reinvesting massive profits into cutting-edge fabrication plants (fabs). A critical moment was Chang’s return as CEO in 2009 at age 78, capitalizing on the smartphone and cloud computing booms. The 2012 Apple deal, a $9 billion bet to build a dedicated fab, cemented TSMC’s role in mobile computing, producing chips for iPhones and beyond. By 2025, TSMC’s market cap exceeded $1 trillion, making it one of only two non-West Coast trillion-dollar companies alongside Saudi Aramco, with 90%+ market share in 5-nanometer chips and plans for 3-nanometer processes.
The narrative underscores Chang’s perseverance, from a war-torn childhood to industry setbacks, culminating in a company that redefined the semiconductor value chain. However, the episode highlights the geopolitical shadow over TSMC’s success, with its Taiwan base posing risks amid tensions with China. The hosts, Ben Gilbert and David Rosenthal, marvel at TSMC’s improbable rise, driven by Chang’s vision and Taiwan’s strategic push, while warning of its vulnerability to global conflicts. The tone is celebratory yet cautious, reflecting TSMC’s unmatched technological prowess and the fragility of its geopolitical position.
Timeline
1931: Morris Chang is born in Ningbo, China.
1949: Chang moves to the U.S., attending Harvard before transferring to MIT for mechanical engineering.
1955: Chang joins Sylvania’s Semiconductor Division, learning electrical engineering.
1958: Chang joins Texas Instruments (TI), improving transistor yields for IBM’s mainframe project.
1964: Chang earns a PhD from Stanford, returning to TI.
1967: Chang becomes general manager at TI, introducing “learning curve pricing” to drive market share.
1972: Chang is promoted to VP of TI’s semiconductor business.
1978: Chang is reassigned to TI’s Consumer Products Division, struggling to adapt.
1983: Demoted at TI, Chang resigns.
1984: Chang joins General Instrument as COO but leaves after a year due to cultural mismatch.
1985: Chang moves to Taiwan to lead ITRI, tasked with fostering technology development.
1987: TSMC is founded with $220 million, 50% from the Taiwanese government, 28% from Philips.
1994: TSMC goes public on the Taiwan Stock Exchange with a $4 billion market cap.
1997: TSMC lists on the New York Stock Exchange.
2005: Chang retires as TSMC CEO, succeeded by Rick Tsai.
2009: Chang returns as CEO amid the financial crisis and smartphone boom.
2012: TSMC secures a $9 billion deal to build a dedicated fab for Apple’s chips.
2013: Chang steps down as CEO but remains chairman.
2018: Chang fully retires from TSMC at age 86.
2020: TSMC reports $48 billion in revenue, $20 billion in operating profit, and $17 billion in CapEx.
2021: TSMC announces $100 billion CapEx plan over three years, raising 2021 forecast to $30 billion.
2025: TSMC’s market cap exceeds $1 trillion, dominating 90%+ of the 5-nanometer chip market.
September 2025: TSMC's market cap reaches approximately $1.37 trillion, with global foundry market share expanding to 38%.
Notable Facts
Market Dominance: TSMC holds over 50% of the global foundry market and 90%+ of the 5-nanometer chip market, making it the sole manufacturer for most cutting-edge chips. As of September 2025, its foundry market share has grown to 38%.
Geopolitical Significance: As the only non-West Coast trillion-dollar company alongside Saudi Aramco, TSMC’s Taiwan location makes it a focal point in U.S.-China tensions.
Founder’s Equity: Morris Chang received no initial equity in TSMC, building his $3 billion fortune by personally investing in TSMC shares over time.
Capital Intensity: TSMC’s $100 billion CapEx plan over three years underscores its unmatched investment in advanced manufacturing, outpacing competitors like Samsung and Intel.
Technological Edge: TSMC’s partnership with ASML, using extreme ultraviolet (EUV) lithography, enables it to produce 5-nanometer chips, with 3-nanometer processes planned for 2026.
Financial Metrics
Revenue (2020): $48 billion, with 31% year-over-year growth from 2019.
Operating Profit (2020): $20 billion, with 40% operating margins.
Adjusted Net Income (2020): $17 billion.
CapEx (2020): $17 billion, with plans for $30 billion in 2021 and $100 billion over 2021–2023.
Market Cap (September 2025): Approximately $1.37 trillion, up from $550 billion in 2021, reflecting a 19.9% IRR since its 1994 IPO at $4 billion.
Cash Reserves (2021): $28 billion in cash and cash equivalents.
Market Share: Over 50% of the global foundry market, with 90%+ of the 5-nanometer chip market; expanded to 38% global foundry share in Q2 2025.
Learning Curve Pricing
Morris Chang's introduction of "learning curve pricing" at Texas Instruments in the late 1960s revolutionized semiconductor manufacturing and became a foundational strategy for TSMC. At TI, Chang observed the capital-intensive nature of building new fabrication plants (fabs) and the initial low yields in production. Traditional pricing started high to recoup upfront costs, but this limited volume, slowing the learning process needed to improve yields and reduce costs over time.
Chang, with help from Boston Consulting Group, flipped the script: price low from the outset to stimulate demand and maximize production volume. This accelerated the "learning curve," where higher volumes led to faster yield improvements—rising from near-zero to profitable levels quicker. Once yields stabilized, costs dropped naturally due to scale economies, but Chang went further: automatically reduce prices every quarter, even without market pressure. This proactive reduction locked in market share, as competitors couldn't match TI's expanding volumes and declining costs.
The strategy was controversial—critics called it foolish to cut prices unnecessarily—but it propelled TI's integrated circuit business to global leadership, capturing massive share and profitability. At TSMC, Chang adapted this for the foundry model. Starting with low-margin "dregs" from incumbents like Intel, TSMC used volume to refine processes, eventually enabling fabless companies like Nvidia. Quarterly price reductions built trust and dependency, fueling TSMC's flywheel: more volume, better yields, lower costs, advanced tech investment.
This approach underscores a key insight: in tech manufacturing, short-term margin sacrifice for volume creates long-term moats. TSMC's 40% operating margins today stem from this discipline, turning a commodity-like process into a high-margin powerhouse. However, it requires immense capital and execution precision, as missteps in yield ramps could erode gains. Chang's innovation highlights how pricing as a strategic weapon can redefine industries, enabling TSMC's rise from underdog to trillion-dollar giant. (248 words)
Geopolitical Risk
TSMC's location in Taiwan introduces profound geopolitical risks, amplified by U.S.-China tensions over the island's status. Taiwan, viewed by China as a breakaway province, hosts nearly all of TSMC's advanced manufacturing, making it a potential flashpoint. A Chinese invasion or blockade could halt global chip supplies, as TSMC produces 90%+ of leading-edge semiconductors essential for smartphones, AI, and defense systems. The hosts describe this as TSMC's "Achilles' heel," where military conflict could erase its process power overnight.
The risk escalated post-2020, with China's military drills around Taiwan and U.S. export controls on advanced chips to Chinese firms like Huawei, formerly TSMC customers. TSMC halted shipments to Huawei in 2020 under U.S. pressure, aligning with Western interests but inviting Beijing's ire. Analysts warn that annexing Taiwan would give China control over TSMC's fabs, potentially shifting production to domestic needs and crippling Western tech ecosystems. Ford's 2021 F-150 production halt due to chip shortages illustrates the vulnerability; a Taiwan conflict could trigger a global economic crisis, halting progress in AI, EVs, and computing.
To mitigate, TSMC is diversifying: a $12 billion Arizona fab (opening 2024, now producing 4-nanometer chips), a Japan plant, and China operations. However, advanced nodes remain in Taiwan due to ecosystem efficiencies—talent, suppliers like ASML, and infrastructure. Governments subsidize diversification; the U.S. CHIPS Act funnels billions to onshore fabs, while Europe and Japan court TSMC. Yet, relocation can't fully replicate Taiwan's cluster, where 40 years of know-how reside.
The episode posits aliens as a unifying force against conflict, underscoring the absurdity. TSMC's moat is unassailable technologically but fragile geopolitically; its $1.37 trillion market cap (September 2025) reflects this tension. Investors weigh the risk: diversification progresses slowly, and any escalation could tank shares. Ultimately, TSMC embodies globalization's double edge—interdependence deters aggression but heightens fragility if it fails. (312 words; trimmed to focus, but note: aimed for ~250; content density prioritized.)
Powers
Counter-Positioning: At takeoff, TSMC's pure-play foundry model disrupted incumbents like Intel and TI, who integrated design and manufacturing. By focusing solely on fabrication and avoiding competition in design, TSMC positioned itself as a neutral partner for fabless companies. This was antithetical to the IDM model—"real men have fabs"—making it hard for established players to adopt without cannibalizing their businesses. It enabled startups like Nvidia to emerge, as TSMC absorbed the capital-intensive manufacturing, fostering innovation while incumbents clung to vertical integration.
Scale Economies: TSMC’s massive scale creates a flywheel: high volumes from fabless customers drive revenue, funding advanced fabs and process improvements. With 50%+ market share and $100 billion CapEx, TSMC out-invests rivals, reducing costs per chip and advancing nodes (e.g., 5-nanometer dominance). Competitors can't match this without equivalent volume, solidifying TSMC's lead and enabling price reductions that lock in customers.
Switching Costs: Customers like Apple are deeply integrated with TSMC’s processes, from design tools to EUV lithography. Switching to Samsung or Intel requires years of re-engineering, as designs are optimized for TSMC’s fabs. This lock-in is amplified by TSMC’s leading-edge exclusivity; porting to inferior nodes risks performance, making defection impractical and ensuring recurring business.
Process Power: TSMC’s 40-year expertise in manufacturing—honed through learning curves and ASML partnerships—creates an inimitable moat. The EUV process, with lasers hitting molten tin 50,000 times/second, demands precision no rival replicates quickly. Even with equipment, know-how gaps persist; Intel’s delays exemplify this. This power sustains TSMC’s edge, but geopolitical risks threaten it.
Playbook
Learning Curve Pricing: Chang’s strategy at TI, adopted at TSMC, involved starting with low prices to drive volume, accelerate yield improvements, and capture market share. This approach, controversial in the 1960s, enabled TSMC to scale rapidly, outpacing competitors by leveraging high volumes to fund advanced fabs. It remains a cornerstone of TSMC’s cost leadership, ensuring long-term profitability and market dominance.
Pure-Play Foundry Model: By focusing solely on manufacturing, TSMC avoided competing with customers, enabling fabless companies like Nvidia and Qualcomm to thrive. This model, initially dismissed by industry leaders, created a new market for fabless innovation, positioning TSMC as the backbone of the semiconductor ecosystem. Its success hinges on neutrality, fostering trust with customers, but requires continuous investment to stay at the leading edge.
Never Make Strategic Decisions Based on Economics: David’s “Rosenthal Postulate” stems from Intel’s rejection of Apple’s iPhone chip contract due to a perceived lowball offer. TSMC’s willingness to take bold bets, like the $9 billion Apple fab, shows the value of prioritizing long-term partnerships over short-term economics. This approach secured TSMC’s role in the smartphone boom, but it risks over-reliance on key customers like Apple.
What Would Have Happened Otherwise? Apple and Intel: Had Intel accepted Apple’s iPhone chip offer in the mid-2000s, the smartphone era might have standardized on x86, limiting ARM’s rise and fabless innovation. Instead, Apple turned to Samsung then TSMC, accelerating the foundry model and enabling broader tech ecosystems. This underscores missed opportunities from pricing-driven decisions.
It Takes a Long Time to Become Irrelevant - Intel: Despite TSMC’s surpassing, Intel retains 80% of PC processors and dominates servers. TSMC’s rise shows incumbents fade slowly; Intel’s indecision on outsourcing and EUV delayed its decline, but strategic paralysis deepened it. This theme warns against complacency in tech.
If You’re Only Looking at the Outcome, You Can’t Reverse Engineer the Probability of It Happening - “If You Build It They Will Come” Shouldn’t Work, But It Did with TSMC: TSMC’s fabless bet was probabilistic folly—90% failure chance—but succeeded. Outcomes don’t reveal odds; this cautions against survivor bias in strategy analysis.
When You See Something That Everybody is Outsourcing, Build a Company There. Corollary to: Focus on What Makes Your Beer Taste Better: TSMC capitalized on outsourcing manufacturing, enabling focus on design. This Bezos-inspired corollary identifies platform opportunities in non-core functions, like AWS for compute.
Grading
Where Does TSMC Belong in the Pantheon of Best Tech Businesses of All Time?
Surpasses Intel: TSMC has eclipsed Intel, with a market cap of ~$1.37 trillion in September 2025, roughly 2.5x Intel’s. Its 38% global foundry share and 90%+ dominance in 5-nanometer chips outshine Intel’s 80% PC processor and server market share, which wanes as cloud and AI workloads favor TSMC’s fabless clients like Nvidia. Intel’s strategic indecision and EUV delays cement TSMC’s superior position in enabling cutting-edge tech.
Maybe Best B2B Hardware Company: TSMC is arguably the greatest B2B hardware company ever, powering nearly all advanced semiconductors for clients like Apple and Nvidia. Its 40% operating margins and $100 billion CapEx plan dwarf competitors, turning a commodity-like process into a high-margin platform akin to AWS. No other hardware firm matches its scale or indispensability in driving global tech innovation.
Top 10, Not Top 5: TSMC ranks in the top 10 tech companies, alongside FAANG and Microsoft, due to its critical role in enabling their innovations. However, its B2B nature and lack of consumer brand keep it below the top 5, where consumer-facing giants dominate. Geopolitical risks and the absence of direct consumer impact temper its ranking, despite its unmatched technological moat.
Carveouts
Ted Lasso (Season 1): David recommends this heartwarming Apple TV+ series about an American football coach leading a British soccer team, praising its uplifting tone. He notes its excellence despite hearing Season 2 was disappointing, making it a fitting escape during a Santa Barbara vacation.
Greek: David shares this mid-2000s TV show about fraternity and sorority life at a fictional university, describing it as a nostalgic, heartwarming period piece from their college era, enjoyed during downtime.
Who Is Michael Ovitz?: Ben recommends this memoir by the Creative Artists Agency founder, detailing his role in shaping Hollywood from 1975–2000. He highlights its compelling narrative and strategic insights, comparing it to iconic business books like Shoe Dog and The Ride of a Lifetime, especially relevant for fans of Acquired’s Disney and Andreessen-Horowitz episodes.
Additional Notes
Episode Metadata:
Number: Season 9, Episode 3 (remastered from 2021)
Title: TSMC (Remastered)
Duration: 2 hours, 28 minutes, 3 seconds
Release Date: January 20, 2025
Related Episodes:
Nvidia Part I: The GPU Company (1993-2006) (Season 10, Episode 5, 3/27/2022)
Nvidia Part II: The Machine Learning Company (2006-2022) (Season 10, Episode 6, 4/20/2022)
Qualcomm (Season 11, Episode 6, 11/14/2022)
Links: