← Back to list
Voss CapitalDeep research10 Sep 2023Source: vosscapital.substack.com

Chip Wars — Summary of 'Chips Wars' by Chris Miller, the History of the Semi Conductor Industry and the Potential Geopolitical (and Investing) Implications

Voss Capital is a Houston hedge fund founded by Travis Cocke in 2011, running value-oriented, bottom-up strategies focused on underfollowed small- and mid-cap special situations through long/short and long-only funds, increasingly turning activist.

Travis Cocke · 2011 · 美国休斯顿Small/mid-cap special situations

In plain words

This report covers the history of the semiconductor industry, from Cold War rivalry sparking chip innovation to Japan, Korea, and Taiwan's rise, and the US-China tech war. For everyday investors, the key takeaway is that chips aren't just market-driven—national strategy and geopolitics (like US restrictions on China or Taiwan's manufacturing role) directly shape which companies win. It's worth reading because it explains why firms like TSMC and Nvidia matter so much, and why investing in tech means also watching who backs it.

AI SummaryAI-generated · may contain errors · verify against the original

This report discusses Voss Capital's analysis of the semiconductor industry through a geopolitical lens, inspired by reading the book Chip Wars. The core argument is that semiconductors—especially advanced logic chips produced by TSMC—have become national strategic assets, akin to "the new oil." Whi

~24 min full read · 30 sections
Deep Analysis

Theme and Background

This chapter reviews the early development of the semiconductor industry from 1945 to 1965. The market backdrop was the Cold War and the Space Race between the United States and the Soviet Union. Geopolitical competition—particularly the Soviet launch of Sputnik—became the core driving force behind the initial innovation of American semiconductor technology.

Core Argument

The report's central thesis is that the birth and early major advances of the semiconductor industry were not purely market-driven or technology-driven, but were directly catalyzed and funded by geopolitical competition (especially the U.S.-Soviet Cold War). A counterintuitive judgment is that although the transistor and the integrated circuit were revolutionary inventions, their initial market was not consumer electronics but state-led aerospace and military programs.

Key Evidence and Data

The report uses historical events and specific cases to support the geopolitical-driven thesis:

1. Technological Origins: The transistor was announced by Bell Labs in 1948 and was analogized to a neuron. The first "integrated circuit" was invented roughly a decade later.

2. Initial Market: After the invention of the integrated circuit, its first key market was the National Aeronautics and Space Administration (NASA), directly stemming from President Kennedy's announcement of the moon landing program in response to Soviet space leadership.

3. Military Applications: In 1962, the U.S. Air Force began seeking a new guidance computer for the "Minuteman II" missile, extending the application of chips into the nuclear weapons domain.

4. Industry Law: In 1965, Gordon Moore proposed "Moore's Law," predicting that the number of components on a chip would at least double every year.

5. Development Model: The industry made significant advances after receiving funding from NASA and the military in succession, and only then was it ready to target mass consumers.

Companies/Assets Involved

  • Bell Labs: Inventor and announcer of transistor technology.
  • Bob Noyce: Co-inventor of the integrated circuit, whose products were initially used in rockets.
  • National Aeronautics and Space Administration (NASA) and U.S. Air Force: As early core customers and funding sources, they directly created market demand and drove technological progress.

Investment Implications

For investors, this means that when evaluating semiconductor companies, one must consider the non-market demand or critical initial support that national strategy and geopolitical factors may create. History shows that competition at the national level can provide a nascent technology industry with essential launch markets and R&D funding. This suggests that investors should focus on hard-tech fields that are highly aligned with current national strategic priorities (such as artificial intelligence, quantum computing, and advanced process nodes).


Theme and Background

This chapter reviews the evolution of the competitive landscape in the semiconductor industry from the 1980s to the early 1990s. It focuses on analyzing how Japan transformed from a technology partner of the United States into a formidable competitor, and the subsequent recovery of the United States through policy adjustments and market changes.

Core Argument

The author's core argument is: Leadership in the semiconductor industry is not static, but rather the result of the dynamic interplay of multiple factors, including national industrial policy, capital costs, geopolitics, and market competition. A counterintuitive assessment is that the U.S., which once faced the risk of being completely overtaken by Japan in the 1980s ("Silicon Valley might repeat Detroit's fate"), ultimately recovered not solely through technological innovation, but thanks to financial policy reforms, competitors' mistakes (Japan's economic collapse), and the establishment of new advantages in specific high-value market segments (such as microprocessors).

Key Arguments and Data

The report uses historical comparisons and data to support this dynamic competition viewpoint:

1. Japan's Rise Strategy and Advantages:

  • Strategy: Collaborating with American innovators, acquiring intellectual property, and leveraging government support to build fabrication plants.
  • Advantages: Access to cheaper loans compared to U.S. companies for subsidies; worker efficiency far higher than U.S. peers; transforming from a "cheap" image to one of innovation and quality.

2. Initial U.S. Struggles and Response:

  • Struggles: The Reagan administration was initially indifferent to industry support, comparing subsidies for silicon chips to "subsidies for potato chips." Japan could produce the same quality of products at a lower cost.
  • Response: By deregulating the financial industry and lowering capital gains taxes, a large influx of venture capital was brought into Silicon Valley.

3. Reversal of the Competitive Landscape (Late 1980s to Early 1990s):

  • U.S. Wins in Market Segments: Intel began to dominate the microprocessor market; Qualcomm rose in the telecommunications market; Micron carved out a niche in the DRAM market.
  • Multiple Blows to Japan: Japan simultaneously faced a revival of U.S. protectionism, an overheated economy, and new Asian competition (South Korea). Its semiconductor advantage was built on an unsustainable foundation of government-supported overinvestment. After the 1990 economic crisis, major Japanese semiconductor companies still doubled down on DRAM, while low-cost producers like Micron and Samsung weakened their Japanese rivals.
  • Military Technology Demonstration: The U.S. demonstrated its technological advantage in chip-embedded missiles during the Iraq War, hailed as a "victory of silicon over steel."

4. Evolution of Dominance in the DRAM Sector by Country/Region:

Period Dominant Force Key Companies/Countries
1980s Japan Major Japanese semiconductor companies
Early 1990s to Present South Korea & U.S. (Niche) Samsung (South Korea), Micron (U.S.)

Companies/Assets Involved

  • Japanese Companies: Sony is a representative of Japan's image transformation and the consumerization of technology. Major Japanese semiconductor companies (unnamed) once dominated in DRAM but later declined due to strategic errors and cost competition.
  • U.S. Companies:
  • Intel: Established a dominant position in the microprocessor market, a key factor in the U.S. recovery.
  • Micron: Successfully carved out and defended a niche in the DRAM market against Asian competition.
  • Qualcomm: Rose in the telecommunications market, representing U.S. success in an emerging field.
  • South Korean Company: Samsung began its rise, initially following a strategy similar to Japan's earlier approach, but ultimately became a global leader in DRAM and intensified competitive pressure on Japan.

Investment Implications

The implication for investors is: Leadership transitions and shifts in the supply chain landscape are normal in the semiconductor industry, driven by complex non-market factors. When investing, focus should be placed on:

1. National Industrial and Financial Policy: Cheap capital and active government support can rapidly alter the competitive landscape (e.g., Japan in the 1980s).

2. A Company's Strategic Focus: In hyper-competitive or cyclical markets (e.g., DRAM), blindly pursuing volume over profit can lead to failure (e.g., the lesson of Japanese companies). Success often belongs to those that build barriers in high-value market segments (e.g., Intel in microprocessors).

3. Geopolitical and Macroeconomic Risk: A competitor's economic collapse (e.g., Japan's 1990s crisis) or geopolitical events can create opportunities for other players (e.g., the benefit to South Korea and the U.S.).

4. Technology Spillovers into Other Areas: Semiconductor advantages can be directly translated into strategic advantages in other fields, most notably military technological superiority, which in turn reinforces its status as a strategic national asset.


Theme and Background

This section reviews the formation and consolidation of the global division of labor in the semiconductor industry from the early 1990s to 2015. The core background is that, driven by globalization, the design, manufacturing, and assembly stages of the chip industry became geographically separated, giving rise to a global supply chain where the US leads in design, Taiwan in manufacturing, and China in assembly.

Core Argument

The author's central thesis is that the global division of labor led to a high concentration of semiconductor manufacturing capacity in Asia (especially Taiwan and South Korea), making the US's leading position in chip design "fragile." A counterintuitive judgment is that early significant investments by US companies (e.g., Intel's funding of ASML's EUV technology) ultimately mainly strengthened the manufacturing dominance of their Asian competitors (e.g., TSMC), thereby eroding America's own advanced manufacturing capabilities.

Key Arguments and Data

1. TSMC's Rise: Founder Morris Chang created the "foundry" and "fabless" models, lowering the startup costs for chip design companies and spurring dozens of new design firms. TSMC leveraged scale economies to become a low-cost producer, and its business thrived throughout the 1990s.

2. Consolidation of the Global Division of Labor:

  • United States: Maintains a strong lead in cutting-edge chip design (including semiconductor design software and tools).
  • Taiwan: Becomes a powerhouse in chip manufacturing.
  • China: Occupies a critical niche in equipment assembly (e.g., Foxconn iPhone assembly plants).

3. Monopoly in Lithography: The market for the most advanced lithography machines is monopolized by the Dutch company ASML, with each machine worth hundreds of millions of dollars. ASML's deep relationship with TSMC (rooted in Philips's early investment and technology transfer to TSMC) is key to its success, which in turn reinforces TSMC's leadership.

4. Intel's Strategic Missteps:

  • Although Intel invested $200 million in ASML's extreme ultraviolet (EUV) lithography technology (a significant portion of ASML's total funding), it failed to leverage that technology advantage.
  • Intel missed the semiconductor architectural shift required for artificial intelligence and encountered problems in its manufacturing processes, failing to keep pace with Moore's Law.
  • As of 2020, half of the world's EUV lithography tools (funded and nurtured by Intel) were installed at TSMC.

5. Asian Dominance in Manufacturing:

  • The memory chip market is dominated by Asia.
  • Mobile phone chips are primarily manufactured at TSMC and Samsung.
  • Nvidia, the biggest "fabless" success story, relies mainly on TSMC to manufacture its GPUs.

Companies/Assets Mentioned

Company Role/Key Data Author's Implied View
TSMC Global chip manufacturing hegemon; pioneered the foundry model; one of the largest customers of ASML's EUV lithography tools; symbiotic relationship with fabless design firms like Nvidia. Bullish on its industry position, but highlights geopolitical concentration risk.
ASML Monopolist in cutting-edge lithography machines (EUV); machines worth hundreds of millions of dollars; deep ties with TSMC. Bullish on its technology monopoly position; a critical bottleneck equipment maker.
Intel Previously invested $200 million in ASML's EUV development; retains advantages in server chips, etc.; but missed the mobile and AI architecture transition, and fell behind in manufacturing. Bearish on its past strategy; sees two possibilities over the next five years—either regaining leadership or going bankrupt—concerning the survival of US manufacturing capability.
Nvidia Biggest "fabless" success story; its GPUs are widely used in the most advanced data centers; chips are primarily manufactured by TSMC. Bullish on its design capability, but its success is highly dependent on TSMC's manufacturing.
Samsung One of only two companies globally capable of manufacturing high-end processors (the other being TSMC). Highlights its critical position in the global manufacturing landscape.
China (Assembly) Occupies a critical niche in equipment assembly, e.g., Foxconn is the primary assembler of iPhones. Points out its key but non-cutting-edge role in the global supply chain.

Investment Implications

1. Evaluate Supply Chain Risk: When investing in chip design companies (such as Nvidia) that are highly reliant on TSMC or Samsung for cutting-edge manufacturing, the risk of supply chain disruption due to geopolitics must be incorporated into valuation models.

2. Focus on "Rebalancing" Opportunities: US industrial policies (e.g., subsidies, support) aimed at ensuring supply chain security are a potential recovery catalyst for US-based manufacturers like Intel. Investors should closely monitor relevant US government actions and Intel's execution progress.

3. Focus on Key Bottlenecks: As a monopolist in lithography technology, ASML's capacity allocation and technology roadmap directly influence the pace of innovation across the entire semiconductor industry, making it a leading indicator of industry health and a core investment target.

4. Distinguish Design vs. Manufacturing Risk: In semiconductor investing, it is critical to separate the asset-light, high-growth logic of design companies from the asset-heavy, geopolitically sensitive logic of manufacturing companies, as they face distinctly different risks and drivers.


Theme and Background

This section focuses on a series of strategic initiatives China has undertaken since 2016 to enhance its position and influence in global technology competition, particularly in the semiconductor sector. The backdrop is the intensifying US-China technology rivalry, as China seeks to reduce dependence on foreign sources and gain dominance in future technology fields such as artificial intelligence and advanced military equipment.

Core Argument

The author’s central thesis is that China is leveraging a comprehensive state-led strategy—including industrial policy, technology-for-market exchanges, overseas investments, and acquisitions—to boost its meager 6% share of the global semiconductor supply chain. However, these efforts have not yet led to a fundamental shift in global technology leadership. A counterintuitive insight is that despite China’s frequent moves and considerable determination, the strategic importance of Taiwan (TSMC), as a critical node in semiconductor manufacturing, has not diminished; on the contrary, it has become more prominent in the current competitive landscape.

Key Arguments and Data

The report uses specific data and cases to illustrate China’s strategic path to increasing semiconductor self-sufficiency and the challenges it faces.

1. Extremely Low Starting Market Share: Across the entire semiconductor supply chain (including chip design, intellectual property, tools, manufacturing, etc.), Chinese companies hold a total market share of only 6%. The disparity with other major participants is stark:

Country/Region Market Share
United States 39%
South Korea 16%
Taiwan 12%
China 6%

2. State Strategic Direction: One core goal of the “Made in China 2025” plan is to significantly reduce reliance on imported chips, indicating a strategic orientation toward reducing rather than deepening economic integration.

3. Specific Implementation Methods:

  • Technology Transfer and Market Access: For example, IBM’s then-CEO Ginni Rometty announced the opening of chip technology to Chinese partners, aiming to help China build a local computer ecosystem. Similar collaborations exist with Qualcomm and AMD. China tends to partner with US companies that are struggling or not in a leadership position, in exchange for technology.
  • Capital Infiltration and Acquisition Attempts: Tsinghua Unigroup, closely linked to the Chinese government, attempted to acquire a significant stake in TSMC and MediaTek. This move placed both companies, which rely heavily on the Chinese market, in a dilemma.
  • Circuitous Acquisition Blocked: Tsinghua Unigroup first acquired a 6% stake in Lattice Semiconductors, then subsequently sold it. Later, Canyon Bridge, a US company with Chinese government backing, made a takeover bid for Lattice, but the transaction was ultimately rejected.
  • Cultivating National Champions: The report details the rise of Huawei, viewing it as China’s attempt to replicate the successful model of Samsung.

Companies/Assets Involved

  • Chinese Entities/Companies:
  • Chinese Government: Promoting semiconductor self-sufficiency through industrial policies such as “Made in China 2025.”
  • Tsinghua Unigroup: As government-affiliated capital, attempted to invest in or acquire TSMC, MediaTek, and Lattice Semiconductor.
  • Huawei: Described as a representative of China’s cultivated tech companies aiming for global success.
  • US Companies:
  • IBM: Opened technology in exchange for China market access; the case demonstrates a cooperative strategy.
  • Qualcomm, AMD: Mentioned as having similar technology cooperation or technology-for-market transactions.
  • Lattice Semiconductors: Became the target of a circuitous acquisition by Chinese capital; the deal was ultimately blocked.
  • Taiwan Companies:
  • TSMC: A key target for Chinese capital seeking to acquire a stake; the majority of its chips are assembled in China, indicating high market dependency.
  • MediaTek: Also a target of Chinese capital’s investment attempts, facing a trade-off between market and capital sources.

Investment Implications

For investors, this implies the following:

1. Geopolitical risk is an essential dimension for evaluating the semiconductor supply chain: Any company with significant market exposure or manufacturing dependence in China (e.g., TSMC, MediaTek) must incorporate a geopolitical friction premium into its valuation.

2. Beware of long-term value erosion in “technology transfer” deals: Western companies engaging in technology cooperation with China in exchange for market access (e.g., IBM in the historical case) may face long-term depreciation of their core intellectual property; the pros and cons of such transactions need careful assessment.

3. Focus on protected strategic assets: Targets that the US and its allies have clearly blocked from acquisition (e.g., the Lattice Semiconductor case), as well as indigenous champion companies (e.g., leading US semiconductor firms in various subsegments), may benefit from policy protection and thus offer more certain strategic value.

4. Recognize the long-term nature of the catch-up: Despite China’s significant investment, a shift in semiconductor leadership is far from easy. In the foreseeable future, the moats of current leaders—especially in advanced manufacturing—remain deep, and the stability of key supply chain nodes (such as Taiwan) remains the Achilles’ heel of the global technology industry.


Theme and Background

This chapter focuses on the geopolitical game where, after the Trump administration took office in 2016, U.S. semiconductor policy toward China turned aggressive. The background is that globalization did not lead to the diffusion of chip manufacturing technology; instead, it created a "Taiwanized" landscape dominated by a few irreplaceable companies (e.g., TSMC), a situation the U.S. policy had previously failed to fully recognize.

Core Argument

The author's core thesis is that the Trump administration's aggressive restrictive measures (e.g., export controls targeting Huawei) successfully slowed China's progress in the semiconductor sector and forced China to recognize that establishing a fully autonomous or fully non-U.S. supply chain is unrealistic in the foreseeable future. A counterintuitive judgment is that despite harsh rhetoric, China has not truly pursued a fully self-sufficient supply chain but rather seeks to reduce dependence on the U.S. in specific areas.

Key Arguments and Data

The author supports the argument with concrete cases and data:

1. The U.S. Government's Tough Actions: Three specific measures are listed:

  • ZTE: Restricted from purchasing U.S. goods for selling equipment to Iran and North Korea, and fined nearly $1 billion.
  • Fujian Jinhua: Accused of stealing Micron Technology's trade secrets, cut off from purchasing necessary equipment from U.S. and Japanese suppliers.
  • Huawei: In May 2020, the U.S. Commerce Department escalated rules, prohibiting not only the sale of U.S. products to Huawei but also any goods made with U.S. technology, effectively severing Huawei's ties with the global chip manufacturing infrastructure.

2. The Absolute Height of Technological Barriers: Taking ASML's EUV lithography machine as an example, merely replicating its laser requires perfectly identifying and assembling 457,329 parts. Even if China invests hundreds of billions of dollars and over a decade to duplicate existing technology, ASML might already have released the next-generation "High-NA EUV" technology.

3. The Potential Impact of Cost and Innovation Slowdown: The author points out that the laws of physics will eventually limit further transistor miniaturization, and the rate of decline in manufacturing costs has already slowed significantly. If Moore's Law ultimately loses steam, it may allow latecomers (like China) to believe they can catch up through the window of slower innovation.

Companies/Assets Involved

Company/Entity Role and Key Data Author's Implied Stance/Impact
ZTE Chinese telecom equipment maker sanctioned by U.S., fined nearly $1 billion. As an early case of U.S. sanctions policy, demonstrates the determination of U.S. actions.
Fujian Jinhua Chinese DRAM maker, "strangled" after being accused of stealing Micron technology. Case shows U.S. capability to block specific Chinese companies' development through technology restrictions.
Huawei Primary sanction target, cut off from global chip manufacturing infrastructure. Core case, proving the "long-arm jurisdiction" effect of U.S. export controls and its impact on leading Chinese tech companies.
ASML Global leader in EUV lithography machine supply, its technology extremely complex (laser contains 457,329 parts). Symbol of technological barriers, illustrating the extreme difficulty for China to catch up in cutting-edge equipment.
TSMC Global leader in chip manufacturing, located in geopolitically focal Taiwan. Viewed as a key strategic asset and potential flashpoint for conflict; its security is critical to global supply chain stability.
Micron U.S. memory chip company; its trade secret dispute was one pretext for sanctions on Fujian Jinhua. Representing U.S. technological advantage, its IP serves as a legal basis for sanctions.

Investment Implications

For investors, this means:

1. Geopolitical Risk Premium: The risk of geopolitical disruption to the semiconductor supply chain (especially advanced processes reliant on Taiwan)—a "black swan"—must be factored into investment evaluations, as such a risk could trigger a global recession.

2. Value of Technology Moat: Companies with extremely complex, hard-to-replicate technologies (e.g., ASML's EUV) have their barriers strengthened by technology competition between nations, warranting long-term attention.

3. Beneficiaries of Sanctions and Countermeasures: Investors should monitor alternative hardware and semiconductor companies (non-U.S. or Chinese domestic supply chain firms) that may benefit from U.S.-China tech decoupling or China's push for "de-Americanization" of its supply chain.

4. Impact of Innovation Pace: If the slowdown of Moore's Law becomes reality, it will alter industry competitive dynamics and may provide opportunities for laggards, requiring a reassessment of long-term prospects for industry leaders and challengers.


Theme and Background

This section explores the specific investment implications and potential opportunities that investors should focus on amid the evolving landscape of geopolitical tensions (the "chip war"). The report incorporates recent developments to analyze how policy shifts and technological breakthroughs are reshaping the competitive dynamics in the semiconductor and related hardware sectors.

Core Thesis

The author's central argument is that geopolitical factors are becoming a key variable driving the value of specific hardware and semiconductor companies, potentially creating non-market, structural investment opportunities. A counterintuitive judgment is that the leverage of Western countries to contain China's technology may be weaker than market consensus expects, as evidenced by Huawei's technological breakthroughs.

Key Arguments and Data

The report supports its thesis with specific company and policy cases:

1. Cellular Module Market: The report notes that the North American cellular module market was previously disrupted by Chinese competitors (such as Quectel) using low-price strategies. If the U.S. quickly bans Quectel on national security grounds, its roughly $500 million in annual U.S. revenue market would change hands.

2. Huawei's Technological Breakthrough: Huawei's newly launched Kirin 9000s chip demonstrates technological capabilities (at least in terms of nanometer process) that exceed expectations, only slightly lagging behind the most advanced Western technology. This weakens the leverage exerted by Western technology restrictions.

3. Policy Developments: The U.S. Federal Communications Commission (FCC) has laid the policy groundwork to quickly ban Chinese equipment posing national security threats, and the language of related bans is escalating.

4. Apple's China Risk: China has banned government employees from using Apple iPhones. If this marks the beginning of a full-phase-out of Apple, the impact would be enormous, as Apple manufactures almost all of its iPhones in China at Foxconn factories.

Companies/Assets Involved

Company Name Role & Key Data View/Tendency
Intel Seen as a strategic asset that may receive government aid to revive its glory. Potential beneficiary
ASML Holds a complete monopoly in extreme ultraviolet lithography (EUV); as long as Moore's Law continues, it will remain successful and strategically important. Bullish
Quectel Chinese cellular module manufacturer, dominates the North American market with low-price strategy, roughly $500 million in annual revenue, faces risk of being banned by the U.S. Bearish/Risk
SMTC (Sierra Wireless) After acquiring Sierra Wireless, becomes the only major remaining cellular module vendor in North America. If Quectel is banned, it could take over that $500 million market and improve margins. Potential beneficiary
Huawei Launched the Kirin 9000s chip, exceeding expectations in technological capability, suggesting its semiconductor progress may be faster than Western estimates. Geopolitical variable
Apple Faces risk of gradual phase-out by the Chinese government in its market; its supply chain is highly concentrated in China. Risk warning
Qualcomm Could be an initial loser in the context of Huawei's new chip launch. Potential loser

Investment Implications

For investors, this means the need to:

1. Identify "Strategic Assets" Protected by Geopolitics: Focus on companies like Intel that may receive state support to secure supply chains.

2. Bet on Irreplaceable Monopoly Links: Invest in companies like ASML, which occupy a critical bottleneck in technology with extremely high barriers.

3. Seek Potential Policy Substitution Beneficiaries: In sub-markets (such as cellular modules) that may be reshaped by "de-China" trends or national security reviews, look for existing substitute suppliers like SMTC.

4. Reassess the Effectiveness of Technology Restrictions on China and Related Risks: Huawei's case suggests that technological catch-up may be faster. Investors should be wary of the policy retaliation risk faced by companies (like Apple) that are overly dependent on the Chinese market or supply chain, while also recognizing that the competitive advantages of Western tech companies may be eroded.