Notes on Chip War: The Fight for the World’s Most Critical Technology

book
notes
history
My notes from the book Chip War: The Fight for the World’s Most Critical Technology by Chris Miller.
Author

Christian Mills

Published

August 28, 2024

Book Links

Cast of Characters

Semiconductor Industry Leaders

  • Morris Chang:
    • Founder of Taiwan Semiconductor Manufacturing Company (TSMC), the world’s most important chipmaker.
    • Former senior executive at Texas Instruments.
  • Andy Grove:
    • Former president and CEO of Intel (1980s-1990s).
    • Known for aggressive style and success in reviving Intel.
    • Author of “Only the Paranoid Survive.”
  • Pat Haggerty:
    • Chairman of Texas Instruments.
    • Led the company’s specialization in microelectronics, including for the U.S. military.
  • Jack Kilby:
    • Co-inventor of the Integrated Circuit (1958).
    • Long-time Texas Instruments employee.
    • Nobel Prize winner.
  • Jay Lathrop:
    • Co-inventor of photolithography (the process of patterning transistors using specialized chemicals and light).
    • Formerly of Texas Instruments.
  • Carver Mead:
    • Professor at the California Institute of Technology (Caltech).
    • Advisor to Fairchild Semiconductor and Intel.
    • Visionary thinker about the future of technology.
  • Gordon Moore:
    • Co-founder of Fairchild Semiconductor and Intel.
    • Creator of Moore’s Law (1965), which predicted that computing power on each chip would double every couple of years.
  • Robert Noyce:
    • Co-founder of Fairchild Semiconductor and Intel.
    • Co-inventor of the integrated circuit (1959).
    • Known as the “Mayor of Silicon Valley.”
    • First leader of Sematech.
  • Jerry Sanders:
    • Founder and CEO of AMD.
    • Known as Silicon Valley’s most flamboyant salesman.
    • Aggressive critic of unfair Japanese trade practices in the 1980s.
  • Charlie Sporck:
    • Drove the offshoring of chip assembly while leading manufacturing operations at Fairchild Semiconductor.
    • Later became CEO of National Semiconductor.
  • Ren Zhengfei:
    • Founder of Huawei, China’s telecom and chip design giant.
    • His daughter, Meng Wanzhou, was arrested in Canada in 2018 on charges of violating U.S. law and trying to evade U.S. sanctions.

Other Notable Figures

  • Akio Morita:
    • Co-founder of Sony.
    • Co-author of “The Japan That Can Say No.”
    • Represented Japanese business on the world stage during the 1970s and 1980s.
  • William Perry:
    • Pentagon official (1977-1981), and later Secretary of Defense (1994-1997).
    • Advocated using chips to produce precision strike weapons.

Introduction

A Critical Strait & An Obscure Regulation

  • On August 18, 2020, the USS Mustin sailed through the Taiwan Strait, a waterway increasingly fraught with geopolitical tension.
  • This show of force underscored a new reality: the battle for technological dominance, particularly in semiconductors, has become central to the rivalry between the United States and China.
  • While the U.S. Navy patrolled the strait, Chinese leaders were more concerned about a U.S. Commerce Department regulation called the Entity List, which restricts technology exports.
  • This regulation targeted Huawei, a Chinese tech giant, barring it from buying advanced computer chips made with U.S. technology.
    • This action revealed China’s dependence on foreign-made chips, the essential building blocks of all modern electronics.
    • China now spends more on importing chips than on oil.

The Significance of Semiconductors

  • Semiconductors, also known as integrated circuits or chips, are the foundation of modern technology.
    • They power everything from smartphones and refrigerators to military systems and artificial intelligence.
  • Moore’s Law: In 1965, Gordon Moore observed that the number of transistors on a chip doubles approximately every two years, leading to exponential growth in computing power.
    • This prediction has held true for over half a century, driving the digital revolution.
  • Chips have become ubiquitous and essential:
    • They are found in nearly every device that requires computing power.
    • The global economy and our daily lives are deeply reliant on a steady supply of increasingly powerful chips.

The Rise of Silicon Valley and the Global Chip Supply Chain

  • Silicon Valley became the epicenter of the semiconductor revolution due to a unique confluence of factors:
    • Scientific Expertise: Proximity to universities like Stanford and Berkeley, and access to defense funding.
    • Manufacturing Know-How: Availability of engineers experienced in aviation and radio technologies.
    • Entrepreneurial Spirit: A culture that attracted ambitious individuals from around the world.
  • The complexity of chipmaking led to a highly specialized and interconnected global supply chain:
    • Chip Design: Often involves companies in the US, UK, Japan, and Israel.
    • Raw Materials: Ultra-pure silicon wafers and gases primarily come from Japan.
    • Manufacturing Equipment: Highly specialized machines are produced by a handful of companies, primarily in the Netherlands, Japan, and California.
    • Fabrication: Companies in Taiwan, South Korea, and other Asian countries play a major role.
    • Assembly and Testing: Often done in Southeast Asia.
  • This intricate system has enabled the incredible progress predicted by Moore’s Law but has also created vulnerabilities.

Taiwan: The Geopolitical Linchpin of the Chip Industry

  • Taiwan’s TSMC (Taiwan Semiconductor Manufacturing Company) plays a critical role in the global chip supply:
    • Produces the world’s most advanced processor chips, powering devices like iPhones.
    • Responsible for fabricating a third of the world’s new computing power each year.
  • This reliance on Taiwan has become a major geopolitical concern:
    • The island is claimed by China, which considers it a renegade province.
    • Any disruption to chip production in Taiwan, whether from natural disasters, accidents, or political conflict, would have severe consequences for the global economy.

The Vulnerability of the Semiconductor Supply Chain

  • The COVID-19 pandemic exposed the fragility of the global chip supply chain:
    • Factory shutdowns, logistical bottlenecks, and fluctuating demand led to widespread chip shortages.
    • Industries like automotive manufacturing faced production halts due to the lack of essential chips.
  • This vulnerability stems from the high concentration of production and specialized expertise in a few key locations:
    • Earthquakes, political instability, or even targeted attacks could severely disrupt global chip production.

The Semiconductor Race and the Future of Global Power

  • The U.S. and China recognize the strategic importance of semiconductors:
    • Both countries are investing heavily in domestic chip production and research.
    • The outcome of this technological competition will have major implications for economic and military power.
  • Taiwan’s central role in chipmaking adds a dangerous layer to the US-China rivalry:
    • China’s ambitions towards Taiwan pose a significant threat to the global chip supply.
    • The U.S. is committed to defending Taiwan, raising the stakes of a potential conflict.
  • The complex history of the semiconductor industry, driven by technological innovation, economic efficiency, and government policies, has created a world deeply reliant on a handful of companies and vulnerable to disruption.

Conclusion: A World Defined and Imperiled by Chips

  • Semiconductors have become the foundation of modern life, shaping everything from smartphones to military systems.
  • The intricate global supply chain that produces these chips is a marvel of efficiency but also a point of vulnerability.
  • The concentration of advanced chip manufacturing in Taiwan, particularly by TSMC, has turned the island into a critical geopolitical linchpin.
  • The rivalry between the United States and China, coupled with the strategic importance of semiconductors, has created a situation of unprecedented risk.
  • Understanding the history and interconnectedness of the global chip industry is essential for navigating the challenges and opportunities of the 21st century.

Chapter 1: From Steel to Silicon

World War II: A “Typhoon of Steel”

  • Japanese soldiers described World War II as a “typhoon of steel.”

Akio Morita’s Wartime Experience

  • Akio Morita, a young engineer from a family of sake merchants, narrowly avoided front-line service in World War II due to his assignment to a Japanese Navy engineering lab.
  • Morita witnessed the devastation inflicted on Japan by American B-29 Superfortress bombers, which destroyed much of Tokyo and other urban centers.
  • The American blockade of Japan created widespread hunger and forced the country into desperate measures, such as training Morita’s brothers as kamikaze pilots.

Morris Chang’s Wartime Experience

  • Morris Chang’s childhood was marked by the sounds of gunfire and air raid sirens as Japanese armies swept across China.
  • Chang and his family became refugees, fleeing to Guangzhou, Hong Kong, Chongqing, and finally back to Shanghai after Japan’s defeat.
  • The war’s end didn’t bring peace to China, as communist guerrillas resumed their fight against the Chinese government, forcing Chang to flee to Hong Kong for a second time.

Andy Grove’s Wartime Experience

  • Andy Grove, then known as András Grof, endured multiple invasions of Budapest during World War II.
  • Hungary’s far-right government treated Jews like Grove’s family as second-class citizens. Despite this, Grove’s father was drafted into the army and sent to fight alongside the Nazis against the Soviet Union, where he went missing in action at Stalingrad.
  • In 1944, Nazi Germany invaded Hungary, their supposed ally, with plans to deport Jews like Grove to death camps.
  • Grove and his mother hid in a bomb shelter as the Red Army liberated Budapest, but the Soviet occupation brought its own share of suffering, including the rape of Grove’s mother.

America’s Industrial Might: Deciding Factor in World War II

  • World War II was a conflict of industrial attrition, a struggle that the United States, with its vast manufacturing capacity, was well-positioned to win.
  • The War Production Board in Washington measured success in terms of raw materials like copper, iron, rubber, oil, aluminum, and tin, as America transformed its industrial base to churn out military hardware.
  • The United States outproduced the Axis powers in every major category of war material, building more tanks, ships, planes, artillery, and machine guns than its enemies combined.
  • Convoys of American-made goods flowed across the Atlantic and Pacific, supplying Britain, the Soviet Union, China, and other Allied nations with vital resources.
  • While the war was fought by soldiers and sailors, the fighting power that ultimately decided the conflict was produced on American assembly lines in places like Kaiser shipyards and the River Rouge plant.

The Dawn of a New Technological Era

  • Despite the dominance of industrial output in determining the outcome of World War II, new technologies, particularly in the realm of electronics, were already transforming military power.
  • While mass-producing planes and tanks by the thousands, the great powers invested heavily in research labs that yielded innovations like rockets and radar.
  • The atomic bombs dropped on Hiroshima and Nagasaki dramatically illustrated the potential of emerging technologies to reshape warfare, sparking speculation about a future “atomic age” that would supersede the era of coal and steel.

Akio Morita’s Vision of the Future of Warfare

  • Although too young to have been deeply involved in wartime technological developments, Morris Chang and Andy Grove witnessed the shifting technological landscape. Akio Morita, in his early twenties, had a more direct view.
  • Morita’s experience developing heat-seeking missiles, although far from practical deployment, provided a glimpse into a future where wars might be won not by sheer industrial output but by sophisticated weaponry capable of autonomous targeting and maneuvering.
  • This notion, seemingly like science fiction at the time, was underpinned by Morita’s awareness of advances in electronic computation. These advances hinted at the possibility of machines performing complex calculations, essentially “thinking” by solving mathematical problems.

The Evolution of Computing

  • The concept of using devices to aid computation was not new. Humans had long relied on tools like the abacus to manipulate numbers.
  • The rise of large organizations in government and business during the late 19th and early 20th centuries created a surge in demand for “human computers” - clerks equipped with pen, paper, and mechanical calculators.
  • These human computers performed essential tasks such as processing payrolls, tracking sales, compiling census data, and analyzing statistics for insurance purposes.
  • The Works Progress Administration’s Mathematical Tables Project during the Great Depression illustrated both the potential and limitations of human-powered computation. The project employed hundreds of human computers to generate volumes of tables for complex mathematical functions, a valuable but time-consuming and laborious endeavor.

The Need for More Powerful Computing

  • The demand for faster and more efficient computation continued to grow, particularly in military applications.
  • Mechanical bombsights, for instance, were developed to improve bombing accuracy, but their limited input-output capabilities and reliance on precise conditions restricted their effectiveness.
  • Achieving greater accuracy in targeting and other military applications necessitated more sophisticated calculations, pushing the development of more powerful computing devices.

From Mechanical Gears to Vacuum Tubes: A New Era of Computing

  • The limitations of mechanical computers led to the exploration of electronic alternatives.
  • Early electronic computers employed vacuum tubes, glass-enclosed filaments that could be switched on and off to represent the 1s and 0s of binary code.
  • Binary counting, using only 1s and 0s, could theoretically represent any number, enabling a wide range of computations.
  • Unlike the fixed functionality of mechanical computers, vacuum tube-based computers were reprogrammable, allowing for greater versatility.
  • Despite the advantages of vacuum tubes, they came with significant drawbacks:
    • Attracted insects: The heat generated by the tubes made them attractive to moths and other insects, often leading to malfunctions.
    • Burned out frequently: Vacuum tubes were prone to burning out, just like lightbulbs, requiring constant replacement.

ENIAC: A Powerful but Cumbersome Computer

  • ENIAC, built for the U.S. Army in 1945, exemplified the capabilities and limitations of early vacuum tube computers.
  • ENIAC could perform hundreds of multiplications per second but occupied an entire room due to its 18,000 fist-sized vacuum tubes.
  • The high failure rate of the tubes meant frequent breakdowns, halting operations and requiring technicians to locate and replace faulty components.

The Search for a Better Alternative

  • ENIAC demonstrated the potential computing power of vacuum tubes but also highlighted their impracticality for widespread use.
  • The search for a smaller, faster, more reliable, and less cumbersome alternative to the vacuum tube became paramount in advancing the field of computing.

Chapter 2: The Switch

The Quest for a Better Switch

  • William Shockley, a brilliant but obnoxious physicist at Bell Labs, believed semiconductors held the key to a better switch.

    • Semiconductors: Unique materials like silicon and germanium that can conduct electricity under specific conditions, unlike conductors (e.g., copper) or insulators (e.g., glass).

Shockley’s Solid-State Valve Theory

  • In 1945, Shockley theorized a solid-state valve using a piece of silicon and a 90-volt battery.
    • He hypothesized that an electric field could attract free electrons in the silicon, making its edge conductive and allowing current to flow.
    • His initial experiment failed to produce measurable results due to the limitations of instruments at the time.

Brattain and Bardeen’s Breakthrough

  • Two of Shockley’s colleagues at Bell Labs, Walter Brattain and John Bardeen, built a device with two gold filaments touching a germanium block.
    • On December 16, 1947, they successfully controlled the current flow across the germanium, proving Shockley’s theories.
    • This device was later named the transistor.

Transistor’s Initial Applications and Shockley’s Response

  • AT&T, Bell Labs’ parent company, saw the transistor’s potential for amplifying signals in telephones.
  • Transistors were also recognized as replacements for vacuum tubes in hearing aids and radios.
  • Shockley, furious at being outdone, locked himself in a Chicago hotel room and conceived of a new three-layer transistor design.
    • This design amplified current and functioned as a switch by manipulating a small current to control a larger one.

Public Announcement and Significance

  • Bell Labs announced the transistor in June 1948, but the news was met with little fanfare.
    • “Little brain cell”, Time magazine, 1948.
  • The significance of these “wired blocks of germanium” as replacements for human brains in computing was unimaginable at the time.

Chapter 3: Noyce, Kilby, and the Integrated Circuit

The Need for Mass Production

  • While transistors had been invented and showed promise in replacing vacuum tubes, the challenge of mass production and commercial viability remained.
  • John Bardeen and Walter Brattain, inventors of the transistor, were primarily researchers and less interested in the business aspects of their invention.
  • William Shockley, their colleague, was ambitious and sought both fame and fortune.
    • In 1955, he founded Shockley Semiconductor in Mountain View, California, aiming to capitalize on the potential of transistors.

Shockley Semiconductor and the Transistor Market

  • AT&T, holding the transistor patent, offered licenses for $25,000, enabling companies like Shockley’s to enter the market.
  • The potential of the transistor market was uncertain, as it was unclear whether transistors could surpass vacuum tubes in performance or cost.
  • Shockley, despite winning a Nobel Prize for semiconductor theory, faced the engineering challenge of making transistors practical.
  • Transistors began replacing vacuum tubes in computers, but the complex wiring between thousands of transistors posed a significant hurdle.

Jack Kilby’s Integrated Circuit

  • Jack Kilby, an engineer at Texas Instruments (TI), dedicated himself to simplifying the complex wiring required for transistors.
  • Kilby’s Background:
    • Known for his quiet brilliance, collaborative spirit, and problem-solving approach.
    • One of the first engineers outside Bell Labs to work with transistors at Central Lab in Milwaukee.
    • Joined TI’s transistor unit in 1958.
  • TI’s Background:
    • Originally focused on seismic equipment for oil exploration.
    • Gained electronics expertise during World War II producing sonar devices for the Navy.
    • Hired engineers like Kilby to develop military electronics after the war.
  • Kilby’s Breakthrough (Summer 1958):
    • While working alone at TI during a holiday period, Kilby focused on reducing the number of wires connecting transistors.
    • His idea was to build multiple components on a single piece of semiconductor material instead of using separate pieces for each transistor.
  • Kilby’s invention, the “integrated circuit,” later became known as a “chip” because it was made from a piece of silicon chipped off a circular wafer.

The Rise of Fairchild Semiconductor

  • In Palo Alto, eight engineers, including Robert Noyce, left Shockley Semiconductor due to Shockley’s management style.
  • The “Traitorous Eight”
    • This group, seeking a better working environment, founded Fairchild Semiconductor with funding from an East Coast investor.
    • Their departure is considered a pivotal moment in the formation of Silicon Valley.
    • Eugene Kleiner, one of the eight, later founded the venture capital firm Kleiner Perkins.
    • Gordon Moore, who led Fairchild’s R&D, later formulated “Moore’s Law,” predicting the exponential growth of computing power.
  • Robert Noyce, the group’s leader, possessed a strong vision for microelectronics and a keen understanding of the technical advancements needed to make transistors smaller, cheaper, and more reliable.

Noyce’s Planar Process and the Advancement of the Integrated Circuit

  • While the science of transistors was established, reliable manufacturing was a challenge.
  • Early transistors, using a mesa structure, were prone to impurities affecting their performance.
  • Jean Ernie, Noyce’s colleague at Fairchild, developed a planar process:
    • This method involved building transistors into, rather than on top of, the germanium, and depositing a protective layer of silicon dioxide to prevent impurities.
    • Ernie’s innovation significantly improved transistor reliability.
  • Noyce’s Refinement:
    • Noyce realized Ernie’s planar process could be used to create multiple transistors on a single piece of silicon.
    • Unlike Kilby’s integrated circuit, which used wires, Noyce’s design embedded the transistors and wires within the silicon, further reducing size and complexity.
  • Both Kilby and Noyce independently invented the integrated circuit, but Noyce’s planar process proved to be a significant leap forward in miniaturization and reliability.

Fairchild’s Vision for the Future

  • Noyce, Moore, and the Fairchild team recognized the advantages of their integrated circuits:
    • Reliability: Superior to the complex wiring of other electronic devices.
    • Miniaturization: Planar design facilitated smaller transistors.
    • Electric Efficiency: Smaller circuits consumed less power.
  • They envisioned new applications for their invention, driven by miniaturization and reduced power consumption.
  • The initial cost of Noyce’s integrated circuit was 50 times higher than simpler devices, posing a challenge to market adoption.
    • Despite its ingenuity, the invention needed a market to thrive.

Chapter 4: Liftoff

Sputnik and the Space Race

  • Three days after Fairchild Semiconductor’s founding, the Soviet Union launched Sputnik, the world’s first satellite, on October 4, 1957.
    • This event sparked fear in the United States about falling behind in the space race and the military implications of Soviet technological superiority.
    • Four years later, the Soviet Union also sent the first human, Yuri Gagarin, into space.

The Impact of the Space Race on the U.S. Semiconductor Industry

  • The U.S. government launched a crash program to catch up in the space race, leading to increased funding for rocket and missile programs.
  • President John F. Kennedy declared the goal of sending a man to the moon, further fueling investment in aerospace technology.
  • This created a new market for integrated circuits, with NASA emerging as a major customer.

The Apollo Guidance Computer

The Challenge

  • NASA tasked the MIT Instrumentation Lab with designing the guidance computer for the Apollo spacecraft.
    • This computer would need to be extremely complex, reliable, and lightweight.
    • Initial estimates suggested that a computer powerful enough to guide a spacecraft to the moon would be prohibitively large and power-hungry.

The Solution: Integrated Circuits

  • Transistor-based computers were significantly more advanced than the older vacuum tube computers, but their size and power consumption remained a challenge.
  • MIT engineers had experimented with early integrated circuits from Texas Instruments, purchased in 1959 for $1,000 for 64 chips, for a U.S. Navy missile program.
  • By 1962, Fairchild began marketing its own micrologic chips, attracting the attention of MIT engineers working on the Apollo guidance computer.

Fairchild’s Success

  • Despite being a new company, Fairchild gained the trust of MIT engineers due to the reliability and timely delivery of its chips.
  • Charles Stark Draper, head of the MIT lab, decided to use Fairchild chips for the Apollo guidance computer.
    • Draper calculated that a computer using Fairchild’s integrated circuits would be significantly smaller, lighter, and more energy-efficient than one based on discrete transistors.
  • The final Apollo guidance computer weighed 70 pounds and occupied about one cubic foot of space, a dramatic reduction in size compared to earlier computers.
  • By 1964, Noyce boasted that the integrated circuits in Apollo computers had operated for 19 million hours with only two failures.

The Impact of the Apollo Program on Fairchild

  • The Apollo program transformed Fairchild Semiconductor from a small startup to a company with 1,000 employees.
  • Sales skyrocketed from $500,000 in 1958 to $21 million in 1960.
  • Noyce leveraged increased production for NASA to lower prices for other customers, further expanding the market for integrated circuits.

NASA’s Endorsement

  • NASA’s successful use of integrated circuits in the Apollo program provided a critical stamp of approval for the technology.
  • This demonstrated the reliability and capabilities of integrated circuits in the most demanding environments.

Texas Instruments and the Minuteman II Missile

Pat Haggerty’s Vision

  • Pat Haggerty, president of Texas Instruments, recognized the potential of Jack Kilby’s integrated circuit for military applications.
  • Haggerty, a skilled salesman and visionary leader, believed integrated circuits could be used in all types of military electronics.
  • He actively promoted integrated circuits to the Defense Department and secured funding for research and development.

The Minuteman II Missile Contract

  • In 1962, the Air Force sought a new computer for its Minuteman II missile, aiming to improve its range and accuracy.
  • Haggerty proposed using integrated circuits to create a smaller, lighter, and more powerful guidance computer.
    • His design relied on 22 different types of integrated circuits to handle 95% of the computer’s functions, resulting in significant weight reduction.
  • TI won the contract, marking a turning point for the company’s integrated circuit business.

The Impact of the Minuteman II Program on Texas Instruments

  • The Minuteman II contract led to a surge in demand for TI’s integrated circuits, driving sales from dozens to thousands.
  • By the end of 1964, TI had supplied 100,000 integrated circuits for the Minuteman program.

The Challenge of Mass Production

  • Both Fairchild and TI faced the challenge of mass-producing integrated circuits to meet the growing demand from military and civilian customers.
  • This required developing new manufacturing techniques, improving reliability, and reducing costs.

Chapter 5: Mortars and Mass Production

Jay Lathrop and Photolithography

The Need for Miniaturization

  • Jay Lathrop, an MIT graduate working at a U.S. government lab, was tasked with developing a smaller and more efficient proximity fuse for 81-millimeter mortar shells.
  • Like Fairchild’s engineers, Lathrop struggled with the limitations of mesa-shaped transistors, which were difficult to miniaturize.

A Breakthrough: Printing with Light

  • While examining transistors under a microscope, Lathrop and chemist James Knoll devised a revolutionary technique called photolithography.
    • This process used light to create microscopic patterns on semiconductor material, enabling the production of much smaller transistors with greater precision.

Photolithography Process

  1. Coat: Cover a block of germanium with a photoresist chemical that reacts to light.
  2. Expose: Shine light through a mask containing the desired pattern onto the photoresist-coated germanium.
    • The light passes through the mask’s open areas and shrinks as it passes through an upside-down microscope lens.
  3. Develop: Wash away the exposed photoresist, leaving behind the patterned germanium.
  4. Connect: Add a thin layer of aluminum to create electrical connections.

Advantages of Photolithography

  • Enabled the production of transistors significantly smaller than previously possible.
  • Allowed for greater precision and accuracy in creating transistor features.
  • Opened the door to mass production of transistors and integrated circuits.

Texas Instruments Adopts Photolithography

Recognizing the Potential

  • Pat Haggerty and Jack Kilby recognized the importance of photolithography for mass-producing integrated circuits.
  • They understood that this technique could mechanize and miniaturize chipmaking, making it possible to meet the growing demand from the military and other customers.

Implementing Photolithography at TI

  • Implementing photolithography at TI required significant investment in new materials and processes:
    • Chemicals: TI developed its own high-purity photoresist chemicals.
    • Masks: TI created its own masks for projecting precise patterns of light.
    • Silicon Wafers: TI began producing its own ultra-pure silicon wafers.

Overcoming Production Challenges

  • Mass production of integrated circuits presented numerous challenges due to the microscopic scale and sensitivity of the process:
    • Impurities: Even tiny impurities in chemicals or dust particles could ruin an entire production run.
    • Process Variations: Fluctuations in temperature, pressure, and other factors could affect the quality and reliability of chips.

Trial and Error and Human Computers

  • TI relied heavily on trial and error to optimize its production processes.
    • Engineers conducted thousands of experiments to test different combinations of materials, temperatures, and process parameters.
    • Mary Ann Potter, a TI production engineer, played a crucial role in scaling up chip production for the Minuteman missile.
      • She spent countless hours running experiments, analyzing data, and refining production processes.

Morris Chang and Yield Improvement

From Refugee to Semiconductor Engineer

  • Morris Chang, a Chinese refugee who fled to the U.S., joined TI in 1958.
  • Chang, a brilliant engineer and strategist, was tasked with improving manufacturing yield—the percentage of usable transistors produced.

Systematic Approach to Manufacturing

  • Chang took a methodical approach to improving yield, systematically testing and adjusting production processes:
    • He meticulously tweaked temperatures, pressures, and chemical combinations to find the optimal parameters.
    • He applied his deep understanding of semiconductor physics to solve production problems.

Results and Recognition

  • Chang’s efforts dramatically increased yield on his production line, attracting the attention of IBM and establishing him as a leading figure in semiconductor manufacturing.

Fairchild’s Pursuit of Mass Production

Recognizing Photolithography’s Potential

  • Bob Noyce recognized the transformative potential of photolithography for Fairchild’s future.

Hiring Key Talent

  • Fairchild hired James Knoll, co-inventor of photolithography, to develop and implement the process at the company.

Andy Grove and Process Improvement

  • Andy Grove, a Hungarian refugee with a Ph.D. in chemical engineering, joined Fairchild in 1963 and played a pivotal role in improving manufacturing processes.

The Importance of Engineering and Production

  • The development of the chip industry highlights the crucial role of engineering and production expertise, alongside scientific breakthroughs.
  • While academic research provided the foundation, it was the ingenuity and persistence of engineers like Lathrop, Chang, Grove, and countless others that made mass production possible.

Chapter 6: I Want to Get Rich

Beyond Military and Space: Targeting the Civilian Market

Early Dependence on Military and Space Programs

  • In the early 1960s, the majority of Fairchild’s revenue came from military and space programs.
    • In 1965, these sectors accounted for 72% of all integrated circuit sales.

Noyce’s Vision for a Civilian Market

  • Bob Noyce recognized the potential of integrated circuits to revolutionize civilian products and create a much larger market.
  • He believed that many of the features demanded by the military, such as miniaturization and ruggedness, would also be valuable in consumer products.

Maintaining R&D Independence

  • Noyce deliberately limited Fairchild’s reliance on military research contracts to maintain control over the company’s R&D priorities.
  • He wanted Fairchild to focus on developing innovative products for the civilian market, rather than being bound to the specific requirements of military contracts.

Moore’s Law and the Exponential Growth of Computing Power

Gordon Moore’s Prediction

  • In 1965, Gordon Moore, co-founder of Fairchild, published an article predicting that the number of components on an integrated circuit would double every year for the next decade.
    • This prediction, which came to be known as Moore’s Law, proved remarkably accurate and had profound implications for the future of computing.

Implications of Moore’s Law

  • Exponential Growth in Computing Power: Moore’s Law implied that computing power would increase exponentially over time.
  • Decreasing Costs: As transistors became smaller and more densely packed, the cost per transistor would decrease, making computers more affordable and accessible.
  • Expanding Applications: As computing power became cheaper and more readily available, it would find its way into a wider range of products and applications.

Fairchild’s Strategic Shift

The “McNamara Depression” and Its Impact

  • In the early 1960s, U.S. Defense Secretary Robert McNamara implemented cost-cutting measures that reduced military spending on electronics, impacting companies like Fairchild.
  • This “McNamara Depression” reinforced Noyce’s belief in the importance of developing civilian markets for integrated circuits.

Focus on Civilian Products

  • Fairchild became the first company to offer a full line of off-the-shelf integrated circuits designed specifically for civilian applications.

Price Cuts to Drive Adoption

  • Noyce aggressively cut prices for Fairchild’s integrated circuits, even selling some products below cost, to encourage wider adoption and stimulate demand.

The Rise of the Civilian Computer Market

Rapid Growth of the Computer Industry

  • The computer industry experienced explosive growth in the 1960s, with annual sales increasing from 1,000 units in 1957 to 18,700 units in 1967.

Fairchild’s Dominance

  • By the mid-1960s, nearly all new computers used integrated circuits, and Fairchild captured a dominant share of this rapidly expanding market.

The Burroughs Contract

  • In 1966, Fairchild secured a major contract with Burroughs, a leading computer manufacturer, for 20 million integrated circuits.
  • This contract, far larger than any previous order from NASA or the military, signaled the shift in demand towards the civilian market.

The Allure of Wealth and the Exodus from Fairchild

Financial Incentives Drive Innovation

  • The potential for financial gain became a powerful motivator for engineers and entrepreneurs in Silicon Valley.

Fairchild’s Restrictive Stock Option Policy

  • Fairchild’s owner, Sherman Fairchild, refused to grant stock options to employees, limiting their ability to share in the company’s success.

The Search for Riches

  • As Fairchild’s success attracted competition, many talented engineers and executives left to start their own companies or join rivals, enticed by the promise of greater financial rewards.
  • This exodus of talent led to the formation of numerous new semiconductor companies, further accelerating innovation and competition in the industry.

Chapter 7: Soviet Silicon Valley

The Arrival of Soviet Engineers in Silicon Valley

  • Anatoly Trutko, a Soviet semiconductor engineer, arrived at Stanford University in 1959 as part of a student exchange program.
    • This was surprising given US fears of the USSR catching up in science and technology.
    • Trutko studied advanced technology, even attending lectures by William Shockley.
      • After one class, Trutko asked Shockley to sign his book, Electrons and Holes in Semiconductors.
      • Shockley complained to Trutko that the USSR refused to pay royalties for the book’s Russian translation.
  • The USSR understood the value of semiconductors.
    • They translated Shockley’s textbook into Russian just two years after it was published.
    • US spies were ordered to acquire Soviet semiconductor devices as early as 1956 to assess their quality.
      • A 1959 CIA report found that America was only two to four years ahead of the Soviets in transistor quality and quantity.
  • At least several early Soviet exchange students were later revealed to be KGB agents.
    • This highlights the link between student exchanges and Soviet defense industry goals.
  • Both the Pentagon and the Kremlin understood that transistors and integrated circuits would revolutionize manufacturing, computing, and military power.
    • Starting in the late 1950s, the USSR established new semiconductor facilities and assigned top scientists to this industry.

Yuri Osokyan and the Rise of Soviet Semiconductor Research

  • Yuri Osokyan was a talented young Soviet engineer tasked with advancing Soviet semiconductor technology.
    • He spent his childhood in China, where his father worked in a Soviet military hospital.
    • He was known for his exceptional memory and expertise in semiconductors.
  • Osokyan was assigned to a semiconductor plant in Riga, Latvia.
    • The plant was staffed by top graduates and tasked with building devices for the Soviet space program and military.
  • Osokyan was tasked with building a circuit with multiple components on a single piece of germanium, something never done before in the Soviet Union.
    • He created his prototype integrated circuit in 1962.
    • Osokyan and his colleagues were aware they were at the forefront of Soviet science.
    • They were dedicated to their work, spending evenings discussing solid-state physics.

Khrushchev’s Vision for Soviet Semiconductor Dominance

  • Soviet leader Nikita Khrushchev aimed to surpass the United States in all areas, including technology.
  • Alexander Shokhin, First Deputy Chairman of the Soviet State Committee on Radio-Electronics, saw an opportunity to leverage Khrushchev’s ambition.
    • He told Khrushchev, “Imagine, Nikita Sergeyevich, that a TV can be made the size of a cigarette box,” highlighting the potential of Soviet silicon.
  • Shokhin believed catching up to and surpassing the United States was possible.
  • Similar to their approach with nuclear weapons, the USSR had a secret weapon: espionage.

Joel Barr and Alfred Sarant: From American Spies to Soviet Computer Pioneers

  • Joel Barr and Alfred Sarant were American electrical engineers who became Soviet spies.
    • They were childhood friends, both sons of Russian Jewish immigrants, and active in the Young Communist League.
    • They were recruited into a KGB espionage ring led by Julius Rosenberg.
  • During the 1940s, Barr and Sarant worked on classified radar and military systems at leading American tech firms.
    • They gained knowledge about electronics in new weapon systems.
  • As the FBI closed in on the Rosenberg ring, Barr and Sarant fled to the Soviet Union.
    • They told the KGB they wanted to build advanced computers.
    • Though not computer experts, their spy status gave them access to resources.
  • In the late 1950s, Barr and Sarant began building the UMM computer (“mind” in Russian).
    • Their work got the attention of Shokhin.

Zelenograd: The Dream of a Soviet Silicon Valley

  • Barr, Sarant, and Shokhin partnered to convince Khrushchev to create a city dedicated to semiconductor production.
    • They envisioned a city with its own researchers, engineers, labs, and production facilities.
  • To persuade Khrushchev, Shokhin arranged for him to visit Leningrad’s Special Design Bureau of the Electronics Industry No. 2.
    • During the visit, Sarant and Barr showcased Soviet microelectronics achievements, including a miniature radio and a basic computer.
    • They argued that semiconductors would be crucial for spacecraft, industry, government, aircraft, and even the nuclear missile shield.
  • They presented Khrushchev with plans for a futuristic semiconductor city centered around a 52-story skyscraper.
    • Khrushchev loved grand projects and approved the idea.
  • The Soviet government soon greenlit the construction of Zelenograd (“green city” in Russian) on Moscow’s outskirts.
    • Shokhin envisioned a self-sufficient scientific paradise with research labs, production facilities, schools, daycare, entertainment, libraries, and a hospital.
    • The Moscow Institute of Electronic Technology was built in Zelenograd, modeled after English and American universities.

Chapter 8: Copy It

The “Copy It” Strategy and Its Limitations

  • Around the same time Zelenograd was approved, Soviet student Boris Malin returned from Pennsylvania with a Texas Instruments SN-51, one of the first commercially available integrated circuits.
    • Alexander Shokhin, head of Soviet microelectronics, saw it as vital to acquire.
  • Shokhin ordered a group of engineers, including Malin, to “Copy it, one for one, without any deviations” within three months.
    • Soviet scientists were angered, believing they were capable of independent innovation.
  • While the USSR had leading theoretical physicists, they lagged in practical application and manufacturing, especially in high-volume chipmaking.
    • They excelled in quantity, but not quality or purity, crucial for chipmaking.
  • COCOM restrictions also hindered the USSR’s access to advanced technologies, including semiconductor components.
    • The Soviets used shell companies to bypass COCOM, but this was difficult to do on a large scale.

The Limitations of Espionage in Semiconductor Production

  • Simply stealing a chip didn’t reveal how it was made.
    • The chipmaking process required specialized knowledge, often unwritten, about temperatures, chemicals, and exposure times.
    • Soviet spies couldn’t easily steal this type of knowledge.
  • Additionally, the rapid pace of innovation, driven by Moore’s Law, meant stolen designs quickly became obsolete.
    • American companies introduced new, more powerful chips annually.

The Structural Flaws of the Soviet Semiconductor Industry

  • The Soviet leadership didn’t grasp how the “copy it” strategy ensured their technological backwardness.
  • The entire Soviet semiconductor industry operated like a secretive, top-down defense contractor.
    • There was little room for creativity or market-driven innovation.
    • Civilian products were a low priority.
  • This system stifled innovation and condemned them to follow the US lead, even in such a sensitive industry.
  • Zelenograd, despite its resources and stolen secrets, could not replicate Silicon Valley’s success.
    • Soviet engineers lacked the practical, on-the-factory-floor experience of their American counterparts.
    • Career advancement in the USSR favored bureaucracy over innovation.

Chapter 9: The Transistor Salesman

Japan’s Post-War Economic Transformation

  • In 1962, Japanese Prime Minister Hayato Ikeda gifted French President Charles de Gaulle a Sony transistor radio.
    • De Gaulle, dismissive of Japan as a mere “economic power,” scoffed that Ikeda behaved like a “transistor salesman.”
  • This chapter highlights how Japan’s semiconductor industry would make it wealthy and powerful.
  • Integrated circuits not only revolutionized electronics, but also connected nations, with the US at the center.
  • Unlike the USSR, Japan deliberately integrated itself into the US semiconductor industry, supported by Japanese business leaders and the US government.

The US Role in Japan’s Technological Rebirth

  • After World War II, some in the US favored dismantling Japan’s high-tech industries as punishment.
  • However, US defense officials soon adopted a policy of supporting a strong Japan as a Cold War strategy.
    • The aim was to rebuild Japan’s economy while tying it to the US-led system.
    • Making Japan a “transistor salesman” became central to this strategy.

Makoto Kikuchi and the Introduction of the Transistor to Japan

  • Makoto Kikuchi, a young physicist at the Electrotechnical Laboratory in Tokyo, was among the first in Japan to learn about the transistor.
  • He stayed informed through US scientific journals, which were otherwise unavailable in postwar Japan.
  • In 1953, Kikuchi met John Bardeen, co-inventor of the transistor, during Bardeen’s visit to Tokyo for a physics conference.

Akio Morita and Sony’s Embrace of the Transistor

  • In 1953, Akio Morita, co-founder of Sony, traveled to New York and secured a license from AT&T to produce transistors.
    • AT&T believed the transistor’s use was limited to hearing aids.
  • Morita, however, recognized the transistor’s potential to transform consumer electronics.
    • He aimed to sell these devices not just in Japan, but also in the US, the world’s biggest consumer market.
  • Japan’s government, through MITI (Ministry of International Trade and Industry), supported the growth of its electronics firms.
    • However, their bureaucracy sometimes clashed with the private sector.

Sony’s Innovation and the Rise of Japanese Electronics

  • Unlike the USSR’s “copy it” approach, Sony focused on innovation, product design, and marketing.
  • They licensed transistor technology but excelled at identifying and targeting new markets with desirable products.
    • “Our plan is to lead the public with new products rather than ask them what kind of products they want,” Morita declared. “The public does not know what is possible, but we do.”
  • Sony’s transistor radios were a major success.
  • Japanese firms, leveraging lower wages and US technology, became major players in consumer electronics.

A Symbiotic Relationship

  • The semiconductor industry fostered a complex interdependence between the US and Japan.
  • By 1964, Japan led in discrete transistor production, while US firms made the most advanced chips.
  • The US built the best computers, while Japan excelled in consumer electronics, driving semiconductor demand.
  • Despite some US industry concerns about Japanese competition, the US government understood the strategic importance of a strong Japanese electronics sector.
  • This semiconductor symbiosis benefited both countries:
    • Japan gained access to technology and economic growth.
    • The US strengthened a key Cold War ally and expanded its economic influence in Asia.
  • By the end of the 1970s:
    • Japan’s electronics exports boomed.
    • US-Japan interdependence deepened.
    • Japan emerged as a global economic power.

Chapter 10: Transistor Girls

The Gendered Division of Labor in the Semiconductor Industry

  • This chapter explores the role of women in the semiconductor industry, particularly in assembly line work.
  • The title, “Transistor Girls,” is a play on the 1964 Australian novel of the same name, which depicted Asian women factory workers in a stereotypical and sexualized manner.
  • The chapter highlights the industry’s reliance on women’s labor, often exploited for lower wages and perceived dexterity.

Charlie Sporck and the Drive for Efficiency

  • Charlie Sporck was a production manager at Fairchild Semiconductor known for his focus on efficiency and opposition to unions.
  • Spork’s background:
    • He had an engineering degree from Cornell.
    • He previously worked at GE, where he clashed with unions over his proposed changes to the assembly line.
    • He joined Fairchild in 1959.
  • At Fairchild, Spork was determined to prevent unionization and maximize worker productivity.

Women in Semiconductor Assembly

  • The semiconductor industry in Silicon Valley relied heavily on female labor for assembly line work.
  • Reasons for hiring women:
    • Lower wages: Women were paid less than men for similar work.
    • Perceived dexterity: Managers believed women’s smaller hands were better suited for the intricate tasks of assembling and testing chips.
    • Less likely to unionize: Women were seen as less likely to join unions or demand better working conditions.
  • Women’s history in the Santa Clara Valley:
    • Many had worked in canneries and the aerospace industry, so they were familiar with factory work.
  • The 1965 Immigration Act increased the pool of foreign-born women available for these jobs.

Offshoring to Asia: Hong Kong

  • Faced with labor shortages in the US, Fairchild Semiconductor looked to Asia for cheaper labor.
  • Bob Noyce suggested exploring Hong Kong, where he had invested in a radio factory.
  • Charlie Sporck visited Hong Kong and was impressed by the low wages and high productivity of the (mostly female) workforce.
    • Wages in Hong Kong were around 25 cents an hour, one-tenth of the US average.
  • Fairchild set up an assembly facility in Hong Kong in 1963.
    • This was the beginning of the semiconductor industry’s shift to Asia.

Expanding to Other Asian Countries

  • Fairchild’s success in Hong Kong led them to expand to other Asian countries with even lower wages.
  • Spork targeted countries with:
    • Low wages.
    • A lack of strong labor unions.
  • Fairchild opened facilities in:
    • Singapore
    • Malaysia (Penang).
  • Other US chipmakers followed suit.

The Geopolitical Impact of Offshoring

  • The offshoring of semiconductor assembly had significant geopolitical implications:
    • It strengthened economic ties between the US and Asian countries during the Cold War.
    • It provided jobs and economic development in these countries, potentially mitigating the appeal of communism.
    • It created a globalized supply chain that would shape the future of the electronics industry.

Chapter 11: Precision Strike

The Need for Precision in Warfare

  • This chapter examines how the semiconductor revolution transformed military technology, particularly the development of precision-guided munitions.
  • During the Vietnam War, the US military relied heavily on unguided bombs, which were largely ineffective.
    • Operation Rolling Thunder (1965-1968) dropped a massive amount of bombs but had limited impact.
  • The Shrike missile, which homed in on enemy radar, was somewhat successful, but many other guidance systems were unreliable.
    • A 1985 study revealed the low success rate of air-to-air missiles beyond visual range.
  • The military realized it needed to improve the accuracy of its weapons.

Weldon Word and the Development of Laser-Guided Bombs

  • Weldon Word, a project engineer at Texas Instruments (TI), recognized the potential of microelectronics to improve weapons guidance.
    • He envisioned a system where sensors, missiles, and computers worked together to ensure accurate strikes.
  • Word believed the key to successful military technology was simplicity and affordability.
  • In 1965, Word met with Colonel Joe Davis at Eglin Air Force Base.
    • Davis showed Word a photo of the Thang Hoa Bridge in North Vietnam, which had withstood numerous bombing attempts.
    • This highlighted the need for more accurate weapons.

The Paveway Laser-Guided Bomb

  • Word, inspired by the Thang Hoa Bridge challenge, developed the Paveway laser-guided bomb.
  • How it worked:
    1. A laser designator illuminated the target.
    2. The bomb, equipped with a laser seeker and control surfaces (wings), followed the reflected laser light.
    3. A simple silicon chip with four quadrants detected the laser light and adjusted the wings to keep the bomb on target.
  • The Paveway was:
    • Simple in design.
    • Affordable to produce.
    • Highly accurate.

Impact and Legacy of Precision-Guided Munitions

  • The Paveway was first used in combat in 1972 and proved highly effective in destroying bridges and other targets.
  • The Vietnam War demonstrated the potential of microelectronics to revolutionize warfare.
  • The development of the Paveway marked a turning point in military technology, paving the way for more advanced precision-guided weapons.

Chapter 12: Supply Chain Statecraft

Texas Instruments Expands to Taiwan

  • This chapter explores how semiconductor production became a tool of US foreign policy, particularly in strengthening ties with Taiwan.
  • In 1968, Texas Instruments (TI) executives Mark Shepard and Morris Chang traveled to Asia to scout locations for a new chip assembly facility.
  • Their initial visit to Taiwan was fraught with cultural misunderstandings and disagreements with K.T. Lee, Taiwan’s economy minister.

Taiwan’s Strategic Vulnerabilities and Opportunities

  • Taiwan in the late 1960s:
    • Economically successful, but politically isolated.
    • Faced a growing threat from mainland China, which had developed nuclear weapons.
    • The US, Taiwan’s main ally, was reducing its global commitments after the Vietnam War.
  • K.T. Lee recognized the strategic opportunity presented by TI:
    • Attracting foreign investment could boost Taiwan’s economy and technological development.
    • Deepening economic ties with the US could enhance Taiwan’s security.

A Mutually Beneficial Partnership

  • TI also saw advantages in Taiwan:
    • Low wages.
    • A government eager to attract foreign investment.
  • Chang, a native of mainland China, was initially hesitant about Taiwan but became convinced by former classmates.
  • Despite their initial clashes, Shepard and Lee smoothed over relations.
  • TI’s board approved the Taiwan facility in July 1968, and it began production in 1969.

Semiconductors as a Tool of US Foreign Policy

  • TI’s investment was part of a broader trend of US semiconductor firms setting up assembly plants in Asia.
  • This strategy:
    • Provided low-cost labor for US companies.
    • Boosted economic growth in Asian countries.
    • Strengthened US ties with these countries during the Cold War.
  • Leaders like Singapore’s Lee Kuan Yew saw electronics exports as key to economic growth and political stability.

The Enduring Impact of Semiconductor Supply Chains

  • By the early 1980s, US semiconductor firms employed tens of thousands of workers in Asia, mainly in Korea, Taiwan, and Southeast Asia.
  • This created a complex web of interdependence, with:
    • US firms relying on Asian labor and markets.
    • Asian countries benefiting from investment, technology transfer, and jobs.
    • The US strengthening its economic and strategic position in Asia.
  • This interdependence endured even as the US reduced its military presence in the region after Vietnam.

Chapter 13: Intel’s Revolutionaries

The Birth of Intel

  • The year 1968 was marked by global upheaval, but a seemingly minor event in Silicon Valley would have a revolutionary impact: the founding of Intel.
  • Bob Noyce and Gordon Moore, frustrated with their limited control at Fairchild Semiconductor, decided to start their own company.
  • Their goal:
    • To make transistors ubiquitous and affordable, driving a revolution in computing.

The Invention of DRAM

  • One of Intel’s first major products was the Dynamic Random Access Memory (DRAM) chip.
  • Before DRAM, computers stored data using magnetic cores, which were bulky and difficult to scale down.
  • Robert Dennard at IBM developed the concept of DRAM:
    • Using a transistor and a capacitor to store a single bit of information.
    • Repeatedly charging the capacitor to maintain data storage (“dynamic”).
  • Intel’s DRAM chips were:
    • Smaller.
    • More reliable.
    • Cheaper to produce than magnetic cores.

From Memory to Microprocessors

  • Initially, Intel focused on memory chips, but Bob Noyce, always seeking new challenges, agreed to develop a custom chip for a Japanese calculator company, Busicom.
  • Ted Hoff, an Intel engineer with a background in computer architecture, was assigned to the project.
  • Hoff realized that:
    • Designing custom chips for each device was expensive and time-consuming.
    • Intel’s increasingly powerful memory chips enabled computers to handle more complex software.

The Intel 4004: The World’s First Microprocessor

  • Hoff proposed developing a general-purpose logic chip that could be programmed with software for various tasks.
    • This would be more efficient and cost-effective than custom designs.
  • The result was the Intel 4004, marketed as the world’s first microprocessor.
    • It could be used in a wide range of devices, leading to a revolution in computing.

Carver Mead and the Vision of Ubiquitous Computing

  • Carver Mead, a Caltech professor and Intel consultant, understood the profound implications of Intel’s work.
    • He popularized the term “Moore’s Law.”
  • Mead predicted that microelectronics would lead to the automation of many aspects of society.
    • He envisioned tiny, inexpensive computers embedded in everyday devices.

The Revolutionaries

  • Intel’s founders and engineers saw themselves as revolutionaries.
    • They were not tearing down the existing order through violence but reshaping it through technological innovation.
  • Gordon Moore, reflecting on the societal changes driven by microelectronics, stated in 1973: “We are really the revolutionaries in the world today, not the kids with the long hair and beards who were wrecking the schools a few years ago.”

Chapter 14: The Pentagon’s Offset Strategy

William Perry and the Pentagon’s Technological Turn

  • This chapter explores how the Pentagon, facing a resurgent Soviet Union, turned to microelectronics to regain its technological edge.
  • William Perry became Undersecretary of Defense for Research and Engineering in 1977.
    • He was a Silicon Valley entrepreneur with deep knowledge of semiconductor technology.
    • He previously worked at Sylvania Electronic Defense Laboratories and founded his own defense electronics company.

The Soviet Challenge and the Need for a New Strategy

  • Perry arrived in Washington at a time when the US military faced significant challenges:
    • Defeat in Vietnam.
    • The Soviet Union’s growing military might, particularly its nuclear arsenal and conventional forces.
  • Andrew Marshall, director of the Pentagon’s Office of Net Assessment, argued that the US needed to exploit its technological advantage in computers to offset the USSR’s numerical superiority.

Precision-Guided Munitions and the “Offset Strategy”

  • Perry embraced Marshall’s vision and believed that microelectronics was the key to a new “offset strategy.”
  • He pushed for investments in:
    • Precision-guided munitions (PGMs): Using microchips to dramatically improve the accuracy of weapons.
    • Satellites: For surveillance, communication, and targeting.
    • Next-generation chips: To maintain US technological leadership.

Perry’s Vision for a High-Tech Military

  • Perry believed that miniaturized computing power would revolutionize warfare:
    • “We will be able to put computers, which only 10 years ago would have filled up this entire room, on a chip and field smart weapons at all levels,” he said in 1981.
  • He championed programs like the Tomahawk cruise missile, which relied on sophisticated onboard computers for navigation and guidance.
  • He envisioned a future of networked warfare, with sensors, weapons, and command centers sharing information in real time.

Criticism and Skepticism

  • Perry’s focus on high technology faced criticism from some quarters:
    • Some questioned the reliability and cost-effectiveness of PGMs.
    • Others doubted the Pentagon’s ability to adapt to rapidly changing technologies.
  • Perry dismissed these criticisms, arguing that his opponents underestimated the pace of innovation in microelectronics.

The Enduring Legacy of the Offset Strategy

  • Despite skepticism, Perry’s vision largely prevailed.
  • The Pentagon continued to invest heavily in microelectronics and PGMs, leading to:
    • The development of increasingly sophisticated weapons systems.
    • The transformation of US military doctrine toward precision strike and information warfare.
  • The “offset strategy” helped to maintain US military superiority even as the Soviet Union collapsed.

Conclusion

  • The semiconductor revolution had profound geopolitical implications:
    • It transformed the nature of warfare.
    • It strengthened US military power.
    • It deepened US economic and strategic ties with key Asian countries.
  • By the end of the Cold War, the US, thanks in large part to its dominance in semiconductors, had emerged as the world’s sole superpower.

Chapter 15: That Competition is Tough

The Semiconductor Industry Crisis

  • Richard Anderson, a Hewlett-Packard executive, found himself at the center of the semiconductor industry crisis in the 1980s.
  • Silicon Valley, once dominant, faced fierce competition from Japan.
  • Hewlett-Packard (HP)
    • One of America’s largest tech companies and a major semiconductor buyer.
    • Founded in the 1930s by Stanford graduates Dave Packard and Bill Hewlett.
    • Anderson’s chip assessments held significant weight, influencing purchasing decisions.

Japanese Semiconductor Companies

  • Toshiba and NEC, Japanese companies, entered the DRAM memory chip market.
  • Initially underestimated by Silicon Valley, who believed in Japan’s reputation for imitation (“click-click” of cameras).
  • Anderson’s Findings:
    • Japanese chips demonstrated superior quality compared to American counterparts.
      • Japanese firms reported failure rates below 0.02% in the first 1,000 hours.
      • American firms ranged from 0.09% to 0.26% failure rates.
    • Despite similar functionality and cost, American DRAM chips malfunctioned more frequently.

Japan’s Rise in Consumer Electronics

  • Japan shed its “cheap” image and became synonymous with high-quality products.
  • Akio Morita, Sony’s entrepreneur, spearheaded this transformation.
    • Sony’s transistor radios challenged American dominance.
    • Japanese companies, emboldened by Sony’s success, targeted higher-end markets.
  • Shift in Strategy:
    • Initial success stemmed from replicating and improving upon U.S. products at lower costs.
    • Gradual transition to innovation, exemplified by Sony’s Walkman.

The Walkman: A Japanese Innovation

  • Introduced in 1979, the Walkman revolutionized portable music.
    • Featured five of Sony’s advanced integrated circuits.
  • Global sales reached 385 million units, making it one of history’s most popular consumer devices.
  • The Walkman demonstrated Japan’s capacity for innovation.

U.S. Support for Japan’s Transformation

  • The U.S. played a key role in Japan’s post-war rise as a tech power:
    • Transfer of transistor technology to Japanese physicists during the occupation.
    • U.S. policies facilitated Japanese firms’ access to American markets.

Concerns About U.S. Competitiveness

  • Some Americans questioned if supporting Japan had inadvertently undermined U.S. economic and technological dominance.
  • Charlie Sporck, a semiconductor industry veteran, observed Japan’s productivity with a mix of fascination and fear.
    • Japan’s efficiency surpassed anything his American workforce could achieve.
    • Spork’s Findings from Sending Workers to Japan:
      • Japanese workers displayed exceptional company loyalty.
      • Japanese foremen prioritized the company over their families.
      • Japanese productivity levels were significantly higher.

Chapter 16: At War with Japan

Jerry Sanders and the Competitive Chip Industry

  • Jerry Sanders, CEO of Advanced Micro Devices (AMD), recognized the fierce competition within the chip industry.
    • Known for his aggressive business style (“I can’t walk away from a fight”).
    • AMD frequently engaged in legal battles with Intel over intellectual property.
  • Charlie Sporck described the industry as a battle for survival: “Knock ’em down, fight ’em, kill ’em.”
  • Intense competition, driven by high stakes and personal pride, was characteristic of the U.S. chip industry.

Japanese Competition: A Different Threat

  • Japanese competition, embodied by companies like Hitachi, Fujitsu, Toshiba, and NEC, posed an existential threat.
  • Spork’s Concerns:
    • Japanese success could shift the entire industry to the Pacific Rim.
    • He drew parallels to the decline of the U.S. television industry, warning of similar consequences for semiconductors.
  • “We’re at war with Japan,” Spork declared, emphasizing an economic battle fought with technology, productivity, and quality.

Accusations of Japanese Espionage and Unfair Practices

  • Spork believed Japanese firms benefited unfairly from:
    • Intellectual property theft
    • Protected markets
    • Government subsidies
    • Cheap capital

The Hitachi Espionage Case (1981)

  • Jun Naruse, a Hitachi employee, was caught attempting to steal trade secrets from Pratt & Whitney, an aircraft manufacturer.
    • Naruse obtained a fake badge to access a secure facility and photograph a computer.
    • Hitachi’s involvement in the scheme was revealed, leading to arrests and a public scandal.
  • Similar accusations of espionage were leveled against Mitsubishi Electric.

Toshiba and the Soviet Submarine Scandal

  • Toshiba, a leading DRAM producer, was implicated in selling machinery to the Soviet Union that helped them build quieter submarines.
  • While unrelated to Toshiba’s semiconductor business, the scandal further fueled perceptions of Japanese companies engaging in unethical practices.

Silicon Valley’s Practices: A Double Standard?

  • The chip industry, including American companies, commonly engaged in:
    • Competitor monitoring
    • Employee poaching
    • Accusations of intellectual property infringement
  • Legal battles and talent acquisition were integral to Silicon Valley’s competitive landscape.

Japan’s Protected Market and Government Support

  • Protected Domestic Market:
    • Japan imposed quotas on U.S. chip imports until 1974.
    • Even after quotas were lifted, Japanese companies favored domestic suppliers.
    • NTT, Japan’s state-owned telecom giant, primarily purchased from Japanese companies.
  • Government Subsidies:
    • Japan’s VLSI program (1976) promoted collaboration among chipmakers and provided government funding for R&D.
    • American companies viewed this as an unfair advantage, despite the U.S. government’s own support for the semiconductor industry through DARPA grants.

The Cost of Capital Disadvantage

  • High U.S. Interest Rates:
    • Jerry Sanders: “The Japanese pay 6%, maybe 7% for capital. I pay 18% on a good day.”
    • Building chip fabrication facilities was capital-intensive, making interest rates a crucial factor.
    • U.S. interest rates soared to 21.5% in the 1980s to combat inflation.
  • Japan’s Low Cost of Capital:
    • Japanese chipmakers, often part of large conglomerates, secured low-interest loans from affiliated banks.
    • Japanese banks provided long-term support, even during periods of unprofitability.
  • Factors Contributing to Japan’s Low Rates:
    • High savings rates due to demographic trends and a weak social safety net.
    • Limited investment options channeled savings into banks.

Japan’s Investment and Market Dominance

  • Armed with cheap capital, Japanese firms aggressively pursued market share in the DRAM market.
  • Japanese companies invested heavily in production equipment, outspending U.S. rivals by 60% in the early 1980s.
  • Sustained Losses:
    • Japanese companies endured losses to gain market share, relying on bank loans to stay afloat.
  • Outcome:
    • Five years after the 64K DRAM introduction, Intel’s market share plummeted to 1.7%, while Japanese competitors thrived.
    • Japan’s capital expenditure in semiconductor manufacturing grew exponentially:
      • Hitachi: From ¥80 billion to ¥1.5 billion
      • Toshiba: From ¥3 billion to ¥75 billion
      • NEC: From ¥3.5 billion to ¥110 billion

Conflicting Narratives and the Perception of Unfairness

  • Japanese chipmakers argued that their success was a result of superior quality and efficiency, not unfair practices.
  • American consumers, like HP, favored Japanese chips due to their higher quality.
  • Despite conflicting narratives, Japan’s DRAM market share continued to grow at the expense of American companies.

Chapter 17: Shipping Junk

GCA Corporation: From Dominance to Decline

  • In the early 1980s, GCA Corporation was a leading American manufacturer of photolithography equipment, crucial for chip production.
  • The Evolution of Photolithography:
    • Advanced from rudimentary setups to sophisticated processes involving:
      • Specialized chemicals
      • Precision lenses
      • Laser alignment systems
  • GCA’s Rise:
    • Initially, Perkin-Elmer’s scanner dominated the market.
    • GCA, led by Milt Greenberg, developed the wafer stepper based on input from Texas Instruments.
    • The stepper’s higher resolution enabled smaller transistors.

GCA’s Monopoly and Subsequent Missteps

  • GCA’s stepper gave it a monopoly, leading to:
    • Revenue surge to $300 million.
    • Soaring stock prices.
  • Internal Issues and Mismanagement:
    • Greenberg, despite his technical brilliance, lacked focus on operational efficiency.
    • The company became bloated, with uncontrolled costs and inventory mismanagement.
    • Greenberg ignored warnings of an impending industry downturn.

The Semiconductor Industry Downturn and Loss of Market Share

  • Industry Downturn (Mid-1980s):
    • The semiconductor industry experienced a cyclical downturn.
    • Lithography equipment sales plummeted by 40% between 1984 and 1986.
    • GCA’s revenue dropped by over two-thirds.
  • Rise of Nikon:
    • GCA’s decision to stop using Nikon lenses backfired when Nikon developed its own stepper, eventually surpassing GCA in market share.
  • GCA’s Downfall:
    • Arrogance toward customers.
    • Unreliable equipment.
    • Poor customer service.

Explanations for GCA’s Decline

  • Impact of Japanese Industrial Subsidies:
    • Some argued Japan’s VLSI program, which benefited DRAM producers, also aided equipment suppliers like Nikon.
  • GCA’s Internal Issues:
    • GCA employees acknowledged difficulties with mass production and quality control.
    • Poor customer service and an arrogant attitude towards clients contributed to their decline.

Nikon’s Superiority and GCA’s Continued Struggles

  • Nikon’s Advantages:
    • Nikon’s systems became more reliable and efficient, outperforming GCA’s even in ideal conditions.
    • Nikon steppers offered significantly longer periods of continuous operation (10 times that of IBM’s expectations).
  • GCA’s Leadership Failure:
    • Greenberg failed to address the company’s internal issues, blaming employees instead.
    • The company resorted to accounting gimmicks to meet Wall Street expectations.
  • Decline in Market Share:
    • U.S. firms’ control over the global semiconductor lithography equipment market plummeted from 85% in 1978 to 50% a decade later.
    • GCA’s share dwindled significantly.

The End of GCA Corporation

  • Sale and Closure:
    • GCA’s parent company, General Signal, decided to sell or close GCA in 1993.
  • Sematech’s Withdrawal of Support:
    • Despite providing millions in funding, Sematech withdrew support.
  • Government Inaction:
    • Despite appeals for help, the government ultimately decided against intervening to save GCA.
  • Final Outcome:
    • GCA closed its doors, sold off its assets, and became another casualty of Japanese competition in the semiconductor industry.

Chapter 18: The Crude Oil of the 1980s

A Gathering at Ming’s: Silicon Valley Turns to Washington

  • Setting:
    • A private dining room at Ming’s Chinese restaurant in Palo Alto.
  • Key Players:
    • Bob Noyce (Intel)
    • Jerry Sanders (AMD)
    • Charlie Sporck (Industry Veteran)
  • Purpose:
    • To address the crisis facing America’s semiconductor industry due to Japanese competition.
  • Decision:
    • To seek government assistance, a departure from Silicon Valley’s previous hands-off approach.

Semiconductors: “The Crude Oil of the 1980s”

  • Jerry Sanders’s Argument:
    • “Semiconductors are the crude oil of the 1980s, and the people who control the crude oil will control the electronics industry.”
  • Rationale:
    • Semiconductors were essential components in a vast array of products, from computers and consumer electronics to military equipment.
    • Their importance was analogous to oil’s role as a critical resource.

Strategic Importance of Semiconductors

  • The Computer Revolution:
    • The rapid expansion of the computer industry in the 1980s increased the demand for semiconductors.
  • Ubiquity of Chips:
    • Semiconductors became essential in a wide range of everyday products, making them indispensable.
  • National Security Implications:
    • The reliance on foreign-made semiconductors raised concerns about national security, especially given America’s dependence on technology for military superiority.

Lessons from the Oil Embargoes

  • The 1973 and 1979 Oil Embargoes:
    • Highlighted the vulnerability of relying on foreign suppliers for critical resources.
    • Led to economic recession and political instability in the U.S.
  • Military Intervention:
    • The U.S. responded to the oil crisis with military actions, demonstrating its commitment to securing oil supplies.

Silicon Valley’s Change of Heart

  • Shift from Defense to Commercial Markets:
    • In the 1970s, Silicon Valley had shifted its focus from defense contracts to civilian markets.
  • Return to Washington:
    • The semiconductor crisis forced companies to seek help from the government they had previously ignored.
  • Formation of the Semiconductor Industry Association (SIA):
    • CEOs, including Sanders, Noyce, and Spork, formed the SIA to lobby for government support.

The Pentagon’s Perspective

  • Semiconductors and Military Superiority:
    • Pentagon officials recognized the crucial role of semiconductors in maintaining America’s military edge, particularly in offsetting the Soviet Union’s conventional forces.
  • Defense Strategy Reliance on Semiconductors:
    • Since the mid-1970s, the U.S. had incorporated semiconductors into advanced weapons systems to enhance guidance, communication, and command and control.

Concerns About Japan’s Semiconductor Dominance

  • Growing Japanese Market Share:
    • By 1986, Japan surpassed the U.S. in chip production.
    • Japan’s share of the global lithography equipment market reached 70% by the end of the 1980s.
  • Dependence on Foreign Suppliers:
    • The trend towards Japanese dominance raised concerns about U.S. dependence on foreign suppliers for critical military technology.

The Dilemma of Japan as an Ally and Competitor

  • Cold War Alliance:
    • Japan was a key U.S. ally during the Cold War.
    • The U.S. encouraged limited Japanese rearmament for support against the Soviet Union.
  • Economic Rivalry:
    • Japan’s economic success and technological advancements in areas crucial for U.S. military power created a strategic dilemma.

The Strategic Importance of Semiconductors

  • Charlie Sporck’s Warning:
    • “You don’t want the same thing to happen to semiconductors as happened to the TV industry, to the camera industry…Without semiconductors, you’re in nowheresville.”
  • Semiconductors as a Foundation of Technology:
    • Spork emphasized that semiconductors were fundamental to a wide range of industries, and losing dominance would have severe consequences for the U.S. economy and national security.

Chapter 19: Death Spiral

Bob Noyce’s Concerns about U.S. Competitiveness

  • “We’re in a death spiral,” Noyce stated in 1986, expressing deep concern about the decline of U.S. industries across multiple sectors.
  • Comparison to Detroit:
    • Noyce drew a parallel between Silicon Valley and Detroit, fearing that the semiconductor industry might suffer a similar fate due to foreign competition.

Silicon Valley’s Ambivalent Relationship with the Government

  • Desire for Autonomy vs. Need for Support:
    • Silicon Valley maintained a complex relationship with the government, seeking both independence and assistance.
  • Noyce’s Perspective:
    • As a leading figure, Noyce recognized the need for government help while remaining wary of potential bureaucratic hurdles.
  • Shift in Market Dynamics:
    • By the 1980s, the government was no longer the primary consumer of semiconductors, limiting its influence over the industry.

Debate over Government Intervention

  • Strategic Importance vs. Free Market Principles:
    • The debate centered on whether the government should support a specific industry or allow market forces to determine winners and losers.
  • Defining “Strategic”:
    • There was no consensus on what constituted a “strategic” industry deserving of government intervention.
  • The Potato Chip Analogy:
    • A Reagan administration economist famously (though disputedly) argued, “They’re all chips. A hundred dollars of one or a hundred dollars of the other is still a hundred.”
    • This highlighted the challenge of justifying government support for one industry over another.

Silicon Valley’s Lobbying Efforts

  • Tax Cuts and Intellectual Property Protection:
    • Noyce advocated for:
      • Reducing the capital gains tax.
      • Loosening regulations on venture capital investments.
      • Strengthening intellectual property protections through the Semiconductor Chip Protection Act.
  • Increased Venture Capital Funding:
    • These policy changes led to a surge in venture capital flowing into Silicon Valley.

Trade Tensions and Market Access

  • Japanese Trade Barriers:
    • Despite agreements to eliminate tariffs, Silicon Valley companies faced challenges selling chips in Japan.
    • Japanese companies and government entities often favored domestic suppliers.
  • Dumping Allegations:
    • U.S. companies accused Japanese firms of dumping cheap chips in the American market.

The U.S.-Japan Semiconductor Trade Agreement (1986)

  • Export Quotas:
    • Under pressure from the U.S., Japan agreed to limit DRAM chip exports.
  • Unintended Consequences:
    • The agreement reduced supply and raised DRAM prices globally, benefiting Japanese producers and hurting American computer manufacturers.
  • Limited Impact on U.S. Companies:
    • Most American DRAM manufacturers were already exiting the market, so the agreement provided minimal relief.

Sematech: A Collaborative Effort

  • Formation (1987):
    • Leading chipmakers, with Defense Department support, formed Sematech to revitalize the U.S. semiconductor industry.
  • Structure and Funding:
    • A consortium funded equally by the industry and the Pentagon.
  • Goal:
    • To foster collaboration and innovation in semiconductor manufacturing.
  • Bob Noyce’s Leadership:
    • Noyce came out of semi-retirement to lead Sematech.

Sematech’s Initiatives

  • Focus on Manufacturing Equipment:
    • Assisted equipment manufacturers like GCA in improving reliability and management practices.
  • Production Coordination:
    • Aligned production schedules between chipmakers and equipment manufacturers.

Saving the U.S. Lithography Industry

  • Lithography as a Priority:
    • Noyce allocated 51% of Sematech’s funding to American lithography companies.
  • Rationale:
    • Lithography was essential for chipmaking, and the U.S. industry was struggling to compete with Japanese rivals.

GCA’s Last Chance

  • Noyce’s Initial Skepticism:
    • Noyce initially believed GCA was beyond saving.
  • Change of Heart:
    • After visiting GCA, Noyce agreed to purchase $13 million worth of their equipment to support the company.

Sematech’s Support for GCA

  • Funding for Advanced Equipment:
    • Sematech provided contracts for GCA to develop cutting-edge deep ultraviolet lithography equipment.
  • GCA’s Technological Success:
    • GCA exceeded expectations, producing world-class steppers.
  • Business Model Challenges:
    • Despite technological achievements, GCA still struggled to secure major customers due to its past reputation and financial instability.

GCA’s Demise

  • Financial Losses:
    • GCA continued to incur losses ($30 million between 1988 and 1992) despite Sematech’s support.
  • Loss of Key Supporter:
    • Noyce’s death in 1990 was a significant blow to GCA.
  • Sale and Closure:
    • GCA was put up for sale but failed to find a buyer.
    • Sematech withdrew funding.
    • The U.S. government decided against intervention.
  • GCA ceased operations in 1993.

Chapter 20: The Japan That Can Say No

Akio Morita’s Changing Perspective on America

  • Early Admiration:
    • Morita initially viewed the U.S. with admiration for its technological prowess and economic prosperity.
  • Growing Disillusionment:
    • America’s struggles in the 1970s and 1980s, including economic crises and social unrest, diminished Morita’s perception of U.S. dominance.
  • Shift in Balance of Power:
    • As Sony thrived and Japan’s technological capabilities grew, Morita sensed a shift in the balance of power.

Morita’s Critique of American Business Practices

  • Short-Term Focus vs. Long-Term Vision:
    • Morita criticized American companies for prioritizing short-term profits over long-term investments and strategic planning.
  • Emphasis on Lawyers Over Engineers:
    • He believed the U.S. overemphasized legal professions while neglecting engineering and manufacturing expertise.
  • Inadequate Labor Relations:
    • Morita criticized American management for failing to adequately train and motivate its workforce.

“The Japan That Can Say No”

  • Publication (1989):
    • Morita co-authored “The Japan That Can Say No” with Shintaro Ishihara, a controversial Japanese politician.
  • Central Argument:
    • The book asserted Japan’s growing economic and technological power and urged the country to adopt a more assertive stance in its relationship with the U.S.
  • Ishihara’s Nationalist Rhetoric:
    • Ishihara, known for his nationalist views, called for Japan to shed its subservient role to the U.S.
  • Semiconductors as a Strategic Lever:
    • Ishihara argued that Japan’s dominance in semiconductors, particularly 1-megabit DRAM chips, gave it leverage over the U.S., including in military affairs.

Reactions to the Book

  • Controversy in the U.S.:
    • The book sparked outrage and fueled anxieties about Japan’s economic and technological rise.
  • Morita’s Attempts to Distance Himself:
    • Morita later downplayed his association with Ishihara, claiming their views differed.

The Reality of Japan’s Semiconductor Dominance

  • Facts Supported Ishihara’s Assertions:
    • Japan had undeniably become the leading producer of DRAM chips, essential for computers and military technology.
  • Geopolitical Implications:
    • This shift in technological power had significant potential to reshape the geopolitical landscape.

High-Tech as Foreign Policy

  • Harold Brown’s Analysis:
    • Former U.S. Defense Secretary Harold Brown, in a 1989 article titled “High-Tech is Foreign Policy,” acknowledged the link between technological leadership and geopolitical power.
  • Concerns about U.S. Dependence:
    • Brown admitted that Japan’s dominance in memory chips and advancements in other semiconductor technologies posed a risk to U.S. military superiority.

The Specter of a “Pax Nipponica”

  • CIA’s Assessment (1989):
    • A CIA report predicted the emergence of a “Pax Nipponica,” an East Asian economic and political bloc dominated by Japan.
  • Shift in Regional Dynamics:
    • Japan’s growing economic and technological influence threatened to undermine America’s longstanding dominance in Asia.
  • Unintended Consequences of U.S. Policy:
    • The U.S. strategy of fostering Japan’s post-war economic growth had inadvertently created a formidable competitor.

Chapter 21: The Potato Chip King

  • Jack Simplot, an Idaho billionaire known for his potato empire, invested in Micron Technology, a struggling DRAM chip manufacturer.
    • Simplot’s expertise was in potatoes, evident from his “Mr. Spud” license plate, but he understood business in a way that Silicon Valley often didn’t.
    • Simplot made his fortune by pioneering potato processing techniques that led to a massive contract with McDonald’s.
  • Micron, founded in 1978 by brothers Joe and Ward Parkinson, entered the DRAM market during a time of intense Japanese dominance.
    • Existing American DRAM producers, including AMD, National Semiconductor, and Intel, were abandoning the market due to billion-dollar losses from Japanese competition.
    • Micron, initially focused on chip design, was forced to pivot to manufacturing their own chips after losing their only client, Mostek, to Japanese competitor Fujitsu.
  • Silicon Valley lobbied the government for protection against Japanese competition, claiming that computer chips were strategic assets unlike commodities like potatoes.
  • Micron, however, initially opposed government intervention, believing they could compete with the Japanese through aggressive cost-cutting.
  • Micron secured funding from Alan Noble and other Boise investors. Facing continued struggles, they approached Simplot, who saw an opportunity in the depressed DRAM market and invested $1 million, later increasing his stake to over a billion dollars.
    • Simplot believed that Japanese competition had turned DRAM into a commodity, and the best time to buy a commodity business was during a downturn.
  • Micron’s competitive edge stemmed from their relentless focus on cost reduction and innovative manufacturing processes.
    • Ward Parkinson, Micron’s engineering lead, focused on shrinking the chip size to fit more chips onto each silicon wafer, increasing manufacturing efficiency.
    • Micron simplified manufacturing processes, using fewer steps and less equipment than competitors.
    • They modified existing lithography machines to exceed manufacturer specifications and adapted furnaces to process more wafers simultaneously.
    • Micron’s location in Boise offered lower land and electricity costs compared to California or Japan.
  • Joe Parkinson instilled a culture of cost-consciousness, dimming lights and emphasizing efficiency to employees. This “sweatshop mentality” was driven by the understanding that Micron’s survival was paramount in a place with limited alternative job opportunities.
  • Micron’s strategy ultimately succeeded.
    • They survived the Japanese onslaught, while most American DRAM producers were driven out of the market.
    • Micron acquired Texas Instruments’ DRAM business and became a major player in the memory chip market.
    • Their success was attributed to their engineering ingenuity, relentless cost-cutting, and Simplot’s unwavering support.

Chapter 22: Disrupting Intel

  • Andy Grove, Intel’s President, recognized that the company’s DRAM business was being disrupted by Japanese competition and needed a drastic change.
    • Grove was known for his “paranoid” management style, driven by a fear of competition and failure, which he outlined in his book, Only the Paranoid Survive.
    • Grove’s experiences as a Hungarian refugee escaping Soviet oppression shaped his leadership, emphasizing constant vigilance and a drive to overcome challenges.
  • Despite Intel’s history and identity being rooted in memory chips, Grove acknowledged that the DRAM market was no longer viable and a shift was necessary for survival.
    • The decision to abandon DRAM was difficult and emotionally charged for Intel, but Grove recognized the need to disrupt themselves before being overtaken by competitors.
  • Grove initiated a painful restructuring process.
    • Step One: Laying off over 25% of Intel’s workforce and closing facilities in Silicon Valley, Oregon, Puerto Rico, and Barbados.
    • Step Two: Implementing a ruthless approach to improve manufacturing, inspired by Japanese methods.
      • Craig Barrett, Grove’s deputy, led this effort, pushing for the adoption of rigorous processes.
      • Intel adopted a “copy exactly” manufacturing methodology, replicating the most efficient processes across all facilities to ensure consistency and high yields.
      • This shift from Silicon Valley’s freewheeling engineering culture to a more disciplined and standardized approach was met with internal resistance but ultimately proved successful in increasing efficiency and reducing costs.
  • Several external factors contributed to Intel’s resurgence:
    • The rising value of the Japanese yen against the dollar made American exports cheaper.
    • Falling interest rates in the US lowered Intel’s capital costs.
    • The emergence of Compaq and other IBM PC clones created a new market for Intel’s microprocessors.
    • These “clone” manufacturers utilized Intel’s chips and Microsoft’s software, driving down PC prices and fueling mass adoption.
  • By transitioning from DRAM to microprocessors, Intel secured a near-monopoly in the burgeoning PC market, solidifying their position as a leading chip manufacturer.

Chapter 23: My Enemy’s Enemy, The Rise of Korea

  • Lee Byung-chul, founder of Samsung, saw an opportunity in the semiconductor industry amidst the US-Japan DRAM wars.
    • Lee, a shrewd businessman who started by trading dried fish and vegetables, believed in building businesses that were “big, strong, and eternal”.
    • He recognized the potential of semiconductors after witnessing the success of Japanese companies like Toshiba and Fujitsu.
  • Lee was inspired by Hewlett-Packard’s journey from a garage startup to a tech giant during a visit to California in 1982.
    • He was determined to replicate their success with Samsung.
  • Despite the high capital expenditure and risk involved, Lee decided to invest at least $100 million in semiconductor manufacturing, a move that could have jeopardized his entire business empire.
  • Support from the South Korean government was crucial:
    • The government pledged $400 million to develop the country’s semiconductor industry and pressured banks to provide additional loans to companies like Samsung.
    • This mirrored the Japanese model of government-backed industrial policy.
  • Silicon Valley, eager to undercut Japanese dominance, saw Korean companies as potential allies.
    • Bob Noyce believed that Korean competition would prevent Japan from monopolizing DRAM production.
    • Intel and other Silicon Valley companies formed joint ventures with Samsung, providing them with manufacturing contracts and technology.
    • The logic was “my enemy’s enemy is my friend,” as Jerry Sanders stated.
  • Factors that aided Samsung’s rise:
    • US-imposed tariffs on Japanese DRAM chips created an opening for Korean companies.
    • Lower wages and production costs in South Korea compared to Japan.
    • Access to US technology through licensing agreements, such as the deal for a 64K DRAM design from Micron.
  • The collaboration between Silicon Valley and South Korea, driven by a shared goal of challenging Japan, inadvertently helped establish Korea as a major player in the global memory chip market.

Chapter 24: This is the Future

  • By the 1980s, the limitations of manual chip design were becoming apparent as transistor counts increased.
    • Early microprocessors, like Intel’s 4004, were designed by hand, a time-consuming and error-prone process.
    • Federico Fagan’s experience designing Intel’s first microprocessor in 1971 highlighted the challenges: he spent six months meticulously drawing the design, which was then transferred to rubylith film and used to create masks for photolithography.
  • Carver Mead and Lynn Conway revolutionized chip design by introducing a standardized, algorithmic approach.
    • Carver Mead, a physicist, and Lynn Conway, a computer architect at Xerox PARC, collaborated to create a set of mathematical design rules that enabled software to automate the chip design process.
    • Their methodology allowed designers to use a library of pre-designed components instead of drawing every transistor individually. This “Mead-Conway revolution” mirrored Gutenberg’s invention of the printing press, simplifying chip design and making it more accessible.
  • DARPA (Defense Advanced Research Projects Agency) recognized the strategic importance of the Mead-Conway approach and funded initiatives to promote its adoption.
    • They financed programs for university researchers to design and fabricate chips at advanced facilities, fostering a new generation of chip designers.
    • DARPA’s focus was on building educational infrastructure and supporting research that would maintain America’s lead in microelectronics, recognizing the crucial role of innovation in national security.
  • The Semiconductor Research Corporation (SRC), funded by the chip industry, also played a critical role by distributing research grants to universities.
    • Universities like Carnegie Mellon and the University of California, Berkeley became hubs for semiconductor research, producing graduates who went on to found companies that developed essential software tools for chip design.
    • These DARPA and SRC-funded programs nurtured an ecosystem of innovation that led to the creation of companies like Synopsys, Cadence, and Mentor Graphics, which continue to dominate the chip design software industry today.
  • Erwin Jacobs, a pioneer in wireless communication, envisioned using increasingly powerful chips to revolutionize data transmission.
    • Jacobs, who co-founded Qualcomm, realized that as transistor density increased, chips would be able to encode and decode significantly more data within the limited bandwidth of radio waves.
    • He predicted that this would lead to a future where large amounts of data could be transmitted wirelessly.
  • DARPA funding was instrumental in Qualcomm’s early success, supporting the development of space communication systems before the company shifted focus to the civilian market.
  • Jacobs’ story highlights the importance of government funding in supporting early-stage research and development, particularly in capital-intensive fields like semiconductors.
  • By the end of the 1980s, thanks to these advancements in chip design, manufacturing, and applications, the unthinkable had become reality: Intel released its 486 microprocessor, containing 1.2 million transistors, a testament to the relentless pace of innovation driven by Silicon Valley and its government and academic partners.

Chapter 25: The KGB’s Directorate T

  • The Soviet Union, lagging behind the West in semiconductor technology, relied heavily on espionage to acquire Western chips and manufacturing equipment.
    • Directorate T, a specialized KGB division established in 1963, was tasked with stealing foreign technology to bolster the Soviet semiconductor industry.
    • The KGB employed various tactics, including:
      • Direct theft: Stealing physical chips from companies.
      • Black market purchases: Acquiring chips through intermediaries and smugglers.
      • Blackmail: Targeting Westerners with access to sensitive technology.
  • Soviet espionage efforts were extensive:
    • The KGB reportedly had a team of 60 agents in San Francisco focused on infiltrating Silicon Valley companies.
    • The discovery of a sophisticated Soviet listening device using Texas Instruments chips disguised as a buoy in 1982 highlighted the reach and sophistication of their operations.
  • The limitations of the “copy-it” strategy:
    • While stealing samples or even shipments of chips was relatively straightforward, replicating the complex manufacturing processes was a significant challenge.
    • As chip complexity increased, the Soviets struggled to produce high-quality copies, rendering their stolen technology less valuable.
    • This over-reliance on theft hindered the development of indigenous Soviet semiconductor expertise.
  • Vladimir Vetrov (codename: Farewell): A disillusioned KGB agent who became a double agent for France.
    • Vetrov, dissatisfied with his career and personal life, leaked thousands of pages of classified documents about Directorate T’s operations to French intelligence, revealing the vast scale of Soviet technology theft.
    • He was motivated by a desire for a more exciting life and hoped to rekindle a romantic relationship.
  • Operation Exodus: A US-led initiative to counter Soviet technology theft, launched in response to Vetrov’s revelations.
    • The program implemented stricter export controls, increased customs inspections, and targeted Soviet acquisition networks.
    • By 1985, Operation Exodus had seized $600 million worth of goods and led to numerous arrests.
  • Despite the KGB’s efforts, the Soviet Union remained technologically inferior to the West in microelectronics.
    • By the mid-1980s, Soviet microprocessors were estimated to be half a decade behind their American counterparts.
    • The reliance on the copy-it strategy ultimately proved detrimental, hindering innovation and perpetuating dependence on foreign technology.

Chapter 26: Weapons of Mass Destruction, the Impact of the Offset

  • Marshal Nikolai Ogarkov, chief of the Soviet military, recognized the transformative impact of microelectronics on warfare, predicting that precision-guided weapons would turn conventional explosives into “weapons of mass destruction.”
  • Ogarkov argued that the Soviet Union’s numerical superiority in tanks and troops was becoming increasingly irrelevant in the face of America’s superior precision-guided weapons and surveillance technology.
    • He recognized that Bill Perry’s offset strategy of leveraging technology to counter Soviet numerical superiority was succeeding.
  • Impact of the semiconductor gap on Soviet military capabilities:
    • Missile technology:
      • Soviet ballistic missiles were less accurate than their American counterparts due to the limitations of their onboard computers and guidance systems.
      • The US MX missile was estimated to be significantly more accurate than the Soviet SS-25, giving the US a strategic advantage in a potential nuclear conflict.
    • Anti-submarine warfare:
      • The US Navy’s use of advanced computers, like the ILLIAC-4, for analyzing sonar data significantly enhanced their ability to detect and track Soviet submarines.
    • Conventional warfare:
      • The effectiveness of US precision-guided munitions, such as the Paveway laser-guided bombs and Tomahawk cruise missiles, significantly increased the vulnerability of Soviet forces in a conventional conflict.
  • Challenges faced by the Soviet semiconductor industry:
    • Political interference:
      • The KGB’s influence within the industry hindered efficiency and innovation.
      • Yuri OShokhin’s removal from the Riga semiconductor plant for refusing to dismiss employees targeted by the KGB exemplified this problem.
    • Lack of a robust consumer market:
      • The absence of a large civilian market for electronics limited the growth and innovation of the Soviet semiconductor industry.
      • This contrasted with the US, Europe, and Japan, where booming consumer markets drove demand for advanced chips.
    • Limited international collaboration:
      • The Soviet Union’s isolation during the Cold War prevented it from accessing the global semiconductor supply chain and benefiting from specialization and economies of scale.
      • In contrast, the US and its allies had established an efficient international division of labor in chip production and innovation.
  • The Soviet Union’s efforts to revitalize its microelectronics industry ultimately failed.
    • Despite attempts to modernize Zelenograd, the USSR couldn’t match the technological advancements or the economic dynamism of Silicon Valley.

Chapter 27: War Hero

  • The 1991 Persian Gulf War served as the first major test of US military technology developed after the Vietnam War.
    • The US military’s overwhelming victory against Iraq demonstrated the effectiveness of precision-guided munitions and advanced electronics in warfare.
  • The Paveway laser-guided bomb:
    • Originally developed during the Vietnam War, the Paveway underwent continuous improvement, with each iteration incorporating more advanced microelectronics.
    • Weldon Word, the key figure behind the Paveway’s development, focused on improving its accuracy, reliability, and affordability.
  • Key advantages of the Paveway:
    • High accuracy: Paveway bombs significantly increased hit rates compared to unguided munitions.
    • Versatility: Targets could be chosen in real-time, allowing for flexibility on the battlefield.
    • Cost-effectiveness: Advancements in microelectronics made the Paveway increasingly affordable, enabling widespread adoption.
  • The impact of microelectronics on other military technologies:
    • Improved guidance systems for missiles, like the Sidewinder air-to-air missile, significantly increased their accuracy.
    • Enhanced surveillance and communication systems provided the US military with a decisive information advantage.
  • Validation of the offset strategy:
    • The Persian Gulf War proved the effectiveness of Bill Perry’s offset strategy of investing in technology to counter Soviet numerical superiority.
  • “Silicon over steel”:
    • The war highlighted the shifting balance of power in warfare, with technology and information becoming as important as traditional military hardware.

Chapter 28: The Cold War is Over and You Have Won

  • The 1990s marked a period of decline for the Japanese semiconductor industry and the rise of American dominance.
    • Japan’s economic bubble burst in 1990, leading to a prolonged recession and exposing the weaknesses of its over-investment in certain industries, including semiconductors.
  • Factors contributing to Japan’s semiconductor decline:
    • Overinvestment and lack of profitability:
      • Japanese chipmakers, fueled by easy access to capital, continued investing heavily in DRAM production despite declining prices and competition from lower-cost producers like Micron and Samsung.
      • This overcapacity led to a DRAM glut and falling profits.
    • Missed opportunities in new markets:
      • Japanese companies, with the notable exception of Sony, failed to capitalize on the rising PC market and largely ignored the growing demand for microprocessors.
      • This contrasted with Intel’s successful pivot from DRAM to microprocessors.
    • Slow to adapt to changing market dynamics:
      • The Japanese model of consensus-driven decision-making, which had been an advantage in the past, hindered their ability to adapt quickly to the rapidly changing semiconductor landscape.
  • Sony:
    • Sony’s success in developing specialized chips for image sensors highlighted the potential for innovation within the Japanese semiconductor industry.
    • However, even Sony struggled to maintain profitability in the face of increasing competition and the company’s failure to cut investments in loss-making sectors.
  • The decline of Japan’s semiconductor industry undermined its geopolitical ambitions:
    • Akio Morita’s, co-author of “The Japan That Can Say No,” witnessed firsthand the decline of Japan’s economic and technological prowess.
    • The country’s inability to translate its semiconductor dominance into geopolitical power during the Persian Gulf War further highlighted its limitations.
  • The Soviet Union’s collapse:
    • By 1990, Mikhail Gorbachev, recognizing the Soviet Union’s dire economic and technological situation, sought closer ties with the West.
    • During a visit to Silicon Valley, Gorbachev acknowledged the region’s technological leadership and encouraged US investment in the USSR.
  • Ogarkov’s prediction:
    • Marshal Ogarkov’s earlier prediction that the US had effectively won the Cold War due to its superior technology proved accurate.
    • The Soviet Union’s inability to compete in microelectronics and other advanced technologies significantly contributed to its collapse.
  • The end of the Cold War:
    • The US emerged victorious, with Silicon Valley’s technological prowess playing a decisive role in securing this victory.
    • The semiconductor industry had become a critical driver of both economic and military power.

Chapter 29: We Want a Semiconductor Industry in Taiwan

Taiwan’s Semiconductor Ambitions (1960s-1980s)

  • By the 1980s, Taiwan aimed to transition from semiconductor assembly to chip fabrication to capture a larger share of the industry’s profits.
    • Semiconductor assembly, primarily involving testing and packaging chips made elsewhere, offered lower profits than chip design and production.

Early Efforts and Challenges

  • Taiwan had been strategically integrating itself into semiconductor supply chains since the 1960s.
    • Goals:
      • Job creation
      • Acquisition of advanced technology
      • Strengthened security ties with the United States
  • Early successes included attracting Texas Instruments to build the island’s first semiconductor facility in the 1960s.
    • K.T. Lee, a powerful Taiwanese minister, was instrumental in securing TI’s investment and fostering relationships with its leaders, including Pat Haggerty and Morris Chang.
  • Taiwan further encouraged electronics firms to establish factories on the island.
  • Despite these efforts, Taiwan faced challenges in advancing its chipmaking capabilities:
    • UMC, a Taiwanese chipmaker founded in 1980 using licensed technology from America’s RCA, lagged behind the industry’s cutting edge.
    • Competition: Taiwan faced fierce competition from other Asian economies:
      • South Korea: Samsung and other conglomerates were heavily investing in advanced memory chips.
      • Singapore and Malaysia: These countries were attempting to replicate South Korea’s transition from assembly to fabrication, but with limited success.
    • China’s Emergence: The People’s Republic of China’s integration into the global economy posed a significant threat.
      • China’s low wages and vast workforce attracted basic manufacturing and assembly jobs, directly challenging Taiwan’s economic model.
      • Taiwanese officials perceived this as “economic warfare,” as competing with China on price was deemed impossible.

Morris Chang and the Birth of TSMC (1985)

  • In 1985, Minister K.T. Lee recruited Morris Chang, a seasoned semiconductor industry veteran, to lead Taiwan’s push into chip fabrication.
    • Chang had extensive experience at Texas Instruments, playing a key role in establishing TI’s efficient manufacturing processes and the company’s presence in Taiwan.
    • He had left TI in the early 1980s after being passed over for the CEO position.
    • Chang’s deep understanding of the industry and proven track record made him an ideal candidate for the challenging task.

The Foundry Model: A Radical Idea

  • Chang proposed a radical business model—a “foundry”—to propel Taiwan’s semiconductor industry forward.
    • This model involved manufacturing chips designed by other companies rather than designing and producing chips in-house, which was the prevailing industry model at the time.
  • Chang had first pitched this idea while at TI in 1976, predicting the rise of “fabless” companies that focused solely on chip design.
    • However, TI executives rejected the proposal, deeming it too risky to invest in markets that didn’t yet exist.
  • The concept gained traction due to several factors:
    • Lynn Conway and Carver Mead’s work: Their revolution in chip design, which simplified the separation of design and manufacturing, made Chang’s foundry model more feasible.
    • Growing demand: Chang anticipated that the decreasing cost of computing power would drive demand for semiconductors in a wide range of new applications, creating opportunities for specialized manufacturers.
    • Rising manufacturing costs: As transistors shrank and technology advanced, the cost of manufacturing equipment and R&D became increasingly prohibitive for companies producing smaller volumes of chips.

TSMC’s Founding and Government Support

  • Minister Lee secured government funding for Chang’s ambitious plan, with the Taiwanese government providing 48% of TSMC’s startup capital.
    • The government also granted generous tax benefits to support the company’s growth.
  • Chang’s former colleagues at TI and Intel declined to provide the necessary advanced production technology.
    • Gordon Moore, of Intel, famously told Chang, “Morris, you’ve had a lot of good ideas in your time. This isn’t one of them.”
  • Eventually, Chang secured a deal with Philips, the Dutch semiconductor company:
    • Philips invested $58 million, transferred its production technology, and licensed intellectual property to TSMC in exchange for a 27.5% stake.
  • The remaining capital was raised from wealthy Taiwanese families, often through government pressure and appeals to their past successes with government support.
  • From its inception, TSMC operated as a project of the Taiwanese state, heavily reliant on government support and funding.

TSMC’s Success and the Rise of Taiwan’s Chip Dominance

TSMC’s Symbiotic Relationship with Silicon Valley

  • TSMC’s early success was deeply intertwined with the U.S. chip industry.
    • Many of TSMC’s top executives and engineers had experience working at leading Silicon Valley companies like Motorola, Intel, and TI, bringing valuable expertise to the company.
    • A significant portion of TSMC’s early customers were U.S.-based chip designers.
    • By the mid-1990s, American companies accounted for half of TSMC’s sales.

The Foundry Model’s Impact: Democratization and Monopolization

  • TSMC’s foundry model addressed the challenges faced by early fabless chip design firms, providing them with a reliable manufacturing partner and fostering innovation in the industry.
    • Previously, these smaller firms had to rely on larger chipmakers with spare capacity, often facing second-class status, potential intellectual property theft, and inconsistencies in manufacturing processes.
  • TSMC’s commitment to manufacturing chips designed by its customers, without competing in the design space, built trust and attracted a wide range of clients.
  • This model, mirroring Carver Mead’s “Gutenberg moment” prediction, significantly lowered the barriers to entry in the chip design industry, leading to a surge in new fabless chip design firms.
    • The availability of affordable, high-quality chip manufacturing enabled these companies to focus on innovation and develop specialized chips for a wide range of applications.
  • However, unlike Gutenberg’s printing press, which spread rapidly across Europe, TSMC’s foundry model, while democratizing chip design, ultimately led to the monopolization of chip manufacturing.
    • The economics of chip fabrication favored large-scale production, with companies like TSMC gaining a significant advantage through higher yields and lower costs per chip.
    • This dynamic drove relentless consolidation in the industry, with TSMC emerging as the dominant player in the production of the world’s most advanced chips.

Conclusion

  • Morris Chang’s vision, combined with Taiwan’s unwavering government support, transformed the island into a global semiconductor powerhouse.
  • TSMC’s foundry model revolutionized the chip industry, enabling innovation and driving the widespread adoption of semiconductors in countless devices.
  • However, this success came at the cost of increased industry consolidation, with TSMC wielding unprecedented influence over the production of the world’s most critical technology.

Chapter 30: All People Must Make Semiconductors

The Rise of Taiwan and Huawei

  • In 1987:
    • Morris Chang founded TSMC in Taiwan.
    • Ren Jung-fei founded Huawei, an electronics trading company, in Shenzhen, China.
  • Taiwan:
    • Small island with global ambitions in chipmaking.
    • Deep connections with:
      • Advanced chip companies.
      • Highly educated engineers (Stanford, Berkeley).
  • China:
    • Vast population, but economically and technologically behind.
    • Economic openness boosted trade, especially through Hong Kong.
    • Shenzhen:
      • Located near Hong Kong, became a hub for electronics trade.
      • Huawei bought cheap equipment in Hong Kong and sold it at higher prices within China.
  • Different Approaches:
    • TSMC (Taiwan): Focused on building advanced chips, targeting Silicon Valley giants as customers.
    • Huawei (China): Initially focused on trading electronics, with no chip production aspirations.

China’s Semiconductor Struggle Under Communist Rule

  • 1980s:
    • Chinese government, led by Jiang Zemin, prioritized electronics.
    • China’s most advanced domestic chip was a DRAM over a decade behind the cutting edge.
  • Lost Potential:
    • China possessed factors that attracted American semiconductor investment in other Asian countries:
      • Vast, low-cost workforce.
      • Well-educated scientific elite.
    • However, the communist regime:
      • Viewed foreign connections with suspicion.
      • Discouraged talent like Morris Chang from returning from abroad.
      • Made similar mistakes to the Soviet Union, but more extreme.

Mao’s Disastrous Impact on China’s Chip Industry

  • Early Efforts:
    • Mid-1950s: Beijing identified semiconductors as a scientific priority.
    • 1960:
      • Established its first semiconductor research institute in Beijing.
      • Began manufacturing simple transistor radios.
    • 1965: Chinese engineers created their first integrated circuit, 5 years behind the US.
  • The Cultural Revolution (1966-1976):
    • Mao Zedong’s radical policies devastated China’s scientific and technological progress.
    • Intellectuals and experts were persecuted, sent to work as farmers, or killed.
    • Education system severely disrupted.
    • Mao’s Directive (July 21, 1968):
      • Shorten schooling.
      • Revolutionize education.
      • Prioritize proletarian politics over expertise.
      • Students selected from workers and peasants.
      • Return to production after a few years.
    • This approach hindered technological advancement and innovation.
  • Self-Reliance and Suspicion of Foreign Technology:
    • Mao enforced an embargo on foreign technology, even though China couldn’t produce advanced components.
    • Propaganda promoted “independent and self-reliant development” of the electronics industry.
  • Ideological Opposition to Electronics:
    • Mao viewed electronics as potentially anti-socialist.
    • Prioritized heavy industries like iron and steel over electronics.

The Contrast with Hong Kong and Taiwan

  • Hong Kong:
    • Still under British rule, avoided the Cultural Revolution.
    • Fairchild Semiconductor operated a plant in Hong Kong, employing workers while mainland China faced turmoil.
  • Taiwan:
    • Hosted multiple US chip firms, providing jobs and fostering technological development.

The Aftermath of the Cultural Revolution

  • Early 1970s:
    • Mao’s health declined, leading to a gradual end to the Cultural Revolution.
    • Scientists returned from rural areas, but damage was severe.
    • China’s chip industry lagged far behind its neighbors and the West.
  • 1975:
    • China only produced one usable semiconductor per 1,000 manufactured.
  • John Bardeen’s Visit (September 2nd, 1975):
    • Two-time Nobel laureate in Physics.
    • Visited China as part of a delegation of American physicists.
    • Visit signaled a thaw in relations after the Cultural Revolution.
    • Observed the state of China’s semiconductor industry and found it lacking.

Deng Xiaoping and the Four Modernizations

  • Post-Mao Era:
    • Mao died in 1976.
    • Deng Xiaoping rose to power.
    • Implemented the Four Modernizations, including advancements in science and technology.
  • National Science Conference (March 1978):
    • Semiconductors were central to the agenda.
    • Aim: Develop weapons systems, consumer electronics, and computers.
  • Challenges:
    • Decades of isolation and ideological opposition had severely hampered China’s technological capabilities.
    • China remained heavily reliant on foreign chips.

The “Made in China” Obsession

  • 1980s:
    • China pushed for domestic production of semiconductors.
    • This drive was hindered by:
      • Technological backwardness.
      • Bureaucracy.
      • Lack of access to advanced foreign technology.
  • Huawei’s Rise:
    • Despite the push for domestic production, companies like Huawei remained reliant on foreign chips for their electronics assembly.

Key Takeaways:

  • The Cultural Revolution had a devastating and long-lasting impact on China’s semiconductor industry.
  • Despite efforts to catch up, China remained reliant on foreign technology.
  • The “Made in China” ambition faced significant challenges due to the legacy of past policies and the global dominance of other chip-producing regions.

Chapter 31: Sharing God’s Love with the Chinese

Richard Chang and Semiconductor Manufacturing in China

  • Richard Chang, a devout Christian and semiconductor engineer, aimed to bring advanced chip-making to China.
    • Born in Nanjing in 1948, his family fled to Taiwan when he was one year old after the Communist takeover.
    • Earned a graduate degree in Buffalo, New York, and worked at Texas Instruments, gaining expertise in operating fabs.
  • In 2000, Chang founded Semiconductor Manufacturing International Corporation (SMIC) in Shanghai.
    • Secured over $1.5 billion from international investors, with an estimated half coming from U.S. investors.
    • Hired hundreds of foreign experts, including at least 400 from Taiwan, to operate SMIC’s fab.
  • Chang’s strategy mirrored TSMC’s successful approach in Taiwan:
    • Recruit the best engineers with experience at advanced chip firms.
    • Invest in the best equipment.
    • Prioritize employee training in industry best practices.
    • Leverage government tax and subsidy benefits.
  • SMIC’s aggressive hiring strategy:
    • Focused on recruiting experienced engineers from overseas chipmakers, primarily Taiwan.
    • “One Old Staffer Brings Along Two New Ones”: A company slogan emphasizing the need for experienced foreign-trained employees to train local engineers.
    • Foreign-trained workforce proved crucial for SMIC’s success in domesticating technology.
  • Workforce Composition (Doug Fuller’s analysis):
    • In 2001: 650 local engineers and 393 recruited from overseas (mostly Taiwan and the U.S.).
    • Throughout the 2000s: Approximately one-third of engineering employees were hired from abroad.
  • SMIC’s Successes:
    • Benefited from significant government support (five-year corporate tax holiday, reduced sales tax on chips sold in China).
    • Focused on manufacturing quality and adopting near cutting-edge technology.
    • Secured contracts to build chips for industry leaders like Texas Instruments.
    • Listed its shares on the New York Stock Exchange in 2004.
  • By the late 2000s, SMIC was only a couple of years behind the world’s technology leaders, on track to potentially rival TSMC.

The Global Shift in Semiconductor Fabrication (1990s-2000s)

  • Declining U.S. Market Share:
    • 1990: U.S. fabs produced 37% of the world’s chips.
    • 2000: This figure fell to 19%.
    • 2010: Further declined to 13%.
  • Japan’s Collapsing Market Share: Experienced a significant decline (specific figures not provided).
  • Rise of East Asian Competitors:
    • South Korea, Singapore, and Taiwan made substantial investments in their chip industries, leading to rapid output increases.
  • Singapore:
    • Government-funded fabrication facilities and chip design centers in partnership with companies like Texas Instruments, Hewlett-Packard, and Hitachi, fostering a thriving semiconductor sector.
    • Chartered Semiconductor: A government-backed foundry established to emulate TSMC (though it never achieved the same level of success).
  • South Korea:
    • Samsung’s Rise: After becoming the world’s leading memory chip maker in 1992 by dethroning Japanese DRAM producers, Samsung experienced rapid growth.
      • Benefited from government support, including pressure on banks to provide credit.
      • Successfully fended off competition from Taiwan and Singapore in the DRAM market.
  • The DRAM “Chicken Game”:
    • Characterized by intense competition and heavy investment in new factories, leading to overcapacity and price drops.
    • Samsung’s financial strength allowed it to outspend rivals during downturns, leading to increased market share.

China’s Semiconductor Ambitions

  • 1990s:
    • China emerged as the “world’s workshop,” with cities like Shanghai and Shenzhen becoming hubs for electronics assembly.
    • However, chip manufacturing lagged behind Taiwan, South Korea, and the U.S.
    • Smuggling chips from Hong Kong remained profitable due to limited domestic production.
  • Government Efforts to Develop Domestic Chip Industry:
    • Early attempts were often characterized by:
      • Reliance on foreign investment and technology transfer agreements that offered limited benefits to China.
      • Examples:
        • Hua Hong and NEC Joint Venture: Japanese experts retained control, leaving Chinese workers with basic tasks.
        • Grace Semiconductor: A venture involving politically connected individuals (Jiang Minhang, son of Chinese President Jiang Zemin, and Winston Wang) and Neil Bush (brother of U.S. President George W. Bush). Despite this, the company struggled with technology and market share.
  • Shift from Assembly to Components:
    • Chinese leaders recognized the greater profitability of producing semiconductors compared to assembling electronics.
    • Richard Chang’s vision of bringing chip manufacturing to China aligned with this ambition.

The Rise of Fabless Semiconductor Designers and the Smartphone Revolution

  • The emergence of fabless semiconductor firms that designed chips but outsourced manufacturing.
  • Benefits of Offshoring: Lower manufacturing costs and increased competition, resulting in lower prices for consumers.
  • The Smartphone Revolution:
    • Fabless firms played a key role in developing smartphones, which required numerous complex chips.
    • The combination of offshoring, competition, and technological innovation benefited consumers with affordable and advanced devices.

Conclusion

  • Richard Chang’s efforts to establish SMIC as a leading chip manufacturer in China exemplified the global shift in semiconductor production.
  • Government support, foreign investment, and the relentless pursuit of technological advancement characterized this period.
  • The rise of fabless firms, coupled with the increasing demand for complex chips in devices like smartphones, further fueled the growth of the semiconductor industry.

Chapter 32: Lithography Wars

The Challenge of Shrinking Transistors

  • Lithography - A process using light to carve circuits onto silicon wafers.
  • As transistors shrunk due to Moore’s Law, traditional lithography methods faced limitations.
    • Moore’s Law: The observation that the number of transistors on a microchip doubles approximately every two years, leading to exponential increases in computing power.
  • Existing ultraviolet (UV) light techniques with wavelengths of 248 or 193 nanometers would soon be insufficient.
  • Extreme ultraviolet (EUV) light, with a wavelength of 13.5 nanometers, was proposed as the solution for producing smaller circuits.
    • Smaller wavelengths enable carving smaller features onto chips.
  • Challenge: Most experts believed mass-producing EUV light was impossible.

Intel’s Gamble on EUV

  • 1992: John Carruthers, Intel’s R&D leader, requested $200 million from CEO Andy Grove to develop EUV lithography.
    • Grove was skeptical, as EUV’s viability was uncertain.
    • Carruthers argued it was a necessary “research” investment.
    • Gordon Moore (former CEO and advisor) agreed, highlighting the lack of alternatives if Moore’s Law were to continue.
  • Outcome: Grove approved the funding, ultimately leading to billions invested in EUV R&D and implementation.
  • Intel’s Strategy:
    • Focused on ensuring at least one lithography company successfully brought EUV machines to market.
    • Intel needed these tools to continue shrinking circuits on their chips.

Three Battles over Lithography’s Future

1. The Engineering Battle: Finding the Right Beam

  • The Challenge: Transistors were approaching sizes where the wavelength of light used in lithography significantly impacted precision.
  • Competing technologies:
    • Electron beam lithography: Precise but too slow for mass production.
    • X-ray lithography: Used X-rays and specific photoresist chemicals.
    • EUV lithography: Utilized extreme ultraviolet light and its corresponding photoresists.
  • “Lithography Wars”: Intense competition between engineering groups advocating for different technologies.

2. The Commercial Battle: Dominance in Lithography Equipment

  • High Stakes: Developing new lithography equipment was incredibly expensive, leading to industry consolidation.
  • Key Players:
    • Canon (Japan): Market leader.
    • Nikon (Japan): Market leader.
    • ASML (Netherlands): Small but growing competitor.
  • Decline of US Companies:
    • GCA: Liquidated.
    • Silicon Valley Group (SVG): Technologically behind Canon and Nikon.
  • ASML’s Rise:
    • Strategic Sourcing: Assembled systems from globally sourced components, allowing them to utilize the best available parts.
    • Neutrality: Perceived as neutral in US-Japan trade disputes, attracting US customers like Micron.
    • Partnership with TSMC (Taiwan):
      • Philips (ASML’s parent company) had a strong relationship with TSMC.
      • ASML became a key supplier as TSMC’s fabs were compatible with Philips’ technology.

3. The Geopolitical Battle: Control and Access

  • Context:
    • The US emerged from the Cold War as a global superpower.
    • Globalization and interconnectedness were seen as inevitable and positive.
    • Intel dominated the microprocessor market.
  • Intel’s Consortium:
    • 1996: Intel partnered with US Department of Energy labs (e.g., Lawrence Livermore, Sandia) to advance EUV.
    • Goal: Transition EUV from scientific research to mass production.
  • Choosing ASML:
    • US firms (GCA defunct, SVG lagging) couldn’t commercialize EUV.
    • Government opposed Japanese involvement (Nikon, Canon) due to past trade tensions.
    • ASML, despite being foreign, became the only viable option.
  • Concerns and Justifications:
    • US Government Concerns: Giving a foreign company access to sensitive national lab research raised concerns about future access to critical technology.
    • Counterarguments:
      • No immediate military application for EUV.
      • EUV’s success was not guaranteed.
      • ASML (and the Netherlands) were seen as reliable partners.
      • Job creation through an ASML facility in the US was prioritized.
    • Industry Perspective: Focused on efficient semiconductor production; ASML was the only company capable of delivering EUV.
  • ASML’s Monopoly:
    • Nikon and Canon, excluded from US research, abandoned EUV development.
    • 2001: ASML acquired SVG (US’s last major lithography firm).
      • Concerns: Further consolidated EUV technology under ASML.
      • Justifications:
        • Necessary for EUV progress and the future of computing (argued by Intel CEO Craig Barrett).
        • Aligned with the Bush administration’s policy of promoting globalization and loosening export controls.
  • The Outcome:
    • EUV development continued with ASML as the sole supplier.
    • Concerns about US reliance on a foreign company for this critical technology were dismissed.
    • The narrative of globalization masked the reality of a single company monopolizing EUV lithography.

Chapter 33: The Innovator’s Dilemma

Intel’s Missed Opportunities

Intel’s Dominance and the x86 Architecture

  • By 2006, Intel supplied the processors for most PCs.
    • They achieved this by fending off AMD, the only other major company producing chips using the x86 instruction set architecture.
      • x86 instruction set architecture: A foundational set of rules governing how chips compute; the industry standard for PCs.
    • Apple was the only major computer maker not using x86-based chips until 2006 when they announced a switch to Intel chips.
  • Intel’s dominance in the PC market stemmed from IBM’s decision to use x86 processors in their first personal computers.
    • This positioned Intel as a controller of a crucial building block in the PC ecosystem, much like Microsoft with their operating system, Windows.
  • While considered complex and bulky compared to the RISC architecture, x86 became entrenched due to the cost of change and the threat to Intel’s dominance.

Andy Grove’s “Castle and Moat” Vision

  • In the early 1990s, then-CEO Andy Grove envisioned Intel’s future as a “castle and moat.”
    • The castle represented Intel’s profitability.
    • The moat was the x86 architecture, protecting the castle.
  • Grove considered, but ultimately rejected, switching to the more efficient RISC architecture.
    • The cost of abandoning the established x86 infrastructure was deemed too high.

Intel’s Server Chip Monopoly

  • Alongside their PC market dominance, Intel leveraged their manufacturing prowess to dominate the server chip market.
  • This market grew rapidly with the rise of large data centers and cloud computing.
  • Today, Intel and AMD, both using x86, control nearly the entire data center chip market.

The Rise of ARM and the Mobile Revolution

ARM’s Alternative Vision

  • ARM (Advanced RISC Machines), a joint venture founded by Apple and two partners in 1990, aimed to challenge x86 with a RISC-based architecture.
  • Robin Saxby, ARM’s first CEO, envisioned ARM as the global standard in chip architecture.
  • ARM’s Business Model:
    • Sell licenses for their architecture to fabless design firms.
    • These firms would customize ARM’s architecture, then outsource manufacturing to companies like TSMC (Taiwan Semiconductor Manufacturing Company).
  • This model directly challenged Intel’s vertically integrated approach.

ARM’s Early Successes and Intel’s Missed Opportunity

  • ARM initially struggled to penetrate the PC market due to the strength of the Intel-Microsoft partnership.
  • However, ARM’s energy-efficient architecture flourished in the mobile device market.
    • Nintendo adopted ARM-based chips for their handheld consoles.
  • Intel, blinded by the profitability of their PC market, failed to recognize the significance of mobile devices.
  • Despite early warnings from executives about the potential of mobile devices, Intel prioritized their highly profitable PC chip business.

The iPhone and the Consequences of Inaction

Apple’s Proposal and Intel’s Refusal

  • Steve Jobs approached Intel’s CEO, Paul Ottolini, to build a powerful, computer-style processor for their new product - the iPhone.
  • Intel, focused on profit margins and underestimating the iPhone’s potential, declined Apple’s offer.

The Rise of Apple and the Decline of Intel’s Mobile Ambitions

  • Apple turned to ARM architecture and Samsung (as the chip manufacturer) for the iPhone’s processor.
  • The iPhone’s success exceeded all expectations, with Apple’s profits from smartphones eventually surpassing Intel’s PC processor profits.
  • While Intel invested heavily in catching up in the mobile market, they failed to gain a significant foothold against Apple.

Intel’s “Innovator’s Dilemma”

  • Despite being familiar with Clayton Christensen’s concept of the innovator’s dilemma, Intel fell victim to their own success.
  • The Innovator’s Dilemma: The tendency of established, successful companies to focus on sustaining innovation for existing markets (and higher profit margins), thus neglecting potentially disruptive innovations in smaller, emerging markets.
  • Intel’s focus on short-term profit margins, driven by a management culture that prioritized financial engineering over technological leadership, blinded them to the transformative potential of mobile devices.

Key Takeaways:

  • Intel’s immense profitability in the PC and server markets became a liability.
  • Their prioritization of high-profit-margin products stifled innovation in new, potentially disruptive markets.
  • The company culture shifted from a focus on technological advancement to a focus on financial optimization.
  • This led to a failure to capitalize on the mobile revolution, costing them a chance to dominate another era of computing.

Chapter 34: Running Faster?

Andy Grove’s Concerns

  • In 2010, Andy Grove, former Intel chairman, expressed concern about the offshoring of advanced manufacturing jobs in Silicon Valley, despite the region’s economic success.

Observations Fueling Grove’s Concerns

  • Offshoring Trend: The iPhone, introduced in 2007, exemplified the offshoring trend, with few components built in the U.S.
  • Loss of Expertise: Grove worried that offshoring, which started with low-skilled jobs, would expand to encompass advanced manufacturing and engineering expertise.
  • Electric Vehicle Battery Industry: He cited the lithium battery industry for electric vehicles as a prime example, where the U.S. lagged significantly despite inventing core technologies.

Grove’s Proposed Solution

  • Levy an additional tax on products manufactured using offshored labor.
  • Be prepared to engage in a “trade war” to protect American manufacturing and expertise.

Dismissive Response to Grove’s Concerns

  • Many dismissed Grove’s concerns as outdated, arguing:
    • Bygone Era: Grove built Intel before the internet, and Intel missed the mobile phone revolution.
    • Reliance on Legacy Business: Intel’s success relied on its x86 processor monopoly, not disruptive innovation.
    • Shrinking Technological Lead: While Intel retained advanced semiconductor process technology, rivals like TSMC and Samsung were closing the gap.
    • Rise of New Tech Giants: Companies like Apple and Facebook, with different business models, surpassed Intel in valuation.
    • Semiconductor Industry Trends: The success of foreign semiconductor foundries like TSMC, primarily producing chips designed by American firms, was seen as a sign of a healthy ecosystem.
    • Historical Precedent: Offshoring to Southeast Asia had been a core part of the semiconductor industry since Fairchild Semiconductor’s Hong Kong assembly plant.

Grove’s Counterarguments

  • Loss of Future Opportunities: Grove argued that abandoning current commodity manufacturing could hinder entry into emerging industries, using the electric battery industry as an example:

    “Abandoning today’s commodity manufacturing can lock you out of tomorrow’s emerging industry,” he declared, pointing to the electric battery industry. (Source: Chapter 34)

  • U.S. Decline in Battery Technology:

    “The U.S. lost its lead in batteries 30 years ago when it stopped making consumer electronics devices,” Grove wrote. Then it missed PC batteries, and now was far behind on batteries for electric vehicles. “I doubt they will ever catch up,” he predicted in 2010. (Source: Chapter 34)

Strengths of the U.S. Semiconductor Ecosystem

  • Profitability of U.S. Chip Design Firms: Numerous American fabless chip design companies were highly profitable.
  • Dominance in Semiconductor Manufacturing Equipment:
    • Applied Materials: Remained the world’s largest semiconductor toolmaking company.
    • LAM Research: Held world-leading expertise in etching circuits onto silicon wafers.
    • KLA: Possessed the world’s most advanced tools for identifying nanoscale errors on wafers and lithography masks.
  • Critical Role of U.S. Equipment: Manufacturing advanced chips without American equipment was practically impossible.
  • Monopoly in Chip Design Software:
    • Cadence, Synopsys, Mentor: Three American firms controlled roughly 75% of the market for chip design software, making their products essential for chip development.
    • U.S. Dominance: Most smaller chip design software providers were also U.S.-based.

Risks of Offshore Manufacturing: The Taiwan Earthquake Example

  • 1999 Taiwan Earthquake: A 7.3 magnitude earthquake in Taiwan caused widespread power outages, impacting TSMC’s fabs and highlighting the risks of concentrated manufacturing.
    • TSMC’s Response: Morris Chang, TSMC’s founder, secured prioritized access to electricity, restoring most fabs within a week.
    • Limited Disruptions: Market disruptions were short-lived, with consumer electronics returning to normal within a month.
  • Seismic Vulnerability: While TSMC claimed earthquake resistance up to a magnitude of nine, the potential for stronger earthquakes remained a concern.

The “Run Faster” Strategy and Its Flaws

Changing Relationship Between Silicon Valley and the Pentagon

  • Reduced Reliance on the Pentagon: Silicon Valley giants, facing less direct competition from Japanese firms, decreased their engagement with the Pentagon in the 1990s and 2000s.
  • Shifting Priorities: The industry’s primary focus shifted to advocating for trade deals and deregulation, rather than government support.

The Rise of the “Run Faster” Consensus

  • Belief in Globalization’s Inevitability: Many officials in Washington believed that strict export controls were impractical and counterproductive in an increasingly globalized economy.
  • China’s Integration: China’s deep integration into the global economy further complicated export control efforts.
  • Focus on Engagement with China: U.S. policymakers prioritized building positive relations with China, viewing trade and investment as tools to encourage responsible behavior.
  • The “Run Faster” Doctrine: The prevailing strategy became to outpace rivals in technological innovation rather than restrict trade.

Evidence Challenging the “Run Faster” Approach

  • Van Atta’s Report (2007): A Defense Department study led by Richard Van Atta warned that offshoring threatened the military’s access to cutting-edge chips. This warning was largely ignored.
  • Historical Counterexamples: Past experiences, such as the rise of the Japanese semiconductor industry in the 1980s, demonstrated that simply “running faster” did not guarantee U.S. leadership.
  • Shrinking Leads: While the U.S. maintained advantages in chip design and manufacturing equipment, the gap was narrowing in key areas like advanced lithography and DRAM production.

Conclusion

Despite the perceived success of the U.S. semiconductor industry, Andy Grove’s concerns about offshoring expertise and the potential for a decline in U.S. technological leadership were supported by emerging evidence. The popular “run faster” strategy, while appealing in theory, lacked sufficient analysis and ignored warning signs.

Chapter 35: Real Men Have Fabs

The Semiconductor Industry in the 2000s

  • The semiconductor industry can be split into three categories:
    • Logic: Processors for smartphones, computers, and servers. Driven by Moore’s Law, requiring smaller transistors for better performance.
    • Memory: DRAM (short-term memory) and Flash/NAND (long-term memory). Also driven by Moore’s Law and the need to shrink transistors.
    • Other: Includes analog chips (sensors, RF chips, power management). Less dependent on Moore’s Law, with design being more important than shrinking transistors.
      • 75% of these chips are produced using older, cheaper 180nm technology (from the late 1990s).
      • This category has major players in the US, Europe, and Japan, with most production also located in these regions.
      • Texas Instruments is the largest analog chipmaker.

The Memory Market: A Shift to East Asia

  • DRAM:
    • Advanced fabs cost $20 billion.
    • The market consolidated from dozens of producers to three:
      • Micron (US) - with fabs in Japan, Taiwan, Singapore, and the US.
      • Samsung (South Korea)
      • SK Hynix (South Korea)
    • This consolidation and government subsidies led to most DRAM manufacturing shifting to East Asia.
  • NAND:
    • Similar to DRAM, the NAND market is also Asia-centric.
    • Major players include:
      • Samsung (South Korea) - largest player with 35% market share.
      • SK Hynix (South Korea)
      • Kioxia (Japan)
      • Micron (US) - some US production, but also fabs in Singapore and Japan.
      • Western Digital (US) - some US production, but also fabs in Singapore and Japan.

The Logic Chip Landscape: Rise of the Foundries

  • Advanced logic fabs also cost around $20 billion, leading to fewer companies able to afford them.
  • Intel remained a major player with its own fabs.
  • Many other US logic chipmakers, like Motorola and National Semiconductor, went bankrupt, were acquired, or saw their market share shrink.
  • This led to the rise of fabless firms, which designed chips in-house but outsourced manufacturing, primarily to TSMC.
    • This model lowered startup costs and allowed companies to focus on design.
  • Jerry Sanders, founder of AMD, championed the “real men have fabs” philosophy, believing that in-house manufacturing was crucial for success.
    • However, the industry was shifting towards a fabless model as foundries became more efficient and cost-effective.

Chapter 36: A Fabulous Revolution

The Rise of Fabless Chip Companies

  • Since the late 1980s, there has been significant growth in fabless chip firms.
  • Chips and Technologies (founded 1984), is considered the first fabless firm.
    • Though initially doubted, their success with graphics chips for PCs proved the viability of the fabless model.
  • Advantages of the fabless model:
    • Lower startup costs, requiring only a good idea and a few million dollars (compared to the billions needed for a fab).
    • Allowed companies to focus on their strengths in chip design.

NVIDIA: A Fabless Success Story

  • NVIDIA, founded in 1993, became the dominant player in graphics chips (GPUs).
  • Early on, they focused on 3D graphics, betting on its future importance.
  • Key Innovations:
    • Developed GPUs (Graphics Processing Units) specifically for handling complex 3D graphics.
    • Created a software ecosystem around their GPUs, including shaders for realistic image rendering.
    • Introduced CUDA in 2006, a software platform that allowed GPUs to be programmed for general-purpose parallel processing, expanding their use beyond graphics.
  • NVIDIA’s success is attributed to:
    • Focusing on their strength in chip design.
    • Outsourcing manufacturing to TSMC, allowing them to avoid massive fab investment costs and focus resources on R&D and software development.
    • Recognizing the potential of parallel processing beyond graphics, leading them to become a major player in AI.

Qualcomm: Another Fabless Powerhouse

  • Founded in 1985 by Erwin Jacobs, Qualcomm became a leader in mobile chip technology.
  • Key Contributions:
    • Pioneered frequency hopping technology for 2G cell phones, enabling more calls within limited spectrum space.
    • Developed key technologies for subsequent generations of cellular technology (3G, 4G, etc.), increasing data transmission capabilities.
    • Designed both modem chips (for connecting to cell networks) and application processors for smartphones.
  • Qualcomm’s success is also attributed to:
    • Focusing on their strengths in wireless communication technology and chip design.
    • Outsourcing manufacturing to companies like TSMC and Samsung, allowing them to avoid building their own expensive fabs.
    • Securing fundamental patents in mobile technology, generating significant revenue through licensing.

Other Notable Fabless Companies and their Innovations

  • Xilinx and Altera: Pioneered field-programmable gate arrays (FPGAs), chips that can be programmed for different uses.

The Impact of the Fabless Model

  • Enabled the emergence of new chip categories and applications.
  • Drove innovation in specialized logic chips, like GPUs for graphics and AI.
  • Made mobile devices, advanced graphics, and parallel processing possible.

Chapter 37: Morris Chang’s Grand Alliance

The Changing Landscape of the Chip Industry

  • By the 2000s, a new generation of CEOs led the semiconductor industry, focusing on financial metrics and shareholder value.
  • Morris Chang, founder of TSMC, remained a prominent figure, recognizing the transformative potential of smartphones.

TSMC’s Grand Alliance

  • Chang’s strategy was to create a Grand Alliance, a partnership between TSMC and various players in the semiconductor ecosystem:
    • Chip designers: Companies like Apple, Qualcomm, and AMD.
    • Intellectual property providers: Companies owning patents and designs essential for chip development.
    • Materials companies: Suppliers of silicon wafers, gases, and chemicals used in fabrication.
    • Equipment manufacturers: Producers of the sophisticated machinery used in chip fabs.
  • Benefits of the Grand Alliance for TSMC:
    • Access to the combined R&D capabilities of its partners, exceeding the spending of Samsung and Intel combined.
    • Setting industry standards for chip manufacturing, as compatibility with TSMC’s processes became essential for most chip designers.
    • Securing a central position in the semiconductor industry, with TSMC as the primary manufacturer for many leading chip design companies.

TSMC’s Investment Strategy During the Financial Crisis

  • During the 2008-2009 financial crisis, TSMC’s CEO Rick Tsai focused on cost-cutting, similar to many companies.
  • Chang, however, saw the crisis as an opportunity to invest and gain market share.
  • He dismissed Tsai and resumed direct control of TSMC.
  • Chang invested heavily in:
    • Expanding production capacity to meet the anticipated demand for smartphone chips.
    • Rehiring laid-off workers.
    • Doubling down on R&D to maintain technological leadership.
  • This strategy paid off, as TSMC emerged from the crisis stronger and better positioned to capitalize on the booming smartphone market.

The Decline of GlobalFoundries

  • GlobalFoundries, formed from AMD’s former manufacturing division, initially showed promise as a competitor to TSMC.
  • The company:
    • Inherited a large fab in Germany.
    • Began building a new advanced facility in New York.
    • Partnered with IBM and Samsung for technology development.
    • Benefited from the demand for a credible alternative to TSMC.
  • However, GlobalFoundries faced challenges:
    • Intense competition from TSMC, particularly in winning Apple’s business for iPhone chips.
    • The 2008-2009 financial crisis, which led to a slump in semiconductor demand.
    • Technical difficulties with its 28nm manufacturing process.
  • In 2014, GlobalFoundries acquired IBM’s microelectronics business, further expanding its size but not necessarily its technological edge.
  • Despite these moves, GlobalFoundries remained significantly smaller than TSMC:
    • In 2015, TSMC had 50% of the global foundry market, while GlobalFoundries and UMC each had around 10%.
    • TSMC’s wafer production capacity was 1.8 million per month, compared to GlobalFoundries’ 700,000.

The Adoption of EUV Lithography

  • The industry shift to EUV (Extreme Ultraviolet) lithography was crucial for continuing Moore’s Law and enabling the production of smaller, more powerful chips.
  • TSMC, Intel, and Samsung committed to adopting EUV, albeit with different strategies.
  • GlobalFoundries initially planned to adopt EUV but ultimately decided against it:
    • The company was already struggling with its 28nm process and had licensed its 14nm technology from Samsung.
    • The enormous cost of developing and implementing EUV was deemed too risky for GlobalFoundries’ financial position.
  • In 2018, GlobalFoundries halted its EUV program, deciding to focus on less advanced manufacturing nodes and prioritize profitability over cutting-edge technology.

The Consequences of GlobalFoundries’ Decision

  • The number of companies capable of manufacturing the most advanced logic chips decreased from four to three (TSMC, Intel, Samsung).
  • GlobalFoundries effectively removed itself from the competition to produce the most advanced chips, solidifying TSMC’s dominance.

Chapter 38: Apple Silicon

Apple’s Approach to Hardware and Software

  • From the beginning, Apple and Steve Jobs believed in the tight integration of hardware and software.
  • Jobs believed that software was for things changing too rapidly, not yet fully understood, or lacking time for hardware implementation.

Apple’s Transition to In-House Chip Design

  • The first iPhone (2007) used Apple’s iOS operating system but relied on chips from various suppliers, including Samsung.
  • Over time, Apple invested heavily in chip design:
    • Acquired PA Semi, a small chip design firm specializing in energy-efficient processing (2008).
    • Hired top chip designers from across the industry.
  • In 2010, Apple released its first in-house designed application processor, the A4, used in the iPad and iPhone 4.
  • Today, Apple designs its own application processors for most devices, as well as specialized chips for accessories like AirPods.

Apple’s Dependence on TSMC for Chip Fabrication

  • Despite designing its chips, Apple remains entirely reliant on TSMC for manufacturing.
  • The complexity and cost of advanced chip fabrication have led to a highly specialized industry, with TSMC possessing the necessary expertise and capacity.
  • Even though iPhones are “Designed by Apple in California, Assembled in China,” the most critical components, the processors, are fabricated solely in Taiwan by TSMC.

The Importance of TSMC in the Smartphone Supply Chain

  • Unlike the PC market, where Intel has long dominated processor manufacturing, the smartphone supply chain is heavily reliant on TSMC and, to a lesser extent, Samsung.
  • This reliance on foundries located in East Asia has significant geopolitical implications, given the proximity to China and its growing technological ambitions.

Chapter 39: EUV

The Long Road to EUV Lithography

  • ASML, a Dutch company, spent nearly two decades developing extreme ultraviolet (EUV) lithography.
  • EUV was a high-stakes gamble, requiring significant investment and technological breakthroughs.
  • Intel, Samsung, and TSMC invested billions in ASML to support EUV development, recognizing its importance for future chip manufacturing.

The Complexity of EUV Lithography

  • EUV lithography utilizes light with a wavelength of 13.5 nanometers, much shorter than previous lithography technologies, to etch smaller features onto silicon wafers.
  • This process requires highly specialized components and intricate engineering:
    • EUV light source:
      • Developed by CYMER, acquired by ASML.
      • Involves blasting tiny tin droplets with a powerful laser 50,000 times per second to generate EUV light.
      • Requires lasers with unprecedented power and precision, developed by the German company TRUMPF.
    • EUV mirrors:
      • Developed by the German company ZEISS.
      • Made of alternating layers of molybdenum and silicon, each a few nanometers thick.
      • Designed to reflect EUV light, which is easily absorbed by most materials.
      • Requires extreme precision and smoothness; irregularities are measured in fractions of a nanometer.
    • EUV system integration and software:
      • ASML manages a complex global supply chain to source and integrate components for its EUV tools.
      • Employs predictive maintenance algorithms and computational lithography techniques to ensure reliability and precision.

ASML’s Key Role and Global Collaboration

  • ASML doesn’t manufacture all components of its EUV tools in-house; it relies on a network of specialized suppliers, highlighting the global nature of the semiconductor industry.
  • ASML’s expertise lies in:
    • System integration: Assembling and fine-tuning components from various suppliers into a functional and reliable EUV tool.
    • Supply chain management: Maintaining strict quality control and ensuring the timely delivery of highly specialized components.
    • Software development: Creating software that manages the complex operation of EUV tools, including predictive maintenance and computational lithography.

The Significance of EUV’s Success

  • EUV’s successful development and deployment were crucial for extending Moore’s Law and enabling the continued miniaturization of transistors.
  • ASML’s EUV lithography tools became essential for manufacturing the most advanced chips, giving the company a near-monopoly in this critical technology.

Chapter 40: There is no Plan B

The Limits of Existing Lithography Technologies

  • By the mid-2010s, existing lithography technologies using deep ultraviolet (DUV) light were reaching their limits.
  • Tony Yen, a lithography expert at TSMC, emphasized that without EUV, there was “no plan B” for continuing to shrink transistors and advancing chip performance.

TSMC’s Commitment to EUV

  • Morris Chang and Xiangyi Qiang, head of R&D at TSMC, strongly believed in EUV’s potential and invested heavily in its development and implementation.
  • TSMC worked closely with ASML to test and improve EUV tools, recognizing the technology’s critical role in their future success.

GlobalFoundries’ Decision to Abandon EUV

  • Facing financial constraints and technical difficulties, GlobalFoundries decided to abandon EUV lithography in 2018.
  • The company:
    • Had already spent $1.5 billion on EUV development.
    • Estimated that bringing EUV-based manufacturing online would require billions more in investment.
    • Decided that the cost of developing and implementing EUV was too high, given their financial position and the competitive landscape.
  • This decision marked a turning point, as GlobalFoundries effectively conceded defeat in the race to produce the most advanced chips.

The Consequences of GlobalFoundries’ Withdrawal

  • The number of companies capable of manufacturing cutting-edge logic chips further decreased, leaving only three: TSMC, Intel, and Samsung.
  • This consolidation of manufacturing capacity in East Asia raised concerns about the global supply chain’s vulnerability and the geopolitical implications of relying on companies based near China.

Chapter 41: How Intel Forgot Innovation

Intel’s Dominant Position in the 2010s

  • Intel entered the 2010s as a dominant force in the semiconductor industry:
    • Held a near-monopoly in processors for PCs and data centers.
    • Possessed advanced manufacturing capabilities and a long history of innovation.
    • Invested heavily in R&D, spending billions annually.

Intel’s Missed Opportunities in Artificial Intelligence

  • Despite its strengths, Intel failed to capitalize on the rise of artificial intelligence (AI):
    • Its CPUs (Central Processing Units), while versatile, were not optimized for the parallel processing demands of AI workloads.
    • NVIDIA, originally a graphics chip company, recognized the potential of its GPUs (Graphics Processing Units) for AI and quickly gained market share.
    • Cloud computing giants like Google, Amazon, and Microsoft also began designing their own AI-optimized chips, further challenging Intel’s dominance.

Intel’s Failed Foray into the Foundry Business

  • In the mid-2010s, Intel attempted to compete with TSMC in the foundry business by opening its manufacturing facilities to outside customers.
  • This effort failed due to:
    • Cultural differences: Intel’s secretive and less customer-centric approach contrasted with TSMC’s open and collaborative model.
    • Lack of internal support: Intel prioritized its own chip design and manufacturing over its foundry business.
    • Competition from TSMC: TSMC’s established expertise, scale, and customer relationships proved difficult to overcome.

Intel’s Manufacturing Stumbles and the Delay of Moore’s Law

  • Compounding its other challenges, Intel experienced significant delays in its manufacturing roadmap:
    • Repeatedly announced delays in its 10nm and 7nm manufacturing processes, falling behind TSMC and Samsung.
    • Faced criticism for its lack of transparency regarding the reasons for these delays.
  • These delays were partly attributed to Intel’s slow adoption of EUV lithography, a technology the company itself had heavily invested in during its development.

The Consequences of Intel’s Decline

  • Intel’s struggles created an opportunity for TSMC and Samsung to further solidify their lead in advanced chip manufacturing.
  • The United States faced the prospect of losing its last remaining manufacturer of cutting-edge processors, with significant implications for its technological competitiveness and national security.

The Current State of the Semiconductor Industry

  • By the end of the 2010s, the semiconductor industry had undergone a dramatic transformation:
    • The rise of fabless chip design companies and the dominance of foundries like TSMC reshaped the industry landscape.
    • The successful development and implementation of EUV lithography proved crucial for continuing Moore’s Law.
    • Intel, once a dominant force, faced an uncertain future after missing key technological shifts and struggling with manufacturing delays.
  • The increasing concentration of advanced chip manufacturing capacity in Taiwan and South Korea raised geopolitical concerns, particularly regarding the reliance on companies located near China.

Chapter 42: Made in China

China’s Digital Paradox: Control vs. Vulnerability

Xi Jinping’s Vision of Digital Dominance

  • Xi Jinping, General Secretary of the Chinese Communist Party, declared in 2014: “Without cyber security, there is no national security. And without informatization, there is no modernization.”

  • Xi’s Background and Political Acumen

    • Son of an early Communist Party leader
    • Studied engineering in college
    • Rose through political ranks by adapting to different audiences:
      • To nationalists, he promised “national rejuvenation” and “great power status” through the “Chinese dream.”
      • To businesses, he pledged economic reform.
      • Some foreigners perceived him as a potential advocate for political reform.
    • Known for his political skill, concealing his true intentions behind a carefully crafted persona.

  • Xi’s Perception of the Digital World

    • Believed the digital world posed a primary risk to China.
    • Aimed to harness the internet for China’s power projection, contradicting Western beliefs in its democratizing potential.
    • Advocated for a global network under Chinese influence, embodied in his “one belt, one road” initiative, which included not only physical infrastructure but also network equipment and censorship tools.

China’s Success in Controlling the Internet

  • Implemented the world’s most effective system of internet control, using thousands of sensors to monitor online activity.
  • Erected the “Great Firewall,” blocking access to major international websites like Google and Facebook.
  • This control disproved Western predictions of the internet’s liberalizing influence in China.
  • Xi mocked the Western belief in the internet’s power to spread democratic values.
  • He stated, “The internet has turned the world into a global village,” while simultaneously restricting access to many global platforms within China.

Taming Tech Giants and Promoting Domestic Champions

  • China effectively tamed American tech giants:
    • Banned Google and Facebook
    • Fostered domestic alternatives like Baidu and Tencent, which rivaled their American counterparts technologically.
  • Granted access to the Chinese market to U.S. tech firms like Apple and Microsoft only after securing their cooperation with Beijing’s censorship demands.
  • Successfully made the internet subservient to the Communist Party’s agenda, compelling foreign internet and software companies to comply with censorship rules or face exclusion from the vast Chinese market.

China’s Achilles’ Heel: Dependence on Foreign Chip Technology

The Semiconductor Dilemma: China’s Hidden Vulnerability

  • Despite its successes, China’s control over its digital sphere masked a critical vulnerability: its dependence on imported semiconductors.
    • Chinese tech giants relied on data centers filled with foreign-made chips, mostly from the U.S.
  • Edward Snowden’s 2013 leaks revealed the extent of American network-tapping capabilities, surprising even Chinese cyber experts.
  • While China excelled at software for e-commerce, online search, and digital payments, it lagged in producing the underlying hardware.

Xi Jinping’s Assessment of the Semiconductor Risk

  • Xi acknowledged this dependence as a strategic weakness: “However great its size, however high its market capitalization, if an internet enterprise critically relies on the outside world for core components, the vital gate of the supply chain is grasped in the hands of others.” (2016)

  • Core Technologies of Concern:

    • Microsoft Windows: The dominant operating system in China, despite attempts to create domestic alternatives.
    • Semiconductors: The chips powering computers, smartphones, and data centers, seen as even more critical than software.

  • The Intel-Microsoft Dependency

    • Xi pointed out that Microsoft Windows relied on Intel chips, creating a dependence on American technology for most Chinese computers.

  • Economic Impact: China’s spending on semiconductor imports often exceeded that of oil, highlighting their importance in fueling economic growth.

  • Geopolitical Risk: Unlike oil, the semiconductor supply chain was dominated by China’s geopolitical rivals.

China’s Semiconductor Aspiration: Achieving Technological Independence

  • The Challenge of Perception: Many struggled to understand China’s anxieties, given its seemingly powerful tech sector.
  • Headline Hype vs. Reality: Media often portrayed China as a leading tech power, particularly in artificial intelligence.
  • “AI Superpower” Status: A book by Kai-Fu Lee, former head of Google China, labeled China as one of the world’s two AI superpowers due to its use of AI for surveillance.
  • Hidden Dependence: Even China’s advanced surveillance systems, used to monitor dissidents and minorities, relied on chips from American companies like Intel and NVIDIA.
  • The Fragile Foundation: All of China’s crucial technology rested on a foundation of imported silicon, a vulnerability acknowledged by Chinese leaders.

The Imperative for Domestic Chip Production

  • Beyond Supply Chain Security: Domestic chip production was not just about mitigating supply chain risks; it was also about capturing a larger share of the profits in the global technology market.
  • “Core Technologies”: China aspired to produce what its leaders termed “core technologies,” products essential to the global economy, enabling it to move beyond low-profit manufacturing.
  • The iPhone Example: China’s role in the iPhone production chain illustrated this challenge; millions of Chinese workers assembled the phones, but most profits went to Apple and chipmakers, not Chinese companies.

Learning from Other Asian Tech Powers

  • China sought to emulate the strategies of Japan, Taiwan, and South Korea, which had successfully entered the high-value segment of the chip industry. Their strategies included:
  1. Government Investment and Bank Lending: Pouring capital into domestic semiconductor companies with government backing and pressure on private banks to provide loans.
  2. Talent Acquisition: Attracting scientists and engineers trained in U.S. universities and Silicon Valley back to their home countries.
  3. Strategic Partnerships: Forming partnerships with foreign firms, but with stipulations requiring technology transfer or training for local workers.
  4. Exploiting Competition: Leveraging competition between Silicon Valley companies, and later between American and Japanese firms, to secure the most favorable deals.
  • Taiwan’s Success Story: KT Lee, a powerful Taiwanese minister, played a key role in establishing TSMC (Taiwan Semiconductor Manufacturing Company), a global leader in semiconductor fabrication. His statement to Morris Chang, TSMC’s founder, “We want to promote a semiconductor industry in Taiwan,” resonated with Xi Jinping’s ambitions for China.

Chapter 43: Call Forth the Assault

Xi’s Vision and the Davos Speech (January 2017)

  • Three days before Trump’s inauguration, Xi Jinping outlined China’s economic vision at the World Economic Forum in Davos.
    • Xi’s message: Win-win outcomes through an innovation-driven growth model.
    • Xi’s warning: “No one will emerge as a winner in a trade war,” a comment interpreted as a dig at Trump.
  • Three days later, Trump’s combative inaugural address contrasted with Xi’s message.
    • Trump’s stance: Protectionism for prosperity and strength.
    • Media portrayal:
      • Xi as a defender of globalization.
      • Xi’s Davos speech praised as a “robust defense of globalization” (Ian Bremmer, via Twitter).
      • Headlines highlighted Xi’s pro-globalization stance (Financial Times, Washington Post).
      • Klaus Schwab (World Economic Forum Chair): The “international community” is looking to China for leadership.

Xi’s Contrasting Message to Chinese Leaders

  • Months before Davos, Xi addressed a different audience in Beijing: Chinese tech leaders, PLA researchers, and the Party elite.
  • Xi’s message:
    • Focus on “breakthroughs in core technology,” specifically semiconductors.
    • “Promote strong alliances and attack strategic passes in a coordinated manner.”
    • “Assault the fortifications of core technology research and development.”
    • “Not only call forth the assault, we must also sound the call for assembly” to “concentrate the most powerful forces.”
    • Form “shock brigades and special forces to storm the passes.”
  • Analysis:
    • Xi’s language reveals a more aggressive approach than his Davos speech suggested.
    • Xi’s use of military metaphors signals a determined campaign for technological dominance.

China’s Technological Vulnerability

  • Growing dependence on foreign semiconductors: China’s semiconductor imports increased annually.
  • Shifting industry landscape:
    • “The scale of investment has risen rapidly and market share has accelerated to the concentration of dominant firms.” - China’s State Council technology policy report
    • Dominant firms (TSMC, Samsung) are difficult to compete with.
  • Exploding demand for chips: Driven by cloud computing, the Internet of Things, and big data.
  • China’s vulnerability:
    • Growing reliance on foreign chips, especially as China pursues AI.
    • Dependence on companies outside of China for advanced chips.
    • Staggering dependence on foreign technology at almost every stage of semiconductor production. This technology is primarily controlled by geopolitical rivals (Taiwan, Japan, South Korea, and the US).

China’s Dependence Across the Semiconductor Supply Chain

  • Software Tools:
    • Dominated by US firms.
    • China holds less than 1% of the global market share. Source: Georgetown University’s Center for Security and Emerging Technology
  • Core Intellectual Property (transistor patterns):
    • China’s market share: 2%.
    • The majority is held by US and British firms.
  • Silicon Wafers and Chip-Making Materials:
    • China supplies 4% globally.
  • Chip Fabrication Tools:
    • China supplies 1% globally.
  • Chip Designs:
    • China holds 5% of the market.
  • Chip Fabrication:
    • China’s market share: 7%.
    • Lacks capacity for high-value, leading-edge technology.
  • Overall Semiconductor Supply Chain Market Share:
    • China: 6%
    • US: 39%
    • South Korea: 16%
    • Taiwan: 12%
    • Source: Georgetown University researchers
  • Key Takeaway: China remains highly reliant on foreign, particularly American, technology for advanced semiconductors.

The “Made in China 2025” Plan

  • Goal: Reduce China’s reliance on imported chips.
    • From 85% (2015) to 30% by 2025.
  • Historical Context: Previous attempts to build a domestic chip industry had limited success.
    • Mao’s Cultural Revolution: Failed attempt at mass transistor production.
    • SMIC (Semiconductor Manufacturing International Corporation):
      • Founded with help from Richard Chang.
      • Faced financial struggles, IP lawsuits, and government interference.
      • Lagged behind TSMC in manufacturing.
    • Other Chinese foundries (Hua Hong, Grace):
      • Limited market share due to government interference and inefficient, geographically dispersed facilities.
  • Foreign Investment:
    • Potential seen in the Chinese market, but hampered by corporate governance issues and reliance on subsidies.
    • “When a Chinese firm said, let’s open a joint venture, … I heard, let’s lose money.” - European semiconductor executive
  • The “Big Fund” (2014):
    • Government-backed fund to support the semiconductor industry.
    • Key investors:
      • China’s Ministry of Finance
      • China Development Bank
      • State-owned firms (including China Tobacco)
      • Municipal governments (Beijing, Shanghai, Wuhan)
    • Estimated investment: Tens of billions of dollars.
  • Challenges:
    • China’s desire for self-sufficiency hindered collaboration with Silicon Valley.
    • Contrast with other Asian nations (Japan, South Korea, Taiwan) that integrated with the US chip industry.
  • Goal Misalignment:
    • Not simply seeking a larger role within the existing ecosystem.
    • Aiming to reshape the global semiconductor industry and reduce dependence on the US and its allies.

The Economic and Geopolitical Stakes

  • Trade Impacts:
    • Made in China 2025 threatened to disrupt global trade flows.
    • China’s chip imports in 2017: $260 billion.
      • Larger than Saudi Arabia’s oil exports or Germany’s car exports.
      • Exceeded the value of the global aircraft trade.
  • Regional Impact:
    • Potential harm to export-dependent economies in Asia.
    • Semiconductors as a significant percentage of exports:
      • South Korea: 15%
      • Singapore: 17%
      • Malaysia: 19%
      • Philippines: 21%
      • Taiwan: 36%
  • Taiwan’s Concerns: The rise of the “red supply chain”—mainland Chinese firms competing in areas Taiwan once dominated.
  • China’s Advantages:
    • Government subsidies.
    • State-backed IP theft.
    • Leverage over foreign firms seeking access to the Chinese market.
  • Conclusion:
    • Xi’s vision for semiconductor independence had the potential to reshape globalization and the semiconductor industry.
    • The plan’s success would have significant economic and geopolitical ramifications, particularly in Asia.

Chapter 44: Technology Transfer

The Drive for Self-Sufficiency

  • China’s goal: To acquire semiconductor technology and reduce reliance on foreign companies, particularly for data center chips.

IBM’s Technology Transfer

  • Context:
    • IBM’s sales in China dropped 20% after the Snowden leaks, suggesting government retaliation.
    • IBM sought to regain market access and leverage its lagging chip technology.
  • Strategy:
    • IBM CEO Rometty proposed opening chip technology to Chinese partners, enabling them to create a “vibrant ecosystem.”
    • This shift focused on selling services rather than hardware.
  • Key Individuals:
    • Rometty met with top Chinese officials, including Premier Li Keqiang and Vice Premier Ma Kai, who oversaw China’s chip industry.
    • Shen Chengsheng, former cybersecurity chief of China’s nuclear missile arsenal, worked with IBM’s chip technology, highlighting the strategic importance of the partnership.
  • Business Rationale:
    • IBM’s technology was considered second-rate and needed government support.
    • Sharing chip designs aligned with IBM’s global shift towards services.
  • China’s Perspective:
    • Viewed the partnership as supporting semiconductor self-sufficiency and national interests.
    • State-run media explicitly linked the deal to “integrated circuit development.”

Other Companies Involved in Technology Transfer

  • Qualcomm:
    • Faced pressure from Chinese regulators to lower licensing fees for smartphone chip technology.
    • Formed a joint venture with Huaqing Tong, a company with ties to a rising political figure, to develop server chips.
    • The venture dissolved in 2019, but expertise may have transferred to other Chinese firms.
  • AMD:
    • Struggled financially and sought cash while developing its “Zen” processor series.
    • In 2016:
      • Sold 85% of its semiconductor assembly, testing, and packaging facilities in Malaysia and China to a Chinese firm for $371 million.
      • Licensed the production of modified x86 chips for the Chinese market to a consortium of Chinese firms and government bodies.
    • Controversy Surrounding the Deal:
      • Intel reportedly warned the U.S. government about potential harm to U.S. interests.
      • The deal bypassed CFIUS review.
      • The Wall Street Journal criticized the deal as selling “crown jewels.”
      • Pentagon officials expressed skepticism and concerns about technology leakage.
    • Sugon’s Involvement:
      • The joint venture involved Sugon, a Chinese supercomputer firm with ties to the Chinese military.
      • Sugon advertised its role in “national defense and security.”
      • Even after being blacklisted by the U.S., Sugon was found using AMD chips, raising questions about continued technology access.

Arm China Spin-Off

  • Context:
    • Arm, a British company acquired by SoftBank (a Japanese firm heavily invested in Chinese tech), designed chip architecture.
  • The Deal:
    • In 2018, Arm spun off its China division, selling 51% to Chinese investors.
    • SoftBank, facing U.S. scrutiny over its China exposure, sold the stake for $775 million, a fraction of Arm’s $40 billion acquisition cost.
  • Rationale:
    • Arm executives acknowledged China’s desire for control over technology, particularly for sensitive applications like military and surveillance.
    • The spin-off allowed Arm to access the Chinese market while complying with China’s demand for domestic control.
  • Lack of Scrutiny:
    • Regulators in Japan, the UK, and the U.S. did not investigate the implications of the deal.

Conclusion

  • China’s large semiconductor market creates strong incentives for foreign firms to transfer technology, even if it risks aiding competitors.
  • Companies facing financial pressures or market share loss are particularly vulnerable to these incentives.
  • While individual deals may have sound business logic, the collective impact raises concerns about technology leakage and China’s growing semiconductor capabilities.
  • U.S. and UK chip architectures, designs, and Taiwanese foundries have played a significant role in advancing China’s supercomputer programs.
  • China has made progress in reducing reliance on foreign chipmakers, particularly for data centers.
  • IBM’s expectation of benefiting from technology transfer proved inaccurate, highlighting the potential for China to leverage such deals for its own strategic advantage.

Chapter 45: Mergers Are Bound to Happen

Zhao Weiguo’s Journey to Chip Billionaire

  • Zhao Weiguo’s path to becoming a chip billionaire, as celebrated by Chinese media, was unconventional.

Humble Beginnings and Early Career

  • Early Life:
    • Zhao’s family ended up in rural China after his father was exiled during the Cultural Revolution for writing subversive poetry.
    • Despite growing up raising livestock, Zhao aspired for more than a rural life.
  • Education and Early Career:
    • Zhao gained admission to Tsinghua University, one of China’s top institutions, where he studied electrical engineering.
      • Note: Tsinghua University has been at the forefront of China’s semiconductor efforts since the industry’s beginnings in the country.
      • It’s unclear how much semiconductor-specific knowledge Zhao acquired during his studies.
    • After graduating, he worked at a tech company before transitioning into investing as a vice president at Tsinghua Uni Group.
      • Tsinghua Uni Group: Established by Tsinghua University to commercialize its research, the company seemingly focused heavily on real estate investments.
  • Investment Success:
    • Zhao earned a reputation for skillful corporate deal-making, setting him on a path to substantial wealth.
    • 2004: Zhao founded his own investment firm, Beijing Jiankun Group.
      • Investment Focus: Real estate, mining, and other sectors where political connections are often critical for success.
      • Reported Returns: Zhao reportedly grew an initial investment of 1 million yuan to 4.5 billion yuan.

Entry into the Semiconductor Industry

  • 2009: Zhao acquired a 49% stake in his former employer, Tsinghua Uni Group, with the university retaining the remaining 51%.
    • Unusual Transaction: This raised eyebrows as it meant a private real estate investment firm now held significant control over a company meant to commercialize technology from a leading research university.
  • Tsinghua Unigroup’s Political Connections:
    • The company wasn’t a typical business entity. It had strong ties to the Chinese Communist Party:
      • The son of former Chinese President Hu Jintao, reportedly a personal friend of Zhao’s, held the position of Communist Party Secretary for the holding company that owned Unigroup.
      • The president of Tsinghua University during the 2000s was a former college roommate of Xi Jinping.
  • 2013: Zhao shifted focus to the chip industry, a move that coincided with the Chinese Communist Party’s announcement of major subsidies for domestic semiconductor companies.
    • Zhao’s Claim: He denies that Tsinghua Unigroup’s semiconductor strategy was a reaction to the government’s plans, stating, “Everyone thinks that the government is pushing the development of the chip sector, but it’s not like that” (Forbes, 2015).
    • Zhao’s Narrative: He presents himself as the driving force behind attracting Beijing’s attention to the sector: “Companies did some stuff first, and then the government started to notice, all our deals are market-oriented.”

Tsinghua Unigroup’s Aggressive Acquisition Strategy

Questionable Investment Approach

  • Analyst Skepticism: Most analysts wouldn’t describe Zhao’s strategy as “market-oriented.” Instead of targeting the top chip firms, he seemed to pursue any available acquisition.
  • Zhao’s Analogy: He compared his investment style to hunting: “If you carry your gun up the mountain, you just don’t know if there’s game there… Maybe you’ll catch a deer, maybe a goat, you just don’t know.”
  • Significant Spending: Despite Zhao’s estimated $2 billion fortune, his spending on building a chip empire was astonishing.

Early Acquisitions and Partnerships

  • 2013:
    • Domestic Acquisitions: Tsinghua Unigroup spent billions to acquire two of China’s leading fabless chip design companies:
      • Spreadtrum Communications
      • RDA Microelectronics
      • Both companies specialized in low-end chips for smartphones.
    • Zhao’s Justification: He predicted significant “synergies” from the merger, both domestically and internationally.
      • Note: Almost a decade later, there’s little evidence to support these claims of synergy.
  • 2014:
    • Intel Partnership: Tsinghua Unigroup partnered with Intel to combine Intel’s wireless modem chips with their smartphone processors.
      • Intel’s Goal: To expand its presence in the Chinese smartphone market.
      • Zhao’s Goal: For his companies to gain knowledge from Intel’s chip design expertise.
    • Zhao’s Stated Aims: He acknowledged that semiconductors were a national priority for China and believed the Intel collaboration would “accelerate the technology development and further strengthen the competitiveness and market position of Chinese semiconductor companies.”

Reckless Spending and Diversification

  • XMC Funding:
    • Background: Tsinghua Unigroup offered to finance XMC (later acquired by YMTC), a Chinese company aiming to enter the NAND memory chip market.
    • Excessive Investment: The XMC CEO admitted publicly that while he initially requested $15 billion for a new fabrication plant (“fab”), he was given $24 billion with the reasoning that “if they were going to be serious about being a world leader, then they needed to match the world leader’s investment.”
      • This demonstrates Tsinghua Unigroup’s tendency for excessive spending, even when it seemed economically unjustified.
  • Lack of Focus:
    • Beyond Semiconductors: News emerged that Tsinghua Unigroup was investing in other sectors like real estate and online gambling. This lack of focus further fueled concerns about the company’s strategy.
    • Government Backing: China’s state-backed “Big Fund” announced a plan to invest over $1 billion in Tsinghua Unigroup, effectively endorsing the company’s approach.

Targeting Global Semiconductor Leaders

Focus on Taiwan

  • Ambitious Goals: Zhao’s ambitions extended beyond controlling China’s fabless chip companies or attracting foreign investment into China. He set his sights on dominating the global semiconductor industry.
  • Recruiting Taiwanese Talent: He hired several prominent Taiwanese semiconductor executives, including the former CEO of UMC, Taiwan’s second-largest foundry.
  • 2015 Taiwan Visit:
    • Zhao traveled to Taiwan to advocate for lifting restrictions on Chinese investment in areas like chip design and fabrication.
    • He successfully acquired a 25% stake in Powertech Technology, a Taiwanese semiconductor assembler and tester, which was permissible under existing regulations.
    • He pursued stakes and joint ventures with other major Taiwanese chip assemblers.
  • Targeting MediaTek and TSMC: Zhao’s true targets were Taiwan’s industry giants:
    • MediaTek: The world’s leading chip designer outside of the United States.
    • TSMC: The dominant foundry upon which nearly all global fabless chip companies relied.
    • Proposed Acquisitions: He suggested buying a 25% stake in TSMC and merging MediaTek with Tsinghua Unigroup’s chip design operations.
      • These proposals were not legally permitted under Taiwan’s foreign investment laws.
  • Pressure Tactics: Upon returning from Taiwan, Zhao publicly suggested at a Beijing conference that China should ban imports of Taiwanese chips if Taiwan didn’t ease its investment restrictions.
  • Impact on TSMC and MediaTek: This pressure campaign put both companies in a difficult position due to their heavy reliance on the Chinese market.
    • Most of TSMC’s chips were used in electronics assembled in China.
    • The prospect of selling Taiwan’s key technology companies to a state-backed Chinese investor raised serious concerns about Taiwan’s economic independence.

Taiwan’s Response and Internal Debate

  • TSMC and MediaTek’s Cautious Statements:
    • Morris Chang (TSMC): He expressed conditional openness to the deal, stating, “if the price is right and if it is beneficial to shareholders.” However, he also cautioned, “if Chinese investors could appoint members to Taiwanese companies’ boards of directors, it will not be that easy to protect intellectual property.”
    • MediaTek: The company indicated support for collaboration “to join hands and raise the status and competitiveness of the Chinese and Taiwanese enterprises in the global chip industry,” but only if the Taiwanese government permitted it.
  • Taiwanese Government’s Position:
    • John Deng (Taiwan’s Economy Minister): He suggested loosening restrictions on Chinese investment in Taiwan’s chip sector.
    • Deng’s View: He believed that increased Chinese influence in Taiwan’s chip industry was unavoidable, stating, “You cannot escape from this issue.”
    • Delayed Policy Changes: Due to a contentious presidential election in Taiwan, the government postponed any immediate policy decisions.

Targeting US Semiconductor Companies

  • July 2015: Micron Acquisition Attempt:
    • Tsinghua Unigroup proposed buying Micron, a US memory chip manufacturer, for $23 billion. This would have been the largest-ever Chinese acquisition of a US company across any industry.
    • Micron’s Rejection: The company declined the offer, citing concerns about US government security reviews and deeming the deal unrealistic.
  • September 2015: Another Failed Attempt:
    • Tsinghua Unigroup offered $3.7 billion for a 15% stake in another US NAND memory chip company (name not specified in the text).
    • CFIUS Rejection: The Committee on Foreign Investment in the United States (CFIUS), which reviews foreign investments for national security risks, blocked the deal.
  • Spring 2016: Lattice Semiconductor Investment:
    • Tsinghua Unigroup quietly acquired a 6% stake in Lattice Semiconductor, another US chip company.
    • Zhao’s Claim: He insisted to the Wall Street Journal that this was “purely a financial investment,” adding, “We don’t have any intention at all to try to acquire Lattice.”
    • Quick Sale: Within weeks, Tsinghua Unigroup began selling its Lattice shares.
  • Canyon Bridge and Imagination Technologies:
    • Lattice Semiconductor Buyout Attempt: A California-based investment firm called Canyon Bridge, later revealed to be secretly funded by the Chinese government, made a bid to buy Lattice Semiconductor. This deal was also rejected by the US government.
    • Imagination Technologies Acquisition: Canyon Bridge simultaneously purchased Imagination Technologies, a struggling UK-based chip designer. The deal was structured to exclude Imagination’s US assets to avoid US regulatory scrutiny.
      • Notably, three years later, the new owners attempted to appoint board members linked to a Chinese government investment fund, raising red flags for British regulators who had initially approved the acquisition.

Controversial Tactics and Insider Trading

  • Illegal Practices: The issue wasn’t just Chinese government-linked entities acquiring foreign chip firms but doing so through methods that violated regulations on market manipulation and insider trading.
  • Canyon Bridge Insider Trading Case:
    • During the attempted Lattice Semiconductor purchase, a Canyon Bridge co-founder was found guilty of insider trading for sharing confidential deal information with a contact in Beijing via WeChat and meetings at a Starbucks in Beijing. This contact then used the information to profit from stock purchases.
  • Zhao’s Perspective:
    • Business-Focused: Zhao maintained that he was simply a dedicated entrepreneur, claiming that mergers between large US and Chinese companies were inevitable.
    • Zhao’s View: He believed such deals “should be viewed from a business perspective instead of being treated under nationalist or political contexts.”
  • Government-Led Effort: However, the sheer volume of Chinese state-owned and state-financed private equity firms pursuing global semiconductor companies pointed to a coordinated government-led strategy to gain control of these companies.
  • Following Directives: Xi Jinping had publicly called for an “assault” in the semiconductor sector. Tsinghua Unigroup’s actions, along with those of other government-backed investment vehicles, seemed to be a direct response to this directive.

Continued Investment and Expansion

  • 2017 Funding Announcement: Amidst its aggressive acquisition spree, Tsinghua Unigroup secured approximately $15 billion from the China Development Bank and $7 billion from the Integrated Circuit Industry Investment Fund, both entities owned and controlled by the Chinese government. This substantial injection of state funds further emphasized the Chinese government’s backing of Tsinghua Unigroup’s ambitions in the global semiconductor industry.

Chapter 46: The Rise of Huawei

Huawei’s Global Reach and Struggle

  • Huawei, a Chinese technology company, holds a crucial position in the global technology landscape.
    • Its telecom equipment forms the backbone of mobile internet infrastructure.
    • Its smartphone unit rivals Apple and Samsung in sales.
    • It provides various tech infrastructures like undersea cables and cloud computing.
  • Despite its global presence, Huawei faces scrutiny from the US national security state over its alleged ties to the Chinese government and concerns about espionage.

Comparing Huawei’s Trajectory to Samsung’s Success

  • To understand Huawei’s growth, it is helpful to compare it with Samsung, a South Korean tech giant that followed a similar path.
  • Both companies share a similar operating model:
    • Cultivate Political Relationships: Secure favorable regulations and access to capital.
    • Emulate and Outcompete: Identify products pioneered in the West, produce them at comparable quality, and offer them at lower costs.
    • Aggressive Globalization: Expand globally to reach new customers and learn from international competition.
  • Samsung’s success, with revenues equivalent to 10% of South Korea’s GDP, demonstrates the effectiveness of this model.
  • In contrast to Huawei’s global approach, most Chinese tech firms primarily focus on the domestic market, protected by regulation and censorship.

Huawei’s Business Model: Global Ambition and Competition

  • Ren Zhengfei, Huawei’s founder, adopted a distinct business model:
    • Adapting concepts developed abroad.
    • Manufacturing high-quality products at competitive prices.
    • Targeting the global market and directly competing with international rivals.
  • This strategy, similar to Samsung’s, positions Huawei to become a central player in the global tech ecosystem.

Huawei’s Early Days: From Reseller to Manufacturer

  • Ren Zhengfei’s Background:
    • Grew up in a family of teachers in Guizhou province, China.
    • Trained as an engineer and served in the Chinese army.
    • Moved to Shenzhen, a Special Economic Zone bordering Hong Kong, to pursue entrepreneurial opportunities.
  • Huawei’s Founding:
    • Established in 1987 with $5,000 in capital.
    • Initially imported and resold telecom switches from Hong Kong.
    • When suppliers cut him off, Ren decided to manufacture his own equipment.

Huawei’s R&D Focus and Allegations of Intellectual Property Theft

  • By the 1990s, Huawei invested heavily in R&D, focusing on building switching equipment.
  • While acknowledging some past instances, Huawei faces accusations of intellectual property theft:
    • In 2003, admitted to copying 2% of the code in one of its routers from Cisco.
    • Canadian media reports suggest possible Chinese government-backed espionage against Nortel, potentially benefiting Huawei.

Huawei’s R&D Investment and Shift in Ethos

  • Despite allegations, Huawei’s significant R&D spending suggests a shift from simply copying to innovation:
    • Annual R&D budget exceeding $15 billion, rivaling companies like Google, Amazon, and major pharmaceutical and automotive companies.
    • This commitment to R&D sets Huawei apart from many Chinese firms attempting to enter the chip industry cheaply.
  • Huawei claims its R&D focus stems from learning from Silicon Valley:
    • In 1997, Ren led Huawei executives on a US tour, visiting companies like HP, IBM, and Bell Labs, recognizing the importance of R&D and effective management.

Learning from IBM and Cultivating “Wolf Culture”

  • From 1999, Huawei hired IBM consultants to improve its operations:
    • Spent $50 million on consulting, a significant portion of its revenue at the time.
    • Employed over 100 IBM staff to revamp business processes, focusing on supply chain management, demand forecasting, marketing, and global sales.
  • Alongside Western consulting, Huawei adopted a “wolf culture”:
    • Emphasizing a militaristic work ethic and relentless pursuit of success.
    • This approach, while intense, aligns with the competitive spirit in the chip industry, as seen in the philosophies of leaders like Andy Grove and Morris Chang.

Government Support and Huawei’s Global Expansion

  • Huawei’s growth was supported by various levels of the Chinese government:
    • Receiving subsidies, state-backed credit, and tax breaks, amounting to an estimated $75 billion according to The Wall Street Journal.
    • While this level of support is unusual in the West, it is not uncommon for East Asian governments to back strategic industries.
  • The extent of state support raised concerns, especially in the US, about Huawei’s relationship with the Chinese government.
  • Chinese officials actively promoted Huawei’s global expansion, with Vice Premier Wu Banguo visiting the company and supporting its sales efforts in Africa.
  • This support raises questions about whether Huawei received preferential treatment or benefited from China’s mercantilist trade practices.
  • Concerns remain regarding the opaque nature of Huawei’s ownership structure and the role of the Chinese Communist Party in its governance.

Huawei’s Impact on the Global Telecom Market

  • Despite concerns, Huawei gained market share globally, impacting established Western telecom equipment providers:
    • Nortel filed for bankruptcy.
    • Alcatel-Lucent, inheritor of Bell Labs, was sold to Nokia.
  • Expanding beyond infrastructure, Huawei entered the smartphone market:
    • Its smartphones became top sellers, trailing only Samsung by 2019.
    • While making less profit per phone compared to Apple or Samsung, Huawei’s rapid rise challenged these giants.

The 2011 Japan Earthquake and Huawei’s Chip Design Ambitions

  • The 2011 earthquake and tsunami in Japan highlighted Huawei’s supply chain vulnerabilities:
    • The disaster threatened Huawei’s access to crucial components from Japanese suppliers.
    • While Huawei escaped major disruptions, it prompted a reassessment of its supply chain risks.
  • Huawei identified two critical dependencies:
    • Google’s Android operating system: The software foundation for its smartphones.
    • Semiconductor supply: Essential components for all its devices.
  • To mitigate risks, Huawei initiated in-house design of essential semiconductors, including:
    • Application processors for smartphones: Highly complex chips requiring advanced manufacturing.
  • Like Apple, Huawei outsourced the fabrication of these advanced chips to Taiwan’s TSMC, becoming its second-largest customer.

Huawei’s Challenge to US Chip Dominance and Preparedness for 5G

  • By developing in-house chip design capabilities, Huawei challenged the US dominance in this lucrative sector.
  • Huawei’s success mirrored the trajectory of Samsung and Sony, demonstrating its ability to:
    • Master advanced technologies.
    • Capture global market share.
    • Invest heavily in R&D.
    • Compete directly with American tech leaders.
  • This positioned Huawei advantageously for the rollout of 5G, the next generation of telecom infrastructure, and the era of ubiquitous computing it would enable.

Chapter 47: The 5G Future

The Evolution of Telecom and the Rise of 5G

  • Early Telecom:
    • Telephone switches were initially manual, requiring human operators to connect calls.
    • By the 1980s, electronic switches, often using semiconductors, replaced manual systems.
    • Even with these advancements, managing a building’s telephone lines required bulky equipment.
  • Modern Telecom:
    • Today’s telecom providers rely heavily on silicon.
    • Modern equipment, though still physically compact, can process calls, texts, and video, often sent via radio networks instead of landlines.
  • Huawei’s Mastery and the Significance of 5G:
    • Huawei has excelled in developing the latest equipment for transmitting calls and data over cellular networks, known as 5G.
    • 5G represents a significant leap in computing technology, relying heavily on semiconductors.
  • Generations of Mobile Networks:
    • The “G” in 5G refers to the generation of mobile networking standards.
    • Each generation has brought new hardware requirements for phones and cell towers.
    • Advancements in semiconductors, driven by Moore’s Law, have enabled more data transmission (ones and zeros) via radio waves.
  • Past Generations and Their Impact:
    • 2G: Introduced picture texts (MMS).
    • 3G: Enabled mobile web browsing.
    • 4G: Made video streaming widely accessible.
  • 5G’s Transformative Potential:
    • 5G is poised to deliver another major advancement in mobile network capabilities.
    • It will enable significantly faster data speeds and support a massive increase in connected devices.

The Crucial Role of Semiconductors in Mobile Networks

  • Impact of Semiconductors:
    • Advancements in semiconductors have made smartphones incredibly powerful.
    • Features like picture texts and video streaming, once considered revolutionary, are now commonplace thanks to these advancements.
  • Modem Chips:
    • Modem chips manage a phone’s connection with cell networks.
    • They allow the transmission of vast amounts of data (ones and zeros) as radio waves via the phone’s antenna.
  • Semiconductors in Cell Networks:
    • Similar advancements have occurred in chips used in cell towers and network infrastructure.
    • These advancements are crucial for managing data transmission, minimizing dropped calls, and ensuring smooth video streaming.
  • Challenges of Wireless Data Transmission:
    • Sending data wirelessly while maintaining quality and speed is complex.
    • The radio wave spectrum, the range of frequencies used for wireless communication, has limited space.
    • Not all radio wave frequencies are suitable for sending large amounts of data or transmitting over long distances.
  • The Importance of Spectrum and Semiconductors:
    • “Spectrum is far more expensive than silicon.” - Dave Robertson, chip expert at Analog Devices.
    • Semiconductors have been essential in maximizing the use of available spectrum to transmit more data wirelessly.
  • Innovations in Chip Design:
    • Companies like Qualcomm have developed advanced methods to optimize data transmission within the limited radio spectrum.
    • Manufacturers like Analog Devices produce specialized semiconductors called radio frequency transceivers, which send and receive radio waves with greater precision and lower power consumption.

5G Technology: Enabling the Future of Wireless

  • Increased Data Transmission:
    • 5G will significantly increase the amount of data that can be transmitted wirelessly.
    • It achieves this through:
      • Sophisticated Spectrum Sharing: Complex algorithms and increased processing power in devices and cell towers allow for more efficient use of existing spectrum space.
      • New Spectrum Utilization: 5G utilizes new, previously unused parts of the radio frequency spectrum, expanding available bandwidth.
  • Advanced Semiconductor Capabilities:
    • Advanced semiconductors in 5G networks enable:
      • Denser Data Encoding: Packing more ones and zeros into a given radio wave frequency.
      • Increased Range: Transmitting radio waves over longer distances.
      • Precise Targeting (Beamforming):
        • Beamforming allows cell towers to pinpoint a device’s location and direct radio waves specifically towards it.
        • This focused approach reduces power waste, interference, and improves signal strength for all users.
  • Beyond Faster Phones:
    • The impact of 5G extends far beyond faster smartphones.
    • It will revolutionize mobile computing and transform how we interact with technology.

The 5G Revolution: Connecting Everything and Generating Data

  • Shifting Expectations in Mobile Computing:
    • 1G: Mobile phones were a luxury.
    • 2G: Text messaging (SMS) became standard.
    • Present: Smartphones and tablets offer near-PC functionality.
    • 5G Future: Expect seamless connectivity for a vast number of devices, creating a truly interconnected world.
  • The Internet of Things (IoT) and Data Explosion:
    • 5G will connect countless devices to the internet, leading to an explosion of data.
    • This data, when processed effectively, will enhance products and services in countless ways.
  • Example: Smart Coffee Makers:
    • 5G-enabled coffee makers could collect and analyze data on temperature, brewing time, and coffee quality, leading to a consistently perfect cup.
  • Transforming Industries:
    • The combination of increased connectivity and data processing will revolutionize industries such as:
      • Agriculture: Optimizing farming practices (e.g., precision agriculture with connected tractors).
      • Manufacturing: Coordinating robots and automating processes on assembly lines.
      • Healthcare: Enabling remote patient monitoring and advanced diagnostics through connected medical devices and sensors.
  • Tesla: A Case Study in Connected Cars:
    • Tesla’s success highlights the potential of integrating advanced computing and connectivity into traditional products.
    • Tesla cars, often compared to smartphones on wheels, utilize custom-designed chips for a seamless user experience and advanced autonomous driving features.
  • The Future of Cars and Semiconductors:
    • The rise of electric vehicles (EVs), which require specialized semiconductors for power management, coupled with the demand for autonomous driving, will significantly increase the number and importance of chips in cars.

Huawei’s Rise and the Geopolitics of 5G

  • Huawei’s Leading Position:
    • Around 2017, as telecom companies prepared for 5G deployment, Huawei emerged as a dominant player.
    • Their equipment was regarded as high-quality and competitively priced.
    • Huawei was on track to play a larger role in 5G network construction than Ericsson (Sweden) and Nokia (Finland), its main competitors.
  • Huawei’s Semiconductor Reliance:
    • Despite its leadership, Huawei relied heavily on foreign semiconductor companies:
      • A Nikkei Asia analysis revealed that US-made chips, including those from Lattice Semiconductor (partially owned by China’s Tsinghua Unigroup at the time), Texas Instruments, Analog Devices, Broadcom, and Cypress Semiconductor, constituted almost 30% of the cost of Huawei’s radio systems.
      • While Huawei designed its main processor chip in-house, it was fabricated by TSMC (Taiwan Semiconductor Manufacturing Company).
  • China’s Semiconductor Ambitions:
    • Although not fully self-sufficient, Huawei’s advanced chip design capabilities, combined with the rapid growth of China’s semiconductor industry, suggested a future where China could challenge Silicon Valley’s dominance.
  • Geopolitical Implications:
    • This potential shift in the semiconductor landscape had significant implications:
      • It threatened to disrupt established tech companies and global trade flows.
      • It raised concerns about a potential shift in the balance of military power.

Chapter 48: The Next Offset

The Rise of China’s Military Power

  • The future of warfare hinges on computing power, impacting domains like drone swarms, cyberspace, and the electromagnetic spectrum.
  • The U.S. military’s unchallenged dominance is fading.
    • The U.S. no longer enjoys unrivaled access to global seas and airspace, a reality driven home by the 1991 Persian Gulf War.
      • This conflict demonstrated the effectiveness of precision strikes, prompting other nations, notably China, to bolster their military capabilities.
  • Over the past 30 years, China has heavily invested in high-tech weaponry, shifting from Mao-era low-tech strategies to embrace advanced sensors, communication, and computing.
    • China’s goal is not just to match the U.S. but to develop capabilities that offset American advantages. This strategy mirrors the Pentagon’s approach in the 1970s, now turned against them.

China’s Military Advantages: Undermining U.S. Strengths

  • Precision anti-ship missiles: These pose a significant threat to U.S. surface ships, especially in crucial areas like the Taiwan Strait, limiting American naval power projection.
  • Advanced air defense systems: These challenge America’s air superiority, traditionally a cornerstone of U.S. military doctrine.
  • Long-range land attack missiles: These missiles threaten the network of U.S. military bases spanning from Japan to Guam, potentially crippling American force projection in the Pacific.
  • Anti-satellite weapons: These weapons jeopardize U.S. communication and GPS networks, which are vital for coordinating military operations.
  • Cyber warfare capabilities: While untested in full-scale conflict, China’s cyber capabilities could potentially disrupt or disable entire U.S. military systems.
  • Electromagnetic spectrum dominance: China possesses the capability to jam American communications and blind surveillance systems. This could leave the U.S. military unable to effectively engage enemies or cooperate with allies.

“Intelligentized” Warfare: AI as the Future of Combat

  • The Chinese military believes warfare is transitioning from information-based to intelligentized warfare, leveraging artificial intelligence (AI) in weapon systems.
  • While computing power has always been crucial in warfare, its scale has grown exponentially. What’s new is the emergence of China as a credible challenger in this domain.
    • The Soviet Union, while capable of matching U.S. military might in certain areas, lacked the capacity for a similar level of technological advancement. China aims to rival the U.S. in both quantity and quality of military technology.
  • The race for computing power goes beyond mere commerce; it holds significant military implications. The country with superior computing power gains a distinct strategic advantage.

The AI Arms Race: China’s Potential to Surpass the U.S.

Factors Determining the Computing Race

  • 2021 Report by American Tech and Foreign Policy Experts:
    • The report, chaired by former Google CEO Eric Schmidt, predicted that China could become the world’s leading AI superpower.
    • Chinese leadership seems to agree with this assessment, emphasizing the development of military AI.
  • Xi Jinping’s Directive:
    • Xi Jinping has urged the PLA to prioritize the development of “military intelligentization,” further highlighting China’s focus on AI in warfare.

Applications of AI in Military Systems

  • Beyond Killer Robots: While the concept of AI weapons often brings to mind autonomous weapons, AI has broader military applications.
  • Predictive Maintenance: AI helps anticipate equipment failures, enhancing the operational readiness of planes, ships, and other assets.
  • Enhanced Threat Detection: AI-powered sonar systems in submarines and AI-enhanced analysis of satellite imagery improve the accuracy of threat identification.
  • Accelerated Weapons Development: AI can facilitate faster design and development of new weapon systems.
  • Improved Targeting Precision: AI allows for more accurate targeting, particularly for moving targets, enhancing the effectiveness of bombs and missiles.
  • Autonomous Vehicles: In the air, underwater, and on land, autonomous vehicles are being developed with AI capabilities for maneuvering, enemy identification, and target engagement.
    • While this builds on existing technologies like “fire-and-forget” missiles, the increasing autonomy and intelligence of these systems demand significantly more computing power.

China’s Strengths in the AI Race

  • Ben Buchanan’s Triad of AI: Georgetown University’s Ben Buchanan identifies three key elements for AI development: data, algorithms, and computing power. While the U.S. currently leads in computing power, China is rapidly catching up in the other two areas.
  • Data Access:
    • China’s Argument: Proponents of China’s AI development point to the country’s vast surveillance state and massive population as advantages for data collection.
    • Counterargument: The type of data collected through mass surveillance might not be as relevant for military AI applications.
    • Conclusion: Neither China nor the U.S. has a clear advantage in gathering militarily relevant data.
  • Algorithm Development:
    • China’s Strengths: China possesses a substantial pool of AI talent, with 29% of the world’s leading AI researchers being from China, compared to 20% from the U.S. and 18% from Europe (Marco Polo think tank).
    • US Strengths: The U.S. attracts and retains a significant portion of global AI talent, employing 59% of the world’s top AI researchers.
    • Potential Shift: Changes in U.S. visa policies, increased efforts by China to retain its researchers, and growing opportunities within China’s tech sector could erode America’s advantage in attracting top talent.
  • Computing Power:
    • Current Situation: While the U.S. maintains a lead in computing power, China’s dependence on foreign semiconductors, particularly U.S.-designed, Taiwan-manufactured processors, is a major vulnerability.
      • This dependence extends to data centers crucial for AI workloads.
      • A Chinese study estimates that 95% of GPUs in Chinese servers running AI applications are NVIDIA-designed, highlighting reliance on U.S. technology.
    • China’s Domestic Chip Production Efforts: While China is investing heavily in domestic chip production, experts predict it will be at least five years before they can design and several more before they can manufacture chips that rival the West.
  • China’s Acquisition of U.S. Chips: Despite export controls, the PLA has managed to acquire U.S.-designed chips for military use.
    • A Georgetown University study analyzing 343 PLA procurement contracts for AI-related systems found that less than 20% involved companies subject to U.S. export controls.
    • This indicates the effectiveness of China’s civil-military fusion policy, enabling the transfer of advanced civilian technology to military applications.

The Pentagon’s Response: A New Offset Strategy

  • Recognizing the Gap: The Pentagon acknowledges that China’s military modernization efforts have significantly narrowed the capability gap, especially in strategically important regions like the Taiwan Strait.
  • The Need for a New Offset: Inspired by the success of the 1970s offset strategy, which leveraged technology to counter the Soviet Union’s numerical advantage, the Pentagon is pursuing a similar approach to maintain military superiority over China.
    • Bob Work, former Deputy Secretary of Defense, stated that the U.S. will not engage in a direct numbers race but will instead focus on achieving a decisive technological advantage.
  • Focus on AI and Autonomy: The new offset centers around advancements in artificial intelligence (AI) and autonomy.
    • The 1970s offset focused on digital microprocessors, information technologies, new sensors, and stealth technology. Similarly, this new offset emphasizes emerging technologies to maintain a strategic edge.

Key Elements of the New Offset Strategy

  • Autonomous Platforms: The U.S. military is deploying autonomous vehicles like the SailDrone, an unmanned wind-powered vessel designed for long-duration maritime missions, including submarine tracking and communication interception.
    • These platforms are cost-effective compared to traditional naval vessels, allowing for wider deployment and greater coverage.
    • Development and deployment of autonomous surface ships, planes, and submarines are underway.
    • These platforms rely heavily on AI for navigation, decision-making, and mission execution.
  • Distributed Computing and Human-Machine Teaming:
    • DARPA’s Vision: DARPA (Defense Advanced Research Projects Agency) envisions a battlefield network of interconnected devices, from large ships to small drones, all capable of communication and coordination. This distributed computing approach aims to enhance situational awareness and decision-making.
    • Human-Machine Teaming: DARPA is investing in research on how human operators can effectively team with AI-powered systems. For instance, a fighter pilot could be assisted by a squadron of autonomous drones providing additional sensory input and combat support.

The Electromagnetic Spectrum: A New Battlefield

  • The increasing reliance on electronic sensors and communication systems makes the electromagnetic spectrum a critical domain in modern warfare.
  • Control over the electromagnetic spectrum allows militaries to:
    • Transmit critical information.
    • Detect and track enemy forces.
    • Disrupt enemy communications and sensor systems.
  • Examples of Electromagnetic Spectrum Warfare:
    • Russia’s Tactics in Ukraine: Russia has deployed radar and signal jammers to disrupt Ukrainian communications and hinder their military operations.
    • GPS Disruption: Reports suggest that Russia disrupts GPS signals around President Putin during his travels as a security measure, highlighting the vulnerability of satellite-based navigation systems.

U.S. Countermeasures:

  • DARPA’s Focus on Alternative Navigation: Recognizing the vulnerability of GPS, DARPA is researching alternative navigation systems that do not rely on GPS satellites. These systems would allow U.S. missiles to strike targets even if GPS systems are compromised.

The Role of Semiconductors in Electromagnetic Spectrum Warfare:

  • Semiconductors are essential for radar, jamming, and communication systems.
  • Radio Frequency (RF) Chips and Digital Analog Converters: These chips are crucial for:
    • Modulating signals to utilize available spectrum space efficiently.
    • Transmitting signals directionally.
    • Confusing enemy sensors.
  • Digital Signal Processing Chips: These chips run complex algorithms within radar and jamming systems to:
    • Analyze incoming signals.
    • Make real-time decisions on signal transmission.
  • Impact on Military Operations: Mastery of the electromagnetic spectrum is paramount for:
    • Maintaining situational awareness.
    • Coordinating forces effectively.
    • Denying these capabilities to adversaries.

DARPA’s Electronics Resurgence Initiative

  • Addressing the Challenge: The U.S. military’s reliance on electronics makes it imperative to maintain leadership in semiconductor technology, especially in the face of China’s growing capabilities.
  • Reviving Chip Innovation: DARPA launched the Electronics Resurgence Initiative (ERI) in 2017 to bolster the development of next-generation military-relevant chip technology.
  • Historical Context: DARPA has a history of funding groundbreaking research in microelectronics, but its influence has waned in recent decades as the commercial sector has grown.
  • Limited Influence:
    • DARPA’s budget is dwarfed by the R&D spending of major chipmakers like Intel and Qualcomm.
    • The U.S. government’s share of global chip purchases has shrunk, diminishing its leverage over the industry.
    • The immense cost of semiconductor manufacturing, particularly at the leading edge, is prohibitive even for the Pentagon.

The Pentagon’s Dependence on Foreign Chipmakers

  • Outsourcing Production: Both the U.S. military and intelligence agencies rely on commercial foundries for chip production.
    • While the U.S. maintains some domestic capacity, particularly for specialized analog and RF chips, it depends heavily on foreign companies, mainly TSMC in Taiwan and Samsung in South Korea, for advanced logic chips.
    • This dependence raises concerns about:
      • Security Risks: Potential for tampering, backdoors, or deliberate errors introduced during fabrication or assembly.
      • Supply Chain Vulnerabilities: Geopolitical tensions or disruptions in Taiwan or South Korea could severely impact U.S. access to advanced chips.

The “Zero-Trust” Approach to Microelectronics

  • Verifying Chip Integrity: DARPA is investing in technologies that can guarantee chip integrity and verify that chips are manufactured as intended. These measures are meant to mitigate the risks associated with reliance on foreign chipmakers.
  • Zero Trust Philosophy: The core principle is to trust no component or system by default and rigorously verify everything.
  • Implementation: This approach involves using technologies like:
    • On-Chip Sensors: Tiny sensors embedded on chips can detect tampering attempts or unauthorized modifications.
    • Advanced Verification Techniques: Sophisticated testing and analysis methods to ensure chips function exactly as designed and are free from vulnerabilities.

The Risks of the “Run-Faster” Strategy

  • Falling Behind: The U.S. strategy of simply outspending competitors in chip development, known as the “run-faster” strategy, has lost its effectiveness.
  • Dependence on Taiwan: The U.S. has become increasingly reliant on Taiwan for the production of cutting-edge logic chips, a strategic vulnerability given China’s claim over the island.
  • Intel’s Decline: Intel, once a global leader in chip manufacturing, has fallen behind competitors like TSMC, potentially undermining the U.S.’s position in the industry.
  • China’s Rise: China is aggressively investing in its domestic chip industry, aiming to achieve self-sufficiency and technological leadership.
  • Challenges to U.S. Dominance:
    • China’s aggressive pursuit of advanced chip technology.
    • The global interconnectedness of the electronics industry.
    • The mutual dependence of the U.S. and China on Taiwan for chip fabrication.

China’s Semiconductor Ambitions: A Challenge to U.S. Dominance

  • Call to Action: Chinese President Xi Jinping has called for China to become self-reliant in semiconductor technology, recognizing the strategic importance of this industry.
  • Multi-Pronged Strategy: China aims to reshape the global chip landscape through:
    • Acquisitions: Purchasing foreign chip companies to acquire technology and expertise.
    • Technology Theft: Engaging in industrial espionage and intellectual property theft to accelerate its technological progress.
    • Subsidies: Providing billions of dollars in government subsidies to support domestic chipmakers.
  • PLA’s Reliance on Domestic Chip Efforts: The Chinese military is counting on the success of these initiatives to reduce its dependence on foreign suppliers, particularly for its “military intelligentization” programs.
  • Continued Access to U.S. Chips: Despite U.S. export controls, China continues to acquire American chips through various means, including illicit channels.

Conclusion: A Critical Juncture in the Tech War

  • The competition for technological supremacy, particularly in semiconductors, is intensifying between the U.S. and China.
  • The outcome of this competition will have significant implications for:
    • Military dominance in the 21st century.
    • Global technological leadership.
    • The balance of power in the Indo-Pacific region and beyond.
  • The U.S. faces a critical challenge in maintaining its technological edge as China pours resources into closing the gap and reshaping the global semiconductor landscape. The stakes are high, and the outcome remains uncertain.

Chapter 49: Everything We’re Competing On

U.S. Chip Industry Concerns About China

  • Problem: China’s increasing push to dominate the global chip industry.
    • China is a critical market for U.S. chip firms.
      • Firms sell directly to Chinese customers.
      • Chips are assembled into devices in China.
    • China’s government adopted a policy to replace foreign chip firms with domestic ones.
    • U.S. chip firms were pressured to remain silent about Chinese subsidies.
  • Concerns:
    • China’s Massive Semiconductor Subsidies:
      • China pledged $250 billion to support domestic chip makers.
      • A U.S. official expressed concern, stating, “This massive $250 billion fund is going to bury us.”
    • Fear of Chinese Dominance:
      • U.S. officials were alarmed by the fear expressed by Intel CEO Brian Krzanich.
      • Concern arose from China’s success in driving U.S. solar panel manufacturing out of business.

Initial U.S. Government Response

  • Slow Response: Many senior officials in the Obama administration didn’t consider chips a crucial issue.
  • Shifting Perspectives:
    • Trade negotiators: Viewed China’s subsidies as a violation of international agreements.
    • Pentagon: Concerned about China’s use of computing power in weapons systems.
    • Intelligence agencies and Justice Department: Found evidence of China’s efforts to push out U.S. chip firms.
  • Obstacles to Action:
    • Deeply ingrained belief in globalization and the need to “run faster” (innovate more quickly).
    • Strong lobbying by the tech industry.
    • Lack of understanding about semiconductors among Washington officials.

Late Obama Administration Actions

  • Penny Pritzker’s Address (Late 2016):
    • Declared the importance of U.S. leadership in semiconductor technology.
    • Condemned China’s “unfair trade practices,” “massive non-market-based state intervention,” and attempts to acquire companies and technology for government interests rather than commercial objectives.
    • Ordered a study of the semiconductor supply chain.
    • Pledged to challenge China’s $150 billion industrial policy aimed at dominating the industry.
  • Semiconductor Industry Report:
    • Commissioned by the White House and issued in late 2016.
    • Urged the U.S. to “win the race by running faster” – focusing on innovation.

Critique of U.S. Tech Policy

  • Flawed Assumptions:
    • Globalization: The report claimed “unilateral action is increasingly ineffective in a world where the semiconductor industry is globalized.” This ignored the reality of Taiwan’s dominance in fabrication, not true globalization.
    • Technology Diffusion: The report stated, “Policy can, in principle, slow the diffusion of technology, but it cannot stop the spread.” This was not supported by evidence.
  • Consequences of Inaction:
    • Erosion of U.S. Lead: The U.S. lost its technological edge in fabrication and lithography due to a laissez-faire approach while Asian governments actively supported their chip industries.
    • Dependence on Vulnerable Chokepoints: The U.S. became reliant on a few key players, especially Taiwan, for critical semiconductor production.
  • Silicon Valley Complicity: The chip industry, fearing Chinese retaliation, downplayed the severity of the situation and advocated for continued engagement with China.
  • Dangerous Inertia: The focus on “multilateralism,” “globalization,” and “innovation” masked the deteriorating U.S. position and the rise of China.

The National Security Perspective

  • Growing Concerns:
    • China’s leverage over critical technology systems.
    • Potential for espionage through Chinese-manufactured electronics.
    • Reliance on foreign (particularly Chinese) companies for telecom equipment.
  • ZTE and Huawei:
    • Both Chinese telecom equipment providers raised concerns.
    • ZTE: State-owned.
    • Huawei: Private, but with alleged close ties to the Chinese government.
    • Both companies faced accusations of bribery and sanctions violations.
    • Obama Administration Action: Restricted U.S. firms from selling to ZTE in 2016. ZTE signed a plea deal, paid a fine, and the restrictions were lifted.
    • The ZTE case highlighted the global dependence on U.S. chips.

Trump Administration Shift

  • Focus on Technology: Trump’s focus on trade and tariffs was seen as a distraction from the technological competition with China.
  • China Hawks on the NSC:
    • Led by Matt Pottinger.
    • Believed America’s position had weakened.
    • Rejected “running faster” as a viable strategy.
    • Adopted a more aggressive, zero-sum approach.
  • Semiconductors as a Priority: The NSC recognized semiconductors as crucial to 21st-century competition, stating, “Everything we’re competing on in the 21st century, all of it rests on the cornerstone of semiconductor mastery.”
  • Government-Wide Focus: Various government entities, from Congress to the Pentagon, began prioritizing semiconductors in their China strategies.

Industry Friction and Dependence on China

  • Industry Discomfort: Chip industry leaders were wary of Chinese retaliation and preferred less disruptive government intervention.
  • Contradictory Messages:
    • Publicly: Semiconductor CEOs urged cooperation with China.
    • Privately: They admitted this strategy was failing and feared Chinese competition.
  • Dependence on China: The entire chip industry relied on sales to China, making them vulnerable to pressure from Beijing.
  • “Our Fundamental Problem”: One U.S. semiconductor executive stated to a White House official, “Our fundamental problem is that our number one customer is our number one competitor.”

NSC Intervention

  • Saving the Industry from Itself: The NSC believed the U.S. semiconductor industry would continue transferring technology and expertise to China unless drastic measures were taken.
  • Strengthening Export Controls: The NSC advocated for stricter export controls to prevent the transfer of critical chip-making technologies.

Shifting Government Focus

  • Media Focus on Trade War: Media attention was on Trump’s tariffs, but the national security bureaucracy saw this as a sideshow to the larger technological struggle.

The ZTE Case Revisited

  • Violation of Plea Agreement (April 2018): The Trump administration found ZTE in violation of its 2016 plea agreement.
  • Reimposition of Restrictions: The Commerce Department, led by Wilbur Ross, prohibited U.S. companies from selling to ZTE.
  • Trump’s Intervention: Trump saw ZTE’s potential collapse as leverage in trade negotiations with Xi Jinping and intervened to save the company.
  • ZTE Agreement: ZTE paid another fine and regained access to U.S. suppliers.
  • Differing Views:
    • Trump: Believed he gained leverage in trade talks.
    • China Hawks: Believed Trump was manipulated by officials like Treasury Secretary Steven Mnuchin.

Weaponizing Semiconductors

  • The ZTE Case as a Lesson: The ZTE incident revealed the United States’ power over global tech firms due to their reliance on U.S. chips.
  • Shifting Power Dynamics: Semiconductors were no longer just a cornerstone of competition but a potent weapon.

Chapter 50: Fujian Jinhua

The Micron Case

  • Kenny Wang and Intellectual Property Theft:
    • Kenny Wang, an employee at Micron’s Taiwan facility, downloaded 900 confidential files.
    • These files contained Micron’s trade secrets for DRAM chip production, including chip layouts, mask details, and yield information.
    • Wang transferred the files to a USB drive and uploaded them to Google Drive.
    • Micron estimated it would take years and millions of dollars to replicate the stolen information.
  • The DRAM Market:
    • Dominated by three companies:
      • Micron (U.S.)
      • Samsung (South Korea)
      • SK Hynix (South Korea)
    • High barriers to entry due to economies of scale and specialized expertise.

Jinhua’s Strategy

  • Fujian Jinhua:
    • A state-funded Chinese DRAM chip maker.
    • Received over $5 billion in government subsidies.
  • Partnership with UMC:
    • Jinhua partnered with Taiwan’s UMC (United Microelectronics Corporation), a logic chip maker, to gain DRAM expertise.
    • UMC received around $700 million from Jinhua.
  • Poaching Micron Employees:
    • UMC hired key personnel from Micron’s Taiwan facility, including Stephen Chen (former president) and J.T. Ho (process manager).
    • Kenny Wang joined UMC, bringing the stolen Micron files with him.

The Trump Administration’s Response

  • Escalation Beyond Statements: Unlike the Obama administration, the Trump administration was prepared to take concrete action.
  • Export Restrictions on Jinhua: The U.S. banned the sale of critical U.S.-made chip-making equipment to Jinhua.
  • Japanese Support: Japan, a key producer of chip-making equipment, supported the U.S. restrictions.
  • Chokehold on Jinhua: Without access to U.S. equipment, Jinhua’s production halted, effectively destroying China’s most advanced DRAM company.

Weaponizing Interdependence

  • U.S. Leverage: The case demonstrated the U.S. government’s ability to leverage its control over critical chokepoints in the semiconductor supply chain.
  • “Why the Fuck Wouldn’t We Do This”: Commerce Secretary Wilbur Ross, according to an aide, saw the use of export controls as a powerful and necessary tool.

Chapter 51: The Assault on Huawei

Huawei as a National Security Threat

  • “Spyway”: Trump labeled Huawei “the spyway” and accused it of espionage.
  • Beyond Espionage: The Pentagon and NSC saw Huawei as a strategic threat, representing China’s growing dominance in critical technology sectors.
  • Matt Turpin’s View: A Pentagon official, saw Huawei’s access to U.S. technology as a symptom of a larger problem in the U.S.-China tech relationship.
  • “Proxy for Everything We Had Done Wrong”: A senior Trump administration official said Huawei represented the failures in the U.S. approach to tech competition with China.

Global Concerns and Actions

  • Australia’s Ban: Australia banned Huawei from its 5G networks, citing unmitigable security risks.
  • Other Bans: Japan, New Zealand, and other countries followed suit.
  • European Responses:
    • Divided Approach: European countries were divided in their response to U.S. pressure to ban Huawei.
    • Outright Bans: Poland and other Eastern European countries banned Huawei.
    • Restrictions: France imposed strict limitations on Huawei.
    • Seeking Middle Ground: Germany faced Chinese pressure to not ban Huawei and attempted to find a compromise.
    • UK’s Initial Resistance: The UK initially resisted a ban, but later changed its stance.

The Debate Over Huawei

  • Technical vs. Strategic: The debate went beyond technical assessments of Huawei’s security risks to encompass broader strategic concerns about China’s role in global tech infrastructure.
  • Robert Hannigan’s Argument: The former head of the UK’s GCHQ (Signals Intelligence Agency) argued that China’s tech power was inevitable and the West should focus on managing the risks rather than trying to halt its rise.
  • U.S. Zero-Sum Approach: The U.S. adopted a zero-sum perspective, viewing Huawei’s success as detrimental to U.S. interests.

Escalating U.S. Actions

  • Initial Restrictions: The Trump administration initially banned U.S. companies from directly selling chips to Huawei.
  • Targeting Huawei’s Supply Chain: The administration then imposed broader restrictions, prohibiting the sale of any chips made using U.S. technology to Huawei, even if manufactured outside the U.S.
  • Exploiting Chokepoints: This strategy exploited the U.S. dominance in critical areas of chip design and manufacturing:
    • Software: Almost all chips rely on software from U.S. companies (Cadence, Synopsys, Mentor Graphics).
    • Fabrication: Advanced logic chips are primarily fabricated by TSMC (Taiwan) and Samsung (South Korea), both reliant on U.S. military support.
    • EUV Lithography: The most advanced chips require EUV lithography machines from ASML (Netherlands), which depend on critical components from their U.S. subsidiary.

Weaponized Interdependence in Action

  • “Weaponized Interdependence”: Academics Henry Farrell and Abraham Newman coined the term to describe the use of economic interdependence as a tool of coercion.
  • TSMC’s Compliance: Despite being Huawei’s largest customer, TSMC complied with U.S. restrictions.
  • Huawei’s Decline: The restrictions crippled Huawei’s smartphone and server businesses, forcing it to divest and delaying China’s 5G rollout.
  • Blacklisting Other Chinese Firms: The U.S. blacklisted other Chinese tech firms, including Sugon (supercomputers) and Phytium (high-performance computing).

China’s Response and the Future of the Tech War

  • Limited Retaliation: Despite threats, China has not taken significant retaliatory action against U.S. tech firms.
  • U.S. Escalation Dominance: The U.S. appears to hold the upper hand in the semiconductor sector, capable of inflicting greater damage through supply chain disruptions.
  • “A Beautiful Thing”: A former senior official, reflecting on the effectiveness of weaponized interdependence, called it “a beautiful thing.”

Chapter 52: China’s Sputnik Moment

Wuhan Lockdown and YMTC

  • Yangtze Memory Technologies Corporation (YMTC): China’s leading producer of NAND memory chips, located in Wuhan.
    • NAND memory is a type of chip used in smartphones, USB drives, and other devices.
    • Five companies currently make competitive NAND chips, none headquartered in China.
    • YMTC is considered China’s best chance to achieve world-class chip manufacturing capabilities in NAND production.
  • Government Support for YMTC:
    • Received at least $24 billion in funding from Xinhua Unigroup, China’s National Chip Fund, and the provincial government.
    • Allowed to operate during the Wuhan COVID lockdown, highlighting the government’s prioritization of the semiconductor industry.
      • Special train cars transported YMTC employees despite travel restrictions.
      • The company continued hiring even as the rest of the country remained in lockdown.

China’s “Sputnik Moment”

  • US Chip Sanctions: The US ban on selling chips to firms like Huawei is seen as a “Sputnik moment” for China, similar to the US’s reaction to the Soviet Union’s launch of Sputnik in 1957.
  • Impact of US Restrictions:
    • Dan Wang, a China tech policy analyst, argues that US restrictions have inadvertently boosted China’s pursuit of tech dominance.
      • Restrictions catalyzed new government policies and increased funding for the Chinese chip industry.
      • Without US pressure, Wang believes China’s “Made in China 2025” plan would have likely resulted in wasted resources.
  • Debate on US Strategy:
    • Should the US attempt to derail China’s chip ecosystem, risking an inevitable counter-reaction?
    • Or is it more strategic for the US to focus on domestic investment and hope China’s efforts falter?
  • China’s Response:
    • Increased government support for Chinese chipmakers, with Xi Jinping appointing Liu He as “chip czar” to manage semiconductor efforts.
    • Billions of dollars in subsidies for chip firms, although the effectiveness of this funding remains to be seen.

Challenges and Failures in China’s Chip Drive

  • Wuhan Hangxin (HSMC) Scam: An example of the risks associated with excessive subsidies without proper oversight.
    • A group of scammers deceived the Wuhan government into investing in their fake chip company, HSMC.
    • They used the funds to hire TSMC’s former head of R&D, acquired an ASML lithography machine, and raised more capital from investors.
    • The HSMC factory, a poorly constructed copy of an older TSMC facility, went bankrupt before producing a single chip.
  • Tsinghua Unigroup’s Financial Troubles:
    • The semiconductor investment fund experienced a cash shortage after a global acquisition spree and defaulted on some of its bonds.
    • Despite political connections, the firm required government intervention to avoid collapse.
  • Lack of Experience and Talent: A Chinese government official acknowledged the country’s lack of experience, technology, and talent in the chip industry.
  • Wasteful Spending: Billions of dollars have been wasted on unrealistic or fraudulent semiconductor projects in China.

Challenges to China’s Technological Independence

  • Multinational Supply Chains: Achieving complete technological independence in an industry with such complex and interconnected supply chains is incredibly difficult, even for the US.
    • China lacks competitive firms in key parts of the supply chain, including machinery and software, making independence even more challenging.
  • Replicating ASML’s EUV Machines:
    • EUV machines, crucial for producing advanced chips, took ASML almost three decades to develop and commercialize.
    • Replicating just the laser component requires assembling 457,329 parts flawlessly.
    • Even with stolen design specs, the complexity of these machines cannot be easily replicated without decades of engineering experience and expertise.
  • High Cost of Domestication:
    • Domesticating every aspect of the chip supply chain would be astronomically expensive.
    • The global chip industry spends over $100 billion annually on capital expenditures.
    • China would need to match this spending while also building a foundation of expertise and infrastructure, costing well over $1 trillion over a decade.

China’s Pragmatic Approach

  • Focus on Reducing US Reliance: China recognizes the impossibility of complete domestication and instead aims to lessen its reliance on the US in specific areas and increase its overall influence in the chip industry.
  • Embracing RISC-V Architecture:
    • RISC-V is an open-source instruction set architecture, offering a free alternative to the dominant x86 (Intel, AMD) and ARM architectures.
    • Its open nature potentially reduces security risks, fosters faster innovation, and appeals to Chinese companies seeking geopolitical neutrality.
    • Alibaba is among the Chinese firms designing processors based on RISC-V.
  • Leveraging Older Process Technology:
    • While smartphones and data centers demand cutting-edge chips, other devices like cars can function on older, cheaper technology.
    • China is investing heavily in lagging-edge logic chip production, a segment less vulnerable to US export restrictions on older equipment.
  • Investing in Emerging Materials:
    • China is focusing on silicon carbide and gallium nitride, materials with potential in electric vehicle power management systems.
    • Government subsidies in this area could give Chinese companies a competitive price advantage.
  • Growing Market Share:
    • China’s share of global chip fabrication is projected to rise from 15% to 24% by 2030, surpassing Taiwan and South Korea in volume, although potentially lagging behind technologically.
    • This increased market share could give China more leverage in demanding technology transfer and make it costlier for the US to impose export restrictions.

China’s Semiconductor Ambitions: National Goals over Profits

  • Chinese chip firms prioritize national goals over profit-making and going public.
  • Their focus is on building a domestic chip industry and realizing the “Chinese dream,” as stated by a YMTC executive.
  • Their dependence on government support aligns them with national objectives.

Chapter 53: Shortages and Supply Chains

The Global Chip Shortage

  • President Biden’s Meeting:
    • Convened a meeting with CEOs of chipmakers (Intel, TSMC), automakers (Ford, GM), and other companies to address the global chip shortage.
    • Emphasized the need for increased US investment in semiconductor R&D and manufacturing.
  • Causes of the Shortage:
    • Demand Surge: The shortage stemmed primarily from a surge in demand for electronics (PCs, servers, 5G phones) during the pandemic, not just supply chain disruptions.
      • Work-from-home and online trends fueled PC and data center demand.
    • Auto Industry Miscalculations:
      • Carmakers exacerbated the shortage by initially canceling chip orders, anticipating a sales slump that didn’t materialize.
      • Their “just-in-time” manufacturing practices offered little buffer against disruptions.
    • Chinese Stockpiling: Chinese firms, anticipating US sanctions, stockpiled chips, adding to the demand pressure.
  • Supply Chain Resilience:
    • Despite some disruptions (e.g., lockdowns in Malaysia affecting packaging), the semiconductor industry demonstrated remarkable resilience, producing a record 1.1 trillion chips in 2021 (a 13% increase from 2020).
    • The shortage highlighted the industry’s ability to adapt rather than its fragility.

Geopolitics and the Future of Chipmaking

  • Taiwan’s Strategic Importance:
    • Taiwan’s dominance in advanced chipmaking, epitomized by TSMC’s success, underscores the strategic importance of the semiconductor industry.
    • TSMC’s control over a critical chokepoint grants Taiwan significant leverage.
  • US Efforts to Secure Supply Chains:
    • Recognizing the vulnerability of relying on Asian chip production, the US is:
      • Pressuring TSMC and Samsung to build new fabs in the US.
      • Exploring ways to incentivize or pressure TSMC to locate its most advanced production in the US.
  • Global Competition for Chip Dominance:
    • The US, Europe, Asia, and China are all vying for a larger share of the chip market, particularly in high-value design and advanced fabrication.
    • This competition could lead to a realignment of the global semiconductor landscape.

South Korea’s Balancing Act

  • Government Support and Industry Leadership:
    • South Korea, led by Samsung and SK Hynix, benefits from strong government support and aims to maintain its leading position in memory chips while expanding in logic chip production.
  • Samsung’s Ambitious Investments:
    • Samsung plans to invest over $100 billion in its logic chip business by 2030, along with substantial investments in memory chips.
  • Pressure from Both Sides:
    • The US and China are both vying for South Korean chip investment.
    • Samsung faces US scrutiny over plans to upgrade its Chinese facilities and pressure to restrict the transfer of advanced EUV tools to SK Hynix’s facility in China.

Taiwan’s “Silicon Shield”

  • TSMC’s Continued Dominance:
    • Taiwan remains committed to protecting its chip industry, seeing it as a source of leverage and a “silicon shield” against Chinese aggression.
    • TSMC plans to invest over $100 billion between 2022 and 2024, mostly in Taiwan, while also upgrading facilities in Nanjing, China, and opening a new fab in Arizona.
  • Government Support and Strategic Importance:
    • The Taiwanese government actively supports TSMC through measures such as currency manipulation and trade negotiations.
    • TSMC’s founder, Morris Chang, serves as a trade envoy, advocating for free trade in semiconductors.

Other Players in the Global Chip Game

  • European Union:
    • Aspires to increase its share of advanced chip fabrication but faces challenges due to its low market share in advanced logic chips.
    • More realistically, Europe may focus on attracting foreign chipmakers to build fabs that can provide a stable supply for European automakers.
  • Singapore:
    • Continues to attract chipmaking investments with substantial incentives, recently securing a $4 billion investment from GlobalFoundries for a new fab.
  • Japan:
    • Seeking to revitalize its chipmaking industry, Japan is subsidizing TSMC to build a new facility in partnership with Sony.
    • This move aims to prevent the further erosion of Japan’s expertise in chipmaking equipment and materials.

Intel’s Bid for Resurgence

  • New Leadership and Ambitious Strategy:
    • Under CEO Pat Gelsinger, Intel is pursuing a three-pronged strategy to regain its leadership position:
      1. Manufacturing Leadership: Aiming to overtake Samsung and TSMC in advanced chipmaking by securing the first next-generation EUV machine, expected in 2025.
      2. Foundry Business: Launching a foundry business to manufacture chips for other companies, directly competing with Samsung and TSMC. Requires securing customers producing at the cutting edge to be viable.
      3. Government Support: Leveraging US and European government subsidies to build new fabs domestically while promoting concerns about reliance on Asian chip production.
  • Dependence on TSMC: Ironically, while advocating for a more balanced supply chain, Intel is increasingly outsourcing its own advanced chip production to TSMC in Taiwan.

The Taiwan Dilemma: Risk and Interdependence

  • TSMC’s Vulnerability: TSMC’s dominance makes it a strategic target in potential conflicts between China and Taiwan.
    • TSMC’s chairman downplays the risk of war, highlighting the economic interdependence and shared interest in a peaceful Taiwan Strait.
    • However, China’s military exercises near Taiwan and its stated goal of “reunification” raise concerns about the long-term stability of the region.
  • China’s Military Capabilities:
    • The PLA possesses a range of capabilities to pressure or attack Taiwan, including amphibious assaults, blockades, and air and missile strikes.
    • The military balance in the Taiwan Strait has shifted in China’s favor.
  • The “Salami Slicing” Strategy: China could attempt to gradually assert control over Taiwan through incremental actions, such as seizing outlying islands or imposing limited blockades.
  • US Dilemma: The US faces a difficult choice in responding to Chinese aggression against Taiwan.
    • Direct military intervention carries significant risks, including escalation to a larger conflict.
    • Inaction could embolden China and undermine US credibility, potentially leading to a scenario where China gains influence over TSMC’s chip production.

Chapter 54: The Taiwan Dilemma

TSMC and the Global Chip Supply

  • Investor Concerns: The growing geopolitical tensions surrounding Taiwan have made investors increasingly aware of the risks to TSMC’s operations.
  • TSMC’s Reassurance: TSMC’s chairman downplays the risk of war, emphasizing the global economic interdependence and the shared interest in a stable Taiwan Strait.
  • China’s Military Exercises: Despite reassurances, China’s military conducts exercises near Taiwan, simulating amphibious assaults and highlighting the potential for conflict.

Scenarios for Conflict

  • D-Day Style Invasion:
    • While China could attempt a full-scale amphibious invasion of Taiwan, this option is considered the most challenging and least likely due to the high risk of failure.
  • Blockade:
    • A partial air and maritime blockade would be difficult for Taiwan to overcome alone.
    • Even with US and Japanese support, breaking a blockade would be challenging and could escalate into a larger war.
  • Air and Missile Campaign:
    • A sustained air and missile campaign could cripple Taiwan’s military and economy without requiring a ground invasion.
  • Limited Military Pressure:
    • China could apply pressure through tactics like harassing shipping traffic or demanding TSMC produce chips for Chinese companies, testing US resolve and potentially eroding Taiwan’s morale.

The Stakes: Global Economic Impact of a Taiwan Crisis

  • Semiconductor Dominance: Taiwan plays a critical role in the global chip supply chain:
    • Produces 11% of the world’s memory chips.
    • Fabricates 37% of the world’s logic chips, essential for computers, smartphones, and data centers.
  • Economic Devastation: Disrupting Taiwan’s chip production would have severe global consequences:
    • Computing Power: A 37% reduction in global computing power, impacting everything from smartphones to data centers.
    • Electronics Manufacturing: Delays and shortages in all types of electronics, from PCs and smartphones to cars and appliances.
    • 5G Rollout: Near halt to the rollout of 5G networks due to reliance on Taiwan-made chips.
    • Economic Cost: Trillions of dollars in lost economic output, potentially exceeding the impact of the COVID-19 pandemic.
  • Difficult to Replace: Replicating Taiwan’s chipmaking capacity would take years:
    • Building new fabs would require significant time and investment.
    • Finding trained personnel and acquiring essential machinery and materials would be challenging, especially in a crisis scenario.

Taiwan’s “Silicon Shield” - Protection or Liability?

  • Deterrence: Taiwan’s chip industry serves as a “silicon shield,” making the US and its allies more invested in Taiwan’s security.
  • Vulnerability: The concentration of chip production in Taiwan also creates a single point of failure, making the global economy vulnerable to disruptions in the event of a conflict.

Lessons from the Russia-Ukraine War

  • Semiconductors as a Strategic Asset: The war highlights the strategic importance of semiconductors in modern warfare.
  • Supply Chain Leverage: The US and its allies have used their control over the chip supply chain to impose significant costs on Russia.
  • China’s Awareness: China is acutely aware of the leverage the semiconductor industry provides, leading to calls for securing control over TSMC.

Growing Risks in the Taiwan Strait

  • Military Buildup: China continues to enhance its military capabilities in the Taiwan Strait, increasing the potential for conflict.
  • Escalation Potential: Any future crisis in the Taiwan Strait would be far more dangerous than previous confrontations, with higher stakes and greater potential for escalation.
  • The Digital Battlefield: A conflict over Taiwan would have profound implications for the global technology sector, potentially disrupting supply chains and reshaping the balance of power in the digital world.

Conclusion

The Enduring Legacy of the Chip

  • Historical Context: The invention and development of the integrated circuit were deeply intertwined with Cold War rivalries and the pursuit of technological superiority.
  • Transformative Impact: Chips have revolutionized countless aspects of modern life, from communication and entertainment to transportation and healthcare.
  • Globalized Industry: The semiconductor industry evolved into a highly globalized enterprise, with countries and companies interconnected through complex supply chains.

Challenges to Moore’s Law

  • Physical Limits: As transistors shrink to the atomic level, fundamental physical limits may eventually slow down or halt the pace of miniaturization.
  • Cost and Complexity: The cost and complexity of developing and manufacturing ever-more advanced chips continue to increase.

The Future of Computing

  • Specialization: The trend toward specialization, with chips optimized for specific tasks like AI and graphics, is likely to continue.
  • Cloud Computing: Cloud computing will continue to reshape the industry, with companies like Amazon and Google designing their own chips.
  • New Architectures: The industry is exploring new chip architectures and materials to overcome the limitations of traditional silicon-based technology.

Conclusion

  • Chips and Geopolitics: The semiconductor industry will remain a focal point of geopolitical competition, with countries vying for control of this critical technology.
  • Taiwan’s Precarious Position: Taiwan’s dominance in chipmaking makes it both strategically important and vulnerable, highlighting the delicate balance of power in the region.
  • Uncertainty and Opportunity: The future of computing is full of uncertainty but also holds immense potential for innovation and progress.

Conclusion

The Birth of an Industry: From Cold War Tensions to Silicon Valley

  • 1958: A pivotal year for the semiconductor industry unfolds against the backdrop of the Cold War and U.S.-China tensions:
    • The shelling of Kinoy Island by the People’s Liberation Army underscores the escalating global conflict.
    • Jack Kilby demonstrates the world’s first integrated circuit at Texas Instruments in Dallas.
    • Jay Lathrop joins Texas Instruments, having already filed a patent for photolithographic transistor production.
    • Morris Chang arrives at Texas Instruments, renowned for his expertise in minimizing semiconductor fabrication errors.
  • Texas Instruments undergoes a strategic shift under Pat Haggerty’s leadership:
    • Haggerty’s vision focuses on developing electronics for military applications, moving away from oil exploration instruments.
    • Weldon Word and a team of skilled engineers at Texas Instruments work on electronics for advanced weapons and sensors.
  • The Cold War fuels U.S. investment in electronics:
    • Defense spending surges into electronics companies, driven by the need to maintain technological superiority over the Soviet Union and Communist China.
    • The emphasis shifts from traditional military might to technological advancement, particularly in transistors, sensors, and communication equipment.

Global Talent Converges on American Soil:

  • The lure of opportunity and freedom attracts brilliant minds to the United States:
    • Morris Chang, fleeing potential persecution in China, exemplifies the exodus of talent seeking refuge and opportunity.
    • John Bardeen and Walter Bratton’s invention of the transistor at Bell Labs leads to further breakthroughs.
    • Mohamed Atala and Dewan Kang, Bell Labs colleagues, develop a mass-producible transistor structure.
  • Fairchild Semiconductor, a pivotal company in the industry’s rise, benefits from international expertise:
    • Two of the “traitorous eight” founders, who left Shockley Semiconductor to start Fairchild, were born outside the U.S.
    • András Groff, a Hungarian émigré, plays a crucial role in optimizing Fairchild’s chipmaking processes and rises to become CEO.
  • Key semiconductor hubs emerge:
    • Texas, Massachusetts, and California become magnets for global talent drawn to the burgeoning semiconductor industry.

The Visionaries and the Rise of Silicon Valley:

  • Miniaturization as a Catalyst for Change:
    • Engineers and physicists share a conviction that shrinking transistors holds transformative potential, a belief that exceeds even their wildest expectations.
  • Gordon Moore’s Prescient Predictions:
    • In 1965, Gordon Moore’s prediction about home computers and portable communication devices foreshadows the centrality of chips in modern life.
    • The concept of producing more transistors daily than the number of cells in the human body would have been unimaginable to the industry’s pioneers.

From Defense to Global Markets: The Ever-Increasing Need for Scale:

  • The Semiconductor Industry’s Dependence on Global Markets:
    • The industry’s reliance on vast global markets grows alongside the increasing scale of production and shrinking transistor size.
    • Even the Pentagon’s $700 billion budget proves insufficient to support domestic production of cutting-edge defense chips.
  • Shifting Procurement Strategies:
    • The Defense Department, while possessing dedicated facilities for large-scale military equipment, relies on commercial chip suppliers, many based in Taiwan.
    • The exorbitant cost of designing and fabricating advanced logic chips necessitates a shift towards external procurement.
  • Silicon Valley’s Multifaceted Success:
    • Beyond scientific and engineering prowess, Silicon Valley’s success hinges on sales, marketing, supply chain management, and relentless cost reduction.

Entrepreneurship, Innovation, and the Drive for Efficiency:

  • Bob Noyce:
    • An MIT-trained physicist, Bob Noyce’s entrepreneurial vision and ability to identify a market for a yet-to-exist product were instrumental in the industry’s growth.
  • Fairchild Semiconductor:
    • Gordon Moore, in his 1965 article, highlighted how Fairchild Semiconductor’s success in integrating more components onto chips stemmed from a combination of:
      • Scientific and engineering talent
      • Effective management practices, such as maintaining union-free factories and implementing employee stock options, that boosted productivity.
  • Dramatic Cost Reduction:
    • The price of transistors plummets to less than a millionth of their 1958 cost thanks to the relentless drive for efficiency epitomized by a Fairchild employee’s exit survey response: “I want to get rich.”

The Interplay of Technology, Society, and Geopolitics:

  • Shaping the Semiconductor Landscape:
    • The development and trajectory of the semiconductor industry are not solely driven by technological advancements; societal and political forces play a significant role.
  • DARPA’s Influence:
    • DARPA, the Pentagon’s research and development arm, has significantly influenced the industry by funding research into FinFETs, the 3D transistor structures used in advanced chips.
  • China’s Impact:
    • China’s substantial subsidies are poised to reshape the semiconductor supply chain, regardless of whether China achieves its goal of semiconductor dominance.

The End of Moore’s Law? Challenges and the Future of Computing Power

The Limits of Miniaturization:

  • The Potential End of Moore’s Law:
    • Gordon Moore’s Law, a prediction, not a law of physics, faces challenges as shrinking transistors approach physical limits.
  • Industry Leaders Voice Concerns:
    • Prominent figures, including NVIDIA CEO Jensen Huang and former Stanford president John Hennessy, have declared Moore’s Law to be over.
  • Rising Costs and Slowing Progress:
    • The cost of producing ever-smaller chips continues to rise as the pace of cost decline slows significantly.
    • Tools required for manufacturing, such as EUV lithography machines with a price tag exceeding $100 million, contribute to the escalating expenses.

Historical Challenges and Overcoming Barriers:

  • Past Predictions of Moore’s Law’s Demise:
    • This is not the first time Moore’s Law has been declared near its end. In 1988, IBM expert Eric Block predicted its demise when transistors reached a quarter of a micron, a barrier surpassed by the industry a decade later.
  • Gordon Moore’s Own Concerns:
    • In 2003, Gordon Moore himself acknowledged potential barriers to continued progress within the following decade.
  • Overcoming Technological Hurdles:
    • Despite anticipated limitations, the industry has consistently overcome obstacles. The development and widespread use of 3D FinFET transistors exemplify this ability to innovate.
  • Exceeding Expectations:
    • Carver Mead, the originator of the term “Moore’s Law,” once astounded experts with his prediction of chips containing 100 million transistors per square centimeter.
    • Modern fabs can now pack hundreds of times more transistors onto a chip than Mead envisioned.
  • Jim Keller’s Optimistic Outlook:
    • Renowned chip designer Jim Keller envisions a 50-fold increase in transistor density:
      • Thinner fin-shaped transistors could allow for a threefold increase in density.
      • Tube-shaped gate-all-around transistors could double density by enabling electric field application from all directions.
      • Stacking these wires could further increase density eightfold.

A New Era of Chip Design and the Evolution of Computing:

  • Sustained Investment and Innovation:
    • The semiconductor industry continues to attract substantial investment, demonstrating a belief in continued advancement.
    • AI chip startups have raised billions, aspiring to emulate NVIDIA’s success.
    • Tech giants, including Google, Amazon, Microsoft, Apple, Facebook, and Alibaba, are investing heavily in custom chip design.
  • The Rise of Specialized Chips:
    • The traditional dominance of general-purpose microprocessors is being challenged by chips optimized for specific tasks, like NVIDIA’s GPUs for graphics and AI.
  • Neil Thompson and Svenja Spanut’s Argument:
    • These researchers argue that we are witnessing a decline in computers as a general-purpose technology.
    • They predict a future with a divide between applications using powerful customized chips (fast lane) and those relying on general-purpose chips (slow lane).
  • Accessibility and Democratization of AI:
    • While specialized chips represent a shift from general-purpose computing, they have made technologies like AI more accessible and affordable.
    • Companies like NVIDIA have facilitated the widespread adoption of AI through their specialized chips.

The Rise of Big Tech in Chip Design:

  • Amazon and Google’s Entry:
    • Motivated by enhancing their cloud computing infrastructure, both companies have ventured into chip design.
    • Google’s TPU chips, optimized for AI, are available to the public through their cloud platform.
  • A Bifurcated Computing Landscape?:
    • Some view this trend as a division between a slow lane and a fast lane in computing power.
    • However, access to the “fast lane” through specialized chips or AI-optimized clouds is increasingly available.

New Frontiers in Chip Integration and System Design:

  • Heterogeneous Integration:
    • Modern devices often incorporate multiple types of chips, with some focused on general operations and others optimized for specific tasks (e.g., camera processors).
    • Advanced packaging technologies allow for efficient chip interconnection, enabling flexible configurations and adaptability to evolving needs.
  • Holistic System Design:
    • Chip manufacturers are placing greater emphasis on considering the broader systems in which their chips will function, optimizing for performance and efficiency within those specific environments.

The Enduring Importance of Computing Power and the Quest for Progress:

  • Redefining Moore’s Law:
    • The crucial question is not whether the original definition of Moore’s Law (exponential transistor growth on a chip) is reaching its limit.
    • The focus should be on whether we are approaching the peak of computing power that can be cost-effectively achieved on a chip.
  • Continued Investment and Optimism:
    • The significant investments and tireless efforts of countless engineers reflect a strong belief in the potential for further advancements in computing power.

From Humble Beginnings to a Transformative Force

  • 1958: The year that witnessed the convergence of key figures like Morris Chang, Pat Haggerty, Weldon Word, Jay Lathrop, and Jack Kilby at Texas Instruments also saw a pivotal electronics conference in Washington, D.C.
  • Unsung Heroes: Chang, Gordon Moore, and Bob Noyce, future titans of technology, were unknown to the world as they enjoyed a night of camaraderie amidst the snowdrifts.
  • A Lasting Legacy: The chips they invented and the industry they built have profoundly shaped our world, providing the unseen circuitry that powers our lives and will continue to mold our future.
About Me:
  • I’m Christian Mills, a deep learning consultant specializing in computer vision and practical AI implementations.
  • I help clients leverage cutting-edge AI technologies to solve real-world problems.
  • Learn more about me or reach out via email at [email protected] to discuss your project.