Multi-Disciplinary Reading – Book Reviews

Chip War: The Fight for the World’s Most Critical Technology (2022)
Book by Chris Miller

I read this well-researched book on the chip industry over the holidays. This book contends that semiconductors will determine the shape of international politics, the structure of the world economy, and the balance of military power.

I have organized my notes around some themes I observed; the book, however, is organized differently. A highly recommended read if you want to understand the chip industry.

(1) Journey of a chip

• A typical chip is designed with blueprints from the Japanese-owned, UK-based company called Arm, by a team of engineers in California and Israel, using design software from the United States.

• When a design is complete, it’s sent to a facility in Taiwan, which buys ultra-pure silicon wafers and specialized gases from Japan. The design is carved into silicon using some of the world’s most precise machinery, which can etch, deposit, and measure layers of materials a few atoms thick. These tools are produced primarily by five companies, one Dutch, one Japanese, and three Californian, without which advanced chips are basically impossible to make.

• Then the chip is packaged and tested, often in Southeast Asia, before being sent to China for assembly into a phone or computer.

(2) Type of chips

• It is common to split the semiconductor industry into three categories. “Logic” refers to the processors that run smartphones, computers, and servers. “Memory” refers to DRAM, which provides the short-term memory computers need to operate, and flash, also called NAND, which remembers data over time. The third category of chips is more diffuse, including analog chips like sensors that convert visual or audio signals into digital data, radio frequency chips that communicate with cell phone networks, and semiconductors that manage how devices use electricity.

• This third category has not been primarily dependent on Moore’s Law to drive performance improvements. Clever design matters more than shrinking transistors. Today around three-quarters of this category of chips are produced on processors at or larger than 180 nanometers.

• Today, the biggest analog chipmakers are American, European, or Japanese. Most of their production occurs in these three regions, too, with only a sliver offshored to Taiwan and South Korea.

• The largest analog chipmaker today is Texas Instruments, which failed to establish an Intel-style monopoly in the PC, data center, or smartphone ecosystems but remains a medium-sized, highly profitable chipmaker with a vast catalog of analog chips and sensors.

• There are many other U.S.-based analog chipmakers now, like Onsemi, Skyworks, and Analog Devices, alongside comparable companies in Europe and Japan.

• The memory market, by contrast, has been dominated by a relentless push toward offshoring production to a handful of facilities, mostly in East Asia. Rather than a diffuse set of suppliers centered in advanced economies, the two main types of memory chip—DRAM and NAND—are produced by only a couple of firms.

• For DRAM memory chips an advanced fab can cost $20 billion. There used to be dozens of DRAM producers, but today there are only three major producers: Idaho’s Micron and with Korea’s Samsung and SK Hynix.

• The market for NAND, the other main type of memory chip, is also Asia-centric. Samsung, the biggest player, supplies 35 percent of the market, with the rest produced by Korea’s Hynix, Japan’s Kioxia, and two American firms—Micron and Western Digital. The Korean firms produce chips almost exclusively in Korea or China, but only a portion of Micron and Western Digital’s NAND production is in the U.S., with other production in Singapore and Japan.

• The big shift in recent years is the collapse in the share of logic chips produced in the United States.

(3) Chip architectures

• Arm: is a UK company that licenses to chip designers use of an instruction set architecture—a set of basic rules governing how a given chip operates. The Arm architecture is dominant in mobile devices and is slowly winning market share in PCs and data centers.

• x86:an instruction set architecture that is dominant in PCs and data centers. Intel and AMD are the two main firms producing such chips.

• Only three companies have the necessary intellectual property to produce x86 chips: America’s Intel and AMD as well as a small Taiwanese company called Via.

• RISC-V: an open-source architecture growing in popularity because it is free to use, unlike Arm and x86. The development of RISC-V was partially funded by the U.S. government but now is popular in China because it is not subject to U.S. export controls.

(4) Photolithography and EUV

• Photolithography (a.k.a. printing with light) is the process of shining light or ultraviolet light through patterned masks: the light then interacts with photoresist chemicals to carve patterns on silicon wafers.

• This process helps produce transistors much smaller than had previously been possible

• Photolithography made it possible to imagine mass-producing tiny transistors.

• The Dutch company ASML (offshoot of Phillips with Intel as major investor) builds 100 percent of the world’s extreme ultraviolet (EUV) lithography machines, without which cutting-edge chips are simply impossible to make.

• Producing advanced semiconductors has relied on some of the most complex machinery ever made. ASML’s EUV lithography tool is the most expensive mass-produced machine tool in history, so complex it’s impossible to use without extensive training from ASML personnel, who remain on-site for the tool’s entire life span. EUV machines cost over $100 million each.

• EUV technology took almost 30 years and billions of dollars to come to fruition. Some of the best precision engineering companies worked with ASML to make it happen (Cymer for lithographic light sources; Trumpf for carbon-dioxide based lasers)

• ASML will have introduced a new generation tool, called high-aperture EUV, which is scheduled to be ready in the mid-2020s

(5) A highly concentrated industry

• Nearly every chip in the world uses software from at least one of three U.S.-based companies, Cadence, Synopsys, and Mentor. These control around three-quarters of the market. It is impossible to design a chip without using at least one of these firms’ software.

• Excluding the chips Intel builds in-house, all the most advanced logic chips are fabricated by just two companies, Samsung and TSMC, both located in countries that rely on the U.S. military for their security.

• Making advanced processors requires EUV lithography machines produced by just one company, the Netherlands’ ASML.

• Applied Materials remains the world’s largest semiconductor toolmaking company, building equipment like the machines that deposit thin films of chemicals on top of silicon wafers as they were processed.

• Lam Research has world-beating expertise in etching circuits into silicon wafers.

• KLA, based in Silicon Valley, has the world’s best tools for finding nanometer-sized errors on wafers and lithography masks.

(6) Rise and fall of Japan’s semiconductor story

• Japanese society was structurally geared to produce massive savings, because its postwar baby boom and rapid shift to one-child households created a glut of middle-aged families focused on saving for retirement. Japan’s skimpy social safety net provided a further incentive for saving.

• Tight restrictions on stock markets and other investments left people with little choice but to stuff savings in bank accounts. As a result, banks were flush with deposits, extending loans at low rates because they had so much cash on hand.

• Japanese firms got access to far cheaper capital. Chipmakers like Hitachi, Mitsubishi, Toshiba, Fujitsu had close links to banks that provided large, long-term loans. Even when Japanese companies were unprofitable, their banks kept them afloat by extending credit.

• Because of this, five years after the 64K DRAM chip was introduced, Intel—the company that had pioneered DRAM chips a decade earlier—was left with only 1.7 percent of the global DRAM market.

• The biggest error that Japan’s chip firms made was to miss the rise of PCs. None of the Japanese chip giants could replicate Intel’s pivot to microprocessors or its mastery of the PC ecosystem.

• 1998, South Korean firms overtook Japan as the world’s largest producers of DRAM. Japan’s market share fell from 90 percent in the late 1980s to 20 percent by 1998.

(7) Intel

• Due to competition from Japan Intel was about to go out of business. Andy Grove (then CEO) made a pivot to producing microprocessors for PCs and Intel became a turn-around story.

• By the mid-2000s, just as cloud computing was emerging, Intel had won a near monopoly over data center chips, competing only with AMD. Today, nearly every major data center uses x86 chips from either Intel or AMD. The cloud can’t function without their processors.

• Shortly after the deal to put Intel’s chips in Mac computers, Steve Jobs came back to Otellini (Intel’s CEO) with a new pitch. Would Intel build a chip for Apple’s newest product, a computerized phone? Intel turned down the iPhone contract (phew!)

• The early iPhone processors were produced by Samsung, which had followed TSMC into the foundry business. Otellini’s prediction that the iPhone would be a niche product proved horribly wrong. By the time he realized his mistake, it was too late.

(8) Taiwan and TSMC

• Taiwan produces 11 percent of the world’s memory chips. More important, it fabricates 37 percent of the world’s logic chips.

• Taiwan had deliberately inserted itself into semiconductor supply chains since the 1960s, as a strategy to provide jobs, acquire advanced technology, and to strengthen its security relationship with the United States. In the 1990s, Taiwan’s importance began to grow, driven by the spectacular rise of the Taiwan Semiconductor Manufacturing Company (TSMC)

• The Taiwanese government provided 48 percent of the startup capital for TSMC. Philips, the Dutch company provided another 27.5 percent. Rest came from wealthy Taiwanese businessmen.

• A crucial ingredient in TSMC’s early success was deep ties with the U.S. chip industry. Most of its customers were U.S. chip designers, and many top employees had worked in Silicon Valley.

• The founding of TSMC gave all chip designers a reliable partner. TSMC promised never to design chips, only to build them.

• TSMC’s foundry model gave birth to dozens of new “authors”—fabless chip design firms—that transformed the tech sector by putting computing power in all sorts of devices. However, the democratization of authorship coincided with a monopolization of the digital printing press. The economics of chip manufacturing required relentless consolidation.

• The geography of chip fabrication shifted drastically over the 1990s and 2000s. U.S. fabs made 37 percent of the world’s chips in 1990, but this number fell to 19 percent by 2000 and 13 percent by 2010.

• (2015 stat.) Measured by thousands of wafers per month, the industry standard, TSMC had a capacity of 1.8 million while Samsung had 2.5 million. GlobalFoundries had only 700,000. (Top 3 fabs in the world.)

(9) Rise of fabless chip firms – Nvidia, Qualcomm etc.

• Since the late 1980s, there’s been explosive growth in the number of fabless chip firms, which design semiconductors in-house but outsource their manufacturing, commonly relying on TSMC for this service.

• Today Nvidia’s chips, largely manufactured by TSMC, are found in most advanced data centers.

• Nvidia designs chips called graphics processor units (GPUs) capable of handling 3D graphics. Making realistic graphics requires use of programs called shaders, which tell all the pixels in an image how they should be portrayed in a given shade of light.

• Nvidia’s GPUs can render images quickly because, unlike Intel’s microprocessors or other general-purpose CPUs, they’re structured to conduct lots of simple calculations—like shading pixels—simultaneously

• Nvidia’s GPUs are processor of choice for advanced machine learning/AI applications

• Cloud computing companies like Amazon Web Services and Google are designing their own chips that will be fabricated by companies like TSMC (which are pure foundry plays)

• Tesla is also a leading chip designer. The company hired star semiconductor designers like Jim Keller to build a chip specialized for its automated driving needs.

• For each generation of cell phone technology after 2G, Qualcomm contributed key ideas about how to transmit more data via the radio spectrum and sold specialized chips with the computing power capable of deciphering this cacophony of signals. The company’s patents are so fundamental it’s impossible to make a cell phone without them.

• Qualcomm has made hundreds of billions of dollars selling chips and licensing intellectual property. But it hasn’t fabricated any chips: they’re all designed in-house but fabricated by companies like Samsung or TSMC.

• Designing chips as complex as the processors that run smartphones is expensive, which is why most low- and midrange smartphone companies buy off-the-shelf chips from companies like Qualcomm. However, Apple has invested heavily in R&D and chip design facilities in Bavaria and Israel as well as Silicon Valley, where engineers design its newest chips. Now Apple not only designs the main processors for most of its devices but also ancillary chips that run accessories like AirPods.

• Today, no company besides TSMC has the skill or the production capacity to build the chips Apple needs.

(10) Playbook for setting up semiconductor business

• When Japan, Taiwan, and South Korea wanted to break into the complex and high-value portions of the chip industry, they poured capital into their semiconductor companies, organizing government investment but also pressing private banks to lend.

• Second, they tried to lure home their scientists and engineers who’d been trained at U.S. universities and worked in Silicon Valley.

• Third, they forged partnerships with foreign firms but required them to transfer technology or train local workers.

• Fourth, they played foreigners off each other, taking advantage of competition between Silicon Valley firms to get the best deal for themselves.

(11) China’s challenge

• During most years of the 2000s and 2010s, China spent more money importing semiconductors than oil. High-powered chips were as important as hydrocarbons in fueling China’s economic growth. Unlike oil, though, the supply of chips is monopolized by China’s geopolitical rivals.

• China is devoting its best minds and billions of dollars to developing its own semiconductor technology in a bid to free itself from America’s chip choke.

• World War II was decided by steel and aluminum, and followed shortly thereafter by the Cold War, which was defined by atomic weapons. The rivalry between the United States and China may well be determined by computing power.

• Across the entire semiconductor supply chain Chinese firms have a 6 percent market share, compared to America’s 39 percent, South Korea’s 16 percent, or Taiwan’s 12 percent

• For advanced logic, memory, and analog chips, however, China is crucially dependent on American software and designs; American, Dutch, and Japanese machinery; and South Korean and Taiwanese manufacturing.

• China’s government set out a plan called Made in China 2025, which envisioned reducing China’s imported share of its chip production from 85 percent in 2015 to 30 percent by 2025.

• China’s import of chips—$260 billion in 2017—was far larger than Saudi Arabia’s export of oil or Germany’s export of cars. China spends more money buying chips each year than the entire global trade in aircraft.

• One of China’s core challenges today is that many chips use either the x86 architecture (for PCs and servers) or the Arm architecture (for mobile devices); x86 is dominated by two U.S. firms, Intel and AMD, while Arm, which licenses other companies to use its architecture, is based in the UK. However, there’s now a new instruction set architecture called RISC-V that is open-sourced. The idea of an open-source architecture appeals to many parts of the chip industry.

• China’s also investing heavily in emerging semiconductor materials like silicon carbide and gallium nitride, which are unlikely to displace pure silicon.

(12) Smartphone vs PC supply chain

• The smartphone supply chain looks very different from the one associated with PCs. Smartphones and PCs are both assembled largely in China with high-value components mostly designed in the U.S., Europe, Japan, or Korea.

• For PCs, most processors come from Intel and are produced at one of the company’s fabs in the U.S., Ireland, or Israel.

• Smartphones are different. They’re stuffed full of chips, not only the main processor (which Apple designs itself), but modem and radio frequency chips for connecting with cellular networks, chips for WiFi and Bluetooth connections, an image sensor for the camera, at least two memory chips, chips that sense motion, as well as semiconductors that manage the battery, the audio, and wireless charging.

• Around a quarter of the chip industry’s revenue comes from phones.

(13) Chip choke of 2021

• Carmakers spent much of 2021 struggling and often failing to acquire semiconductors. These firms are estimated to have produced 7.7 million fewer cars in 2021 than would have been possible had they not faced chip shortages, which implies a $210 billion collective revenue loss, according to industry estimates.

• The world produced more chips in 2021 than ever before (a 13 percent increase over 2020). The semiconductor shortage is mostly a story of demand growth rather than supply issues. It’s driven by new PCs, 5G phones, AI-enabled data centers.

(14) Misc.:

• Integrated circuits made up 15 percent of South Korea’s exports in 2017; 17 percent of Singapore’s; 19 percent of Malaysia’s; 21 percent of the Philippines’; and 36 percent of Taiwan’s.

• 5G networks will send more data by using a new, empty radio frequency spectrum that was previously considered impractical to fill (all thanks to advanced semiconductors)

• The global chip industry spends over $100 billion annually on capex.

• Javelin anti-tank missiles (used in Ukraine war) rely on over 200 semiconductors each as they home in on enemy tanks.

• Kilby called his invention an “integrated circuit,” but it became known colloquially as a “chip,” because each integrated circuit was made from a piece of silicon “chipped” off a circular silicon wafer.