Why is it so difficult to manufacture a tiny chip?

In the digital age, all of us cannot live without chips. Our computers, mobile phones, and even cars are equipped with a large number of chips. As long as one chip fails to work properly, it will affect our lives, ranging from mobile phone failure to car out of control...

While enjoying the convenience of chips, have we ever thought about why chips are so important to the digital age? Why are their development and manufacturing so difficult? This has to start with the history of chips.

From vacuum tubes to transistors

"In ancient times, people governed by knotting ropes." Since the birth of human civilization, calculation has become an inseparable part of our lives. From the balance of payments of a family to the economic direction of a country, these numbers that determine the fate of a family or a country all require calculation. For this purpose, people have developed many calculation tools, such as abacus that moves beads up and down, or calculators that can press buttons to get the desired results.

Copyright images in the gallery. Reprinting and using them may lead to copyright disputes.

As our demand for computing continued to increase, human-based computing methods soon encountered bottlenecks. The war gave birth to early computers: Turing's computer, which relied on the principles of electromechanics, cracked the German Enigma code; and in order to crack the German Lorenz code, the United Kingdom developed the "Colossus computer", which is also considered the world's first programmable digital computer. These machines can easily complete calculations that are difficult or even impossible for humans to do alone.

The core of the giant computer is the "vacuum tube", which looks like a huge light bulb with some metal wires inside. After the power is turned on, these metal wires have only two fates: power or no power, which corresponds to 1 and 0 in binary. Using these two numbers, any calculation can be performed in theory. Our current virtual world can also be roughly understood as being born from countless 1s and 0s.

Although computers based on vacuum tubes are powerful, they also have their own limitations. On the one hand, vacuum tubes are too large. The ENIAC machine manufactured by the University of Pennsylvania has more than 17,000 vacuum tubes, which takes up a huge area and consumes a lot of power. On the other hand, these massive vacuum tubes also bring various hidden dangers. According to statistics, on average, this machine will have a vacuum tube failure every 2 days, and each troubleshooting takes at least 15 minutes. In order to stably produce various 1s and 0s, people began to look for alternatives to vacuum tubes.

The famous Bell Labs made a breakthrough, and their choice was semiconductors - the conductivity of this material is based on the difference between conductors (allowing current to flow freely, such as copper wires) and insulators (completely non-conductive, such as glass). Under certain conditions, its conductive properties can change. For example, "silicon" (Si), which we have all heard of, does not conduct electricity itself, but as long as certain other materials are added, it can become conductive. The name "semiconductor" comes from this.

William Shockley of Bell Labs first proposed a theory that applying an electric field near semiconductor materials could change their conductivity , but he was unable to confirm his theory experimentally.

Inspired by this theory, his two colleagues John Bardeen and Walter Brattain created a semiconductor device called a "transistor" two years later. Shockley, unwilling to be surpassed, developed a newer transistor a year later. Ten years later, the three of them won the Nobel Prize in Physics for their contributions to the field of transistors. As the field of transistors continues to expand and more new members are welcomed, they have also become the cornerstone of the digital age.

The Birth of Chips and Silicon Valley

As transistors gradually replaced vacuum tubes, their limitations became apparent in practical applications, chief among them being how to wire tens of thousands of transistors into usable circuits.

In order for transistors to achieve complex functions, in addition to transistors, the circuit also requires components such as resistors, capacitors, and inductors, which are then welded and connected. There is no standard for the size of these components, so the workload of making circuits is huge and it is easy to make mistakes. One solution at the time was to specify the size and shape of each electronic component and redefine the design of the circuit using modular means.

Jack Kilby of Texas Instruments was not impressed by this plan, believing that it would not solve the fundamental problem - no matter how much regulation was made, the size could not be reduced. The resulting modular circuits were still large and could not be applied to smaller devices. His solution integrated everything, putting all transistors, resistors, and capacitors on a piece of semiconductor material, saving a lot of subsequent manufacturing time and reducing the possibility of errors.

In 1958, he made a prototype using germanium (Ge), which contained a transistor, three resistors, and a capacitor. When connected with wires, it could generate a sine wave. This new circuit was called an "integrated circuit," and later it was more commonly known as a chip. Kilby himself won the Nobel Prize in Physics in 2000 for his invention.

Copyright images in the gallery. Reprinting and using them may lead to copyright disputes.

Around the same time, eight engineers simultaneously resigned from Shockley and started a business together, establishing Fairchild Semiconductor. These eight resigned were the famous "Eight Rebels" in the history of semiconductors. Robert Noyce was the leader of these eight rebels, and he also thought of producing multiple components on a piece of semiconductor material to manufacture integrated circuits. Unlike Kilby's method, his design integrated the wires and various components into one. This integrated design has greater advantages in production and manufacturing. The only problem is cost - although Noyce's integrated circuit has obvious advantages, the cost is 50 times the original.

Just as the war decades ago gave birth to the prototype of computers, the Cold War also brought unexpected business opportunities to Noyce's chips. As the former Soviet Union launched the first artificial satellite and sent humans into space for the first time, the United States, sensing the crisis, launched a comprehensive catch-up plan. They decided to send people to the moon as a final counterattack, but this task required a huge amount of computing (controlling rockets, manipulating landing capsules, calculating the best time window, etc.), and the National Aeronautics and Space Administration (NASA) bet its fate on Noyce's chip: this integrated circuit is smaller in size and consumes less power. In order to send people to the moon, every gram of weight and every watt of energy must be calculated. For such an extreme project, it is undoubtedly a better choice.

In the human moon landing project, the chip demonstrated its potential to the world - Noyce said that in the Apollo project computer, its chip ran for 19 million hours and only had two failures, one of which was caused by external factors.

In addition, the moon landing also proved that chips can operate normally in the extremely harsh environment of outer space. After the rise of Fairchild, employees from this company also branched out in the local area and established companies such as Intel and AMD. This area with a large number of semiconductor companies later had a more famous name - Silicon Valley.

Photolithography

Integrated circuits are much smaller than circuits made up of scattered transistor components, and a microscope is often needed to see the structure inside and check the quality. Jay Lathrop of Texas Instruments came up with an idea during an observation: if you look down through a microscope, you can magnify things, so if you look up from the bottom, can you make things smaller?

This was not for fun. At that time, the size of integrated circuits was close to the limit of manual manufacturing, and it was difficult to make new breakthroughs. If the designed circuit diagram could be "shrink-printed" onto semiconductor materials, it would be possible to manufacture through automated technology and achieve mass production.

Lathrop quickly tested his idea. First he bought a chemical called photoresist from Kodak and coated it on the semiconductor material. Then he turned the microscope upside down and covered the lens with a plate, leaving only a small pattern.

Finally, he let the light pass through the lens and shine it on the photoresist at the other end of the microscope. Under the action of light, the photoresist undergoes a chemical reaction and slowly dissolves and disappears, revealing the silicon material underneath. The shape of the exposed material is exactly the same as the pattern he originally designed, but it is hundreds or thousands of times smaller. On the exposed grooves, manufacturers can add new materials, connect circuits, and then wash away the excess photoresist. This set of processes is the photolithography technology for manufacturing chips.

Texas Instruments subsequently further improved this process so that each link could have a standard reference, which also ushered in the era of standardized mass production of integrated circuits. As chips become more and more complex, the production of an integrated circuit requires at least dozens of times.

Fairchild followed suit and developed its own photolithography production technology. In addition to Noyce, the other seven founders of this company were also extraordinary people. Among them, Gordon Moore was the best.

In 1965, he predicted the future of integrated circuits, believing that with the continuous update of production technologies such as lithography, the number of components in chips will double every year. In the long run, the computing power of chips will increase exponentially and the cost will drop significantly. An obvious consequence of this is that chips will enter the homes of ordinary people in large numbers and completely change the world. Moore's prediction was later called "Moore's Law" and became known to the world.

The premise for the establishment of Moore's Law is the continuous development and innovation of manufacturing processes. The photolithography technology developed by some early companies is almost perfect, just like drawing light stroke by stroke on the photoresist to carve out a line with a width of only one micron. Moreover, this technology can carve out multiple chips at a time, greatly improving the production capacity of chips. However, with the ever-increasing demand for chip manufacturing precision, micron-level photolithography machines can no longer meet the needs of the industry, and nano-level photolithography machines have become the new darling.

However, it is not easy to develop such a lithography machine - how to perform lithography in increasingly smaller mini spaces has become a bottleneck hindering the development of lithography technology.

Extreme Ultraviolet Lithography

In 1992, Moore's Law was about to fail. If we wanted to maintain this law, chip circuits would need to be made smaller. Both the light source used and the lens through which the light shines had new requirements.

When Lathrop first developed photolithography, he used the simplest visible light. The wavelength of this light is about a few hundred nanometers, and the ultimate limit of the size printed on the chip is also a few hundred nanometers. If you need to print smaller components on the chip (for example, only tens of nanometers), the light source required must also exceed the limit of visible light and enter the field of ultraviolet light.

Some companies have developed manufacturing equipment using deep ultraviolet (DUV), which uses a wavelength of less than 200 nanometers. But in the long run, extreme ultraviolet (EUV) is the area people want to reach - the shorter the wavelength, the more details can be engraved on the chip . In the end, people set their sights on extreme ultraviolet light with a wavelength of 13.5 nanometers, and ASML in the Netherlands became the world's only EUV machine manufacturer.

EUV technology has been developed for nearly 20 years. In order to manufacture a working EUV machine, ASML needs to look for the most advanced parts in the world to meet its needs. As a lithography machine, the first thing you need is a light source: in order to produce EUV, people need to launch a tin droplet with a diameter of only tens of microns, let it travel through a vacuum at a speed of more than 300 kilometers per hour, and hit it accurately with a laser at the same time - not once, but twice.

The first step is to heat it, and the second step is to blast it into plasma at 500,000 degrees, which is several times the temperature of the sun's surface. This process must be repeated 50,000 times per second to produce enough EUV. It is conceivable how many advanced components are needed for such high-precision technology.

Copyright images in the gallery. Reprinting and using them may lead to copyright disputes.

The actual operation is more complicated than the above description. For example, in order to eliminate the large amount of heat generated during laser irradiation, a fan is needed for ventilation, and the rotation speed needs to reach 1,000 times per second. This speed exceeds the limit of physical bearings, so magnets are needed to suspend the fan in the air for rotation.

In addition, the laser emitter has strict requirements on the gas density, and it is also necessary to avoid reflections when the laser shines on the tin droplets, which would affect the instrument. It took more than 10 years of research and development just to develop the machine that emits the laser, and each emitter requires more than 450,000 components.

The EUV produced by bombarding tin droplets is hard to come by, and researchers also need to learn how to collect these rays and guide them to the chip. The wavelength of EUV is so short that it is easily absorbed by the surrounding materials instead of being reflected. Finally, Carl Zeiss developed an extremely smooth mirror that can reflect EUV.

The mirror is so smooth that it is beyond imagination. Officially, if the mirror is enlarged to the size of Germany, the largest irregularity of the mirror is only 0.1 mm. The company is also confident that their mirror can guide the laser to accurately hit the golf ball on the moon.

Such a complex equipment requires not only scientific technology, but also complete supply chain management. ASML itself only produces 15% of the components of its EUV machine, and the rest comes from partners around the world. Of course, they will also carefully monitor these purchased products, and if necessary, they will even buy these companies and manage them themselves. Such a machine is the crystallization of technology from different countries.

The prototype of the first EUV machine was born in 2006. The first commercial EUV machine was shipped in 2010. In the next few years, ASML expects to launch a new generation of EUV machines, each costing $300 million.

Chip Application

With advanced manufacturing technology, a variety of chips have been born. Some people have concluded that in the 21st century, chips can be divided into three categories.

The first type is logic chips, which are used as processors in our computers, mobile phones, or network servers;

The second category is memory chips. Classic examples include the DRAM chip developed by Intel. Before the launch of this product, data storage relied on magnetic cores: magnetized components represent 1, and unmagnetized components represent 0. Intel's approach is to combine transistors and capacitors, charging represents 1, and uncharged represents 0. Compared with magnetic cores, the new storage tool is similar in principle, but everything is integrated into the chip, so it is smaller and has a lower error rate. This type of chip can provide computers with short-term and long-term memory during operation;

The third type of chip is called an "analog chip" that processes analog signals.

Among these chips, logic chips are probably more familiar. Although Intel developed the earliest DRAM memory chips, it was losing ground to Japanese companies. In 1980, Intel and IBM reached a cooperation to manufacture central processing units, or CPUs, for personal computers.

With the advent of IBM's first personal computer, the Intel processor built into this computer became the industry's "standard configuration," just as Microsoft's Windows system became a more familiar operating system to the public. This gamble also allowed Intel to completely withdraw from the DRAM field and re-emerge.

The development of the CPU was not achieved overnight. In fact, as early as 1971, Intel created the first microprocessor (compared with the CPU, it can only handle a single specific task), and the development of the entire design process took a full six months. At that time, this microprocessor had only thousands of components, and the design tools used were only colored pencils and rulers, which was as backward as a medieval craftsman. Lynn Conway developed a program that solved the problem of automated chip design. Using this program, students who have never designed a chip can learn how to design a functional chip in a short time.

In the late 1980s, Intel developed the 486 processor, which could place 1.2 million micro-components on a tiny silicon chip to generate all kinds of 0s and 1s. By 2010, the most advanced microprocessor chips could carry 1 billion transistors. The development of such chips is inseparable from the design software developed by a few oligopolistic companies.

Another type of logic chip, the graphics processing unit (GPU, commonly known as graphics card), has also attracted more and more attention in recent years. In this field, Nvidia is an important player. In its early days, the company believed that 3D graphics were the future development direction, so it designed a GPU that could process 3D graphics and developed a set of corresponding software to tell the chip how to work. Unlike Intel's central processing unit's "sequential calculation" mode, the advantage of GPU is that it can perform a large number of simple calculations at the same time.

No one expected that in the era of artificial intelligence, GPUs would have a new mission. In order to train artificial intelligence models, scientists need to use data to continuously optimize algorithms so that the models can be trained to complete tasks assigned by humans, such as identifying cats and dogs, playing Go, or talking to humans. At this time, GPUs, which were developed to perform multiple operations at the same time and "process data in parallel", have a unique advantage and have taken on a new life in the era of artificial intelligence.

Another important application of chips is communication. Irwin Jacobs saw that chips could process some complex algorithms to encode massive amounts of information, so he and his friends founded Qualcomm and entered the communications field. We know that the earliest mobile phone was also called a big brother, which looked like a black brick.

Subsequently, communication technology has developed rapidly - 2G technology can transmit pictures and texts, 3G technology can open websites, 4G is enough to watch videos smoothly, and 5G can provide a greater leap. Each G here represents a "generation". It can be seen that each generation of wireless technology has allowed us to transmit information through radio waves exponentially. Nowadays, when we watch videos on our mobile phones, we feel impatient if there is a slight lag. Little did we know that more than 10 years ago, we could only send text messages.

Copyright images in the gallery. Reprinting and using them may lead to copyright disputes.

Qualcomm participated in the development of 2G and other mobile phone technologies. Using chips that continue to evolve according to Moore's Law, Qualcomm can put more mobile phone calls in the vast space through unlimited spectrum. In order to upgrade the 5G network, not only new chips need to be put into mobile phones, but also new hardware needs to be installed in base stations. These hardware and chips can transmit data faster by wireless methods with more powerful computing power.

Manufacturing and supply chain

In 1976, almost every company that designed chips had its own manufacturing base. However, if the chip design and chip manufacturing work were separated and the chip manufacturing work was handed over to a specialized foundry, the cost of chip design companies could be greatly reduced.

TSMC was born, promising to only manufacture chips, not design them. In this way, companies that design chips don't have to worry about confidential information being leaked. And TSMC doesn't rely on selling more chips - as long as its customers are successful, his company is successful.

Before TSMC, some American chip companies had set their sights on the other side of the vast Pacific Ocean: In the 1960s, Fairchild established a center in Hong Kong to assemble various chips shipped from California. In the first year of production, the Hong Kong factory assembled 120 million devices, with extremely low labor costs but excellent quality. Within ten years, almost all chip companies in the United States had established assembly plants in Asia. This also laid the foundation for the current chip supply chain structure centered on East Asia and Southeast Asia.

Asia's high efficiency and obsession with quality soon impacted the United States' position in the chip industry. In the 1980s, executives of companies responsible for testing chip quality unexpectedly discovered that the quality of chips produced in Japan had surpassed that of the United States - the failure rate of ordinary American chips was 4.5 times that of Japanese chips, and the failure rate of the worst American chips was 10 times that of Japanese chips! "Made in Japan" is no longer synonymous with cheap but poor quality. What's more terrifying is that even American production lines that have been squeezed to the limit are far less efficient than Japan. "Japan's capital cost is only 6% to 7%, and when I'm at my best, the cost is 18%." AMD CEO Jerry Sanders once said.

The financial environment also played a role in fueling the trend: in order to curb inflation, the US interest rate once reached 21.5%; while Japanese chip companies were supported by financial groups, and the public was used to saving, which enabled banks to provide chip companies with large, low-interest loans for a long time. With the help of capital, Japanese companies could aggressively seize the market.

As a result, companies that are capable of producing advanced logic chips are concentrated in East Asia, and the manufactured chips are then sent to surrounding areas for assembly. For example, Apple's chips are mainly produced in South Korea and Taiwan, and then sent to Foxconn for assembly. These chips include not only the main processor, but also wireless network and Bluetooth chips, chips for taking pictures, chips for sensing motion, etc.

As the ability to produce and manufacture chips is gradually concentrated in a few companies, these original foundry companies have also gained greater power, such as coordinating the needs of different companies and even making rules. Since the companies currently responsible for designing chips do not have the ability to manufacture chips, they can only follow suggestions. These growing powers are also one of the topics of the current geopolitical struggle.

Conclusion

From machines that decrypted World War II codes to spacecraft that sent people to the moon. From portable music players to airplanes and cars for daily travel, to the phones and computers we use to read this article, these devices are inseparable from chips.

Every day, every ordinary person uses at least dozens or hundreds of chips in their lives. All of this is inseparable from the development of chip technology and the production and manufacturing of chips. Chips are one of the most important inventions of this era. To develop new chips, not only scientific and technological support is needed, but also advanced manufacturing and production capabilities, as well as the civilian market for these chips.

The layout of chip design and manufacturing capabilities has undergone decades of changes and has formed the current pattern, which has also produced a different meaning in this era. This article hopes to review some important industrial nodes related to chips in the past few decades for reference by interested readers.

Planning and production

Author: Ye Shi Popular Science Creator

Review丨Huang Yongguang Associate Researcher of Optoelectronic Chips, Institute of Semiconductors, Chinese Academy of Sciences

Planning丨Xu Lai

Editor: Yinuo

The cover image and the images in this article are from the copyright library

Reprinting may lead to copyright disputes