In this blog post, we’ll explore everything from the basic principles of semiconductor devices to how transistors work and their evolution, as well as the limits of miniaturization technology and the future direction of semiconductor technology.
The majors in the Department of Electrical Engineering can be broadly categorized into communications, semiconductors, control systems, computer design, electronic physics, and electrical energy conversion. While there are differences in the specifics, the field that interests me the most is semiconductor devices. Semiconductor devices have become an indispensable and crucial part of our daily lives today. Semiconductor devices are found in nearly all the electronic products we use, such as cell phones, computers, televisions, and radios. Furthermore, sensors and computers made from semiconductor devices play crucial roles even in automobiles and apartments. As such, semiconductor technology plays a vital role in our lives in the 21st century. The leading nations in the semiconductor industry include South Korea, the United States, Japan, and Taiwan. South Korean companies possess global competitiveness in the fields of memory semiconductors and semiconductor manufacturing, while the United States and Taiwan also hold significant positions in advanced semiconductor design and manufacturing. Whereas in the past, advancements in semiconductor manufacturing technology focused on making devices smaller using planar structures, today, three-dimensional integration technology—which improves transistor structures in a three-dimensional manner and vertically connects multiple chips—is advancing alongside advanced packaging technology. Furthermore, the 60 nm figure mentioned in a previous article does not represent the current state-of-the-art manufacturing process. Today, 2 nm-class process technology has entered the actual production phase, and transistor structures are evolving from the traditional planar structure through FinFET to the Gate-All-Around (GAA) structure, in which the gate surrounds the channel. Intel is applying the GAA-based RibbonFET and PowerVia—a back-side power supply technology—in its 18A process, which entered mass production in 2025. To easily grasp just how small these semiconductors are, one can compare them to the thickness of a human hair. Since a human hair is generally about tens of micrometers thick, one can imagine just how small semiconductor structures fabricated on the nanometer scale are. Furthermore, since the wavelength of visible light—which the human eye can perceive—is roughly in the range of several hundred nanometers, semiconductor structures on the nanometer scale correspond to an extremely small scale that cannot be seen directly with the naked eye.
So, what exactly is the operating principle of these semiconductor devices, and how are they manufactured? First, it is necessary to accurately understand the term “semiconductor device.” A semiconductor device refers to an electronic device made using a semiconductor material. Materials can be broadly classified into conductors, semiconductors, and insulators based on their electrical properties. As one might intuitively understand, conductors allow electricity to flow well, insulators allow almost no electricity to flow, and semiconductors are materials with properties that fall somewhere in between. However, semiconductors should not be understood merely as lying somewhere between conductors and insulators. Semiconductors are significant because their electrical properties can be controlled depending on conditions such as the addition of impurities, electric fields, and temperature. Representative semiconductor materials include silicon and germanium, while compound semiconductors include gallium arsenide (GaAs), among others. So, to what extent can a semiconductor material conduct electric current? Taking silicon—a representative semiconductor material—as an example, pure silicon has a very low concentration of free electrons at room temperature, so it does not conduct electricity as well as metals. The key to conducting electricity lies in charge carriers that can move freely; in silicon, these are electrons and holes. However, since it is difficult to control the current to the desired level using pure silicon alone, doping is performed to regulate the current and electrical properties. The original text likened this to the use of performance-enhancing drugs by athletes, but in semiconductors, doping is the process of adding specific impurity atoms at very low concentrations to control the concentration of charge carriers. To be precise, when a Group 5 element is introduced into silicon (a Group 4 element), four of the five valence electrons form covalent bonds with neighboring silicon atoms, while the remaining one becomes an electron that can move relatively easily. A semiconductor in which the electron concentration is increased in this way is called an n-type semiconductor. Conversely, adding Group 3 elements creates holes—vacancies where electrons are missing—resulting in a p-type semiconductor. By adjusting the type and concentration of dopants in this way, the electrical properties of a semiconductor can be tailored as desired. Unlike ordinary conductors, which allow a large current to flow but make it difficult to actively control the current, semiconductors have the distinctive ability to control the flow of current, which is why they play a crucial role in the manufacture of various electronic devices.
Among the devices made using this semiconductor material, the transistor—which played a pivotal role in ushering in the current digital age—is a prime example. A transistor is a semiconductor device capable of controlling the flow of current based on the applied voltage or current. Specifically, when used as a switch, a transistor can allow or block the flow of current under specific conditions, and when used as an amplifier, it can amplify a small signal into a larger one. By utilizing these characteristics of transistors, it is possible to create switches that turn on and off in response to voltage or current, amplifiers that boost signals, and circuits that control current. Furthermore, by using transistors as switches, logic gates can be created, and by combining multiple logic gates, complex digital circuits can be constructed. These digital circuits are integrated to form core semiconductor components such as a computer’s central processing unit (CPU) and memory. In the past, transistors and the circuits that utilized them were very large, so a single computer could be so massive that it filled an entire room or a lecture hall. Unlike today’s computers, early computers used vacuum tubes or early transistors and required a great deal of space and power. Subsequently, as integrated circuit technology advanced, the number of transistors on a single chip increased dramatically, and computers became smaller while their performance improved significantly. One of the key concepts in this development process is what is known as Moore’s Law. Moore’s Law is an empirical observation and long-term prediction that the transistor density in semiconductor integrated circuits has increased at regular intervals; it is generally understood to mean that transistor density doubles approximately every 18 to 24 months. However, rather than being a physical law, it is closer to an empirical trend derived from observations of technological advancements, investment, and manufacturing process developments in the semiconductor industry. While the semiconductor industry has largely followed this trend for a long time, it has recently become difficult to continue with simple miniaturization at the same pace as in the past. Consequently, the industry is evolving toward a approach that combines changes in transistor structure, new materials, advanced packaging, and technologies such as chiplets and 3D integration. Therefore, it is no longer appropriate to judge today’s CPU performance based solely on clock frequency. The performance of the latest processors is determined not only by clock frequency but also by various factors such as the number of cores, instruction-set processing capability, cache architecture, power efficiency, parallel processing capability, memory bandwidth, and accelerators. As it has become difficult to improve performance solely by continuously increasing clock frequency—as was done in the past—multi-core architectures, which integrate multiple CPU cores into a single processor, have become commonplace, and processors with dozens or more cores are now emerging. As computers have become smaller and more powerful than ever before, their price and accessibility have also changed significantly. It can be said that advancements in transistor manufacturing technology and integrated circuit technology have played a crucial role in bringing about these developments.
However, transistor manufacturing technology is entering a new phase where it is becoming difficult to continue following Moore’s Law in the same way as before. The process of continuously shrinking transistor size gives rise to various issues, such as leakage current, heat generation, power supply, and wiring resistance. While it cannot be concluded that further miniaturization is impossible simply due to the size of silicon atoms, physical limitations and manufacturing difficulties increase as the scale approaches the atomic level. Consequently, the semiconductor industry is moving beyond simply reducing the planar size of transistors and is changing the transistor structure itself. A prime example of this evolution is the transition from FinFET to the Gate-All-Around (GAA) structure, in which the gate more effectively surrounds the channel to control current flow. In addition to making transistors smaller, 2.5D and 3D packaging technologies—which involve stacking chips vertically or connecting multiple chips within a single package—are also gaining importance. Intel’s 18A process combines the GAA-based RibbonFET with PowerVia, a back-side power delivery technology, while TSMC is also continuously advancing its 2nm-class processes, 3D integration, and advanced packaging technologies. Therefore, future semiconductor technology will not advance simply by reducing the size of transistors, but will evolve toward the comprehensive utilization of new transistor structures and materials, back-side power delivery, chiplets, 3D integration, and advanced packaging. Furthermore, research into transparent and flexible transistors—as predicted in the original text—is ongoing, evolving into flexible electronic device technologies that can be utilized in wearable devices and various sensors. As semiconductor technology advances further, research is also underway on technologies that use extremely small sensors or chips to measure information inside the human body or for medical applications. However, these technologies cannot be realized simply by making transistors very small; various technologies—including sensors, wireless communication, power supply, and biocompatibility—must advance in tandem. Moving forward, semiconductor technology will enter a new phase driven by advancements in transistor miniaturization, structural innovation, 3D integration, advanced packaging, and low-power technology. These technological advancements will drive change not only in computers and communication devices but also in diverse fields such as automotive, healthcare, industry, and artificial intelligence, serving as a crucial foundation for shaping a new era.
