Over the last few decades, we all got overwhelmed by the growth rate of smartphones, digital cameras, the internet, PC, and many more products. We have also been making projections of the astronomical growth of many other products, from IoTs to wearable devices. On the other hand, the rise of Startups like Sony, Apple, Microsoft, Intel, and many others bewildered us. The rapid monopolization of global industries is also intriguing. Besides, the migration of Innovation epicenters of Radio, Television, data storage, camera, silicon chips, displays, light bulbs, and many more from the USA and Europe to Japan, South Korea, and Taiwan demands clarification. Economists are also confused in explaining the unfolding dynamics. But what is the underlying factor? The answer appears to be semiconductor economics.
In the 1950s, except a few, no one was using transistors. But its adoption kept growing, reaching more than a billion in everyone’s pocket. What is the underlying reason? The answer is simple—semiconductor economics of simultaneous quality improvement and cost reduction.
Higher quality products cost more is a natural phenomenon. Scarcity and trade off are the core challenges for economists to deal with in optimal resource allocation. But semiconductor economics kept proving wrong, making economists confused. On the other hand, Dr. Vannevar Bush’s endless frontier thesis kept marching forward. Semiconductor economics offering increasing quality at decreasing cost kept incremental advancing and reinventing products, resulting in growing diffusion of transistors in society. Consequentially, over the last 70 years, like air and water, semiconductor devices like transistors have become most abundant—making many economic theories questionable! Sustained semiconductor economics have made the trade-off irrelevant, and the market has lost invisible hands, resulting in growing monopolization.
Semiconductor economics gave birth to Moore’s Law and the Natural tendency of monopoly
In 1963, upon trading transistors (with part number 1211) at $150 apiece to the US military, Fairchild was making a healthy $50 profit per piece. But at that price, no product could be developed for civilian applications. Even a simple AM radio receiver needing four transistors would cost over $600. To reach the cost budget of the civilian market (UHF tuner), upon trading off quality and sourcing low-cost labor from Asia, Fairchild managed to reduce the price to $0.5 apiece in 1965. But this trade-off exercise became irrelevant with the invention of the planner integrated circuit (IC) in 1959 by Dr. Moore and Dr. Noyce. The finding of integrating a growing number of transistors on the same area of the silicon die, making them smaller and closer, resulted in quality improvement and cost reduction simultaneously–a critical achievement that bewildered many economists and strategists down the road.
In a 1965 paper, Dr. Moore reflected on the growth of complexity and Transistor density in IC, resulting in cost reduction. He observed that the transistor density for minimum component costs had increased “at a rate of roughly a factor of two per year”. The continuation of this trend would lead to the transistor (component) density increase by a factor of approximately 1,024 over ten years. He also observed that the density advancement was contributing to per-component cost reduction at a rate of about 37% per year. The continuation of this rate of cost reduction would result in a “90% cost decrease in five years and a 99% cost decrease in ten years”.
Of course, for higher component density, that would be growing capital expenditure, but that higher cost would be far more than offset by the annual doubling of density. Hence, the race of R&D started to increase component density to reduce cost.
Consequentially, Moore’s law got birth—doubling transistor density every 12 months. In 1975, Dr. Moore reset forward expectations to a doubling of transistor density every 18 to 24 months versus every year. This aspect of Moore’s law is very much within the comprehension of economic theories.
Race of density increase for cost minimization led to performance improvement—giving birth to the natural tendency of monopoly
Interestingly, unlike many other products, the shrinking size of the transistor not only contributed to cost reduction; most importantly, it contributed to quality improvement significantly. Shrinking size and falling distance between components led to (i) reducing noise, (ii) increasing frequency, (iii) lowering latency, (iv) expanding bandwidth, and (v) falling energy dissipation. As a result, cost reduction led to quality improvement—making the trade-off irrelevant. Hence, competition intensified for offering higher quality at less cost, leading to price setting capability of the winner. Furthermore, minimum efficient scale (MES) kept expanding with the advancement of density.
This trend of quality enhancement with cost reduction has been a significant reason for the exponentially growing demand for transistors—making us “transistor trillionaire”. Due to it, innovators have been reinventing all kinds of products, resulting in the rise of startups and the fall of giants.
Consequently, the semiconductor industry started showing a natural tendency to monopolize, making Adam Smith’s invisible hands increasingly weaker to govern the competition. Such reality has been the underlying cause of falling numbers of leading-edge processor manufacturing firms from 29 in 2001 to 02 in 2020, leading to only TSMC in 2022.
Process equipment, chemicals, device design, and software kept driving semiconductor economics
Process equipment, chemicals, gas, device design, and software have been playing a critical role in driving semiconductor economics and keeping Moore’s Law alive. Equipment kept demanding continued enhancement to increase the density, resulting in growth in number and capability. One of the notable ones has been the lithography machine. From a humble optical projection instrument, it has become a scientific wonder of generation and guiding 13.5nm extreme ultraviolet (EUV) to project images of chip designs on the wafer. Similarly, there has been continued enhancement in vapor deposition, testing, metrology, cleaning, etching, and defect detection equipment. There has been a growing demand for device architecture, the reusable IP module, and design software to leverage the scope of increasing density, reaching billions in a single die. Consequently, there has been a growing specialization in every layer of the semiconductor value chain, leading to a monopolistic situation.
Semiconductor economics kept monopolizing the semiconductor industry
As explained, the underpinning of semiconductor economics is the power of ideas to increase the quality and simultaneously reduce the cost of components. This is a perfect recipe for attaining price-setting capability by winning the race of offering increasing quality at a decreasing cost. Hence, the semiconductor industry has a natural tendency of monopoly at each layer of the industry value chain. For example, a few Japanese companies have Monopolistic market power in wafers, chemicals, and gas. In sub-10 nm lithography, ASML has attained absolute monopoly status.
Similarly, the USA has market power in fabless IC companies, IDMs for PC and server markets, and EDA software tools. On the other hand, both the USA and Japan have dominant positions in the fab process equipment market. Besides, TSMC and Samsung have attained dominant positions in the high-end wafer processing capability. For leading-edge monolithic chip fabrication, TSMC commands as high as 90% of the global market.
Reinvention of products raising startups and failing incumbents
Leveraging transistors started reinventing computers, control systems, telephone switches, and consumer electronics. While US companies like Fairchild and Texas Instruments were busy supplying discrete transistors and ICs to US military and space programs, Japanese Sony led Japan to start reinventing consumer electronics.
Notably, Sony’s success in reinventing Radio and Television by changing vacuum tubes with transistors led to the rise of Sony from a radio repairing shop to a global icon. Due to Sony’s rise, American RCA tumbled as its products out of vacuum tubes could not cope. Such a reality raises the question of why RCA could not pursue the same path of reinvention. Surprisingly the answer lies in latent semiconductor economics.
The transistor’s early emergence was neither better nor cheaper than vacuum tubes. For example, in 1963, RCA’s Nuvistor vacuum tube costing $1.05, was far less costly than Fairchild’s part number 1211, selling for $150 apiece. On the other hand, the analog signal passing through the transistor added noise. Besides, the signal power handling capacity of the transistor was far less than that of vacuum tubes. Hence, for an apparent reason, RCA avoided replacing vacuum tubes in its Radios, TVs, and other consumer electronics with poorer and costlier Transistors. Ironically, to the surprise of RCA and many others, reinvented Japanese Radios, TVs, and other items kept rapidly getting better and cheaper, unleashing creative destruction power on incumbent firms like RCA.
The long-run way of semiconductor economics in making transistors better and cheaper led to the rise of many startups. Notable examples have been Intel, Microsoft, Apple, and many more. On the other hand, such reality kept making incumbent firms bewildered, suffering from Disruptive innovation. Hence, companies like IBM and Kodak suffered from a massive burn.
Rise of Japan and Taiwan in the idea economy during the golden era of semiconductor
In developing Nobel Prize-winning idea and object theory, Prof. Paul Romer argued that the underlying reason of rise of Japan and Taiwan had been due to their graduation from idea importers to exporters. But he did not explain how they got a scalable path of producing and exporting ideas. On the other hand, despite having an import substitution policy and strong engineering institutions like IITs and national laboratories, why could India not succeed in reaching high-income states? Besides, despite having sound infrastructure, education and R&D capacity, a Market Economy, and democracy, why could not Canada or Australia benefit much from an idea economy?
Economists do not know well about Japan and Taiwan’s rise and many others’ failure to keep growing. The failure of the World Bank’s growth commission, headed by Nobel Laureate Michael Spence, in finding reoccurring patterns driving sustained growth underscores this reality. Japan took advantage of reinventing American and European-invented products with transistors. Yes, they emerge in an inferior form. But the continued advancement of semiconductors with locally produced ideas led to getting them better and cheaper. Consequentially, Japan kept leveraging a long runway to create economic value by producing ideas and exporting them as increasingly better products.
Even after 40 years of invention, the semiconductor was still amenable to getting better and cheaper, keeping Moore’s law alive. Curiously, it remained amenable over the next 30 years. Hence, upon entering the industry in 1985, Taiwan kept leveraging the option of generating and profiting from ideas of silicon chip-making–graduating Taiwan a high-income economy. Without semiconductor economics, Japan and Taiwan’s growth path could have been entirely different.
Semiconductor economics makes Economists confused in understanding market and development
Scarcity is one of the challenges for economists to deal with. Hence, they are after the tradeoff for optimal allocation of resources. On the other hand, they rely on invisible hands to govern the market. Furthermore, they believe in collusion to monopolize the market, resulting in curtailing market power through antitrust laws. They also believe in the middle-income trap and are helpless in offering solutions. Surprisingly, semiconductor economics—a long-run way of making transistors increasingly better and cheaper—keep questioning the merits of their beliefs and tools.
For example, semiconductor economics has made the scarcity and cost-quality trade-off irrelevant. You could be a better-quality provider at less cost. On the other hand, the race to offer better quality at decreasing cost out of the exploitation of semiconductor economics instead of collusion has led to growing monopolies in many industries. Unfortunately, economists often fail to understand, leading to the wrongful application of antitrust laws. Ironically, one of the worst victims has been transistor inventor America.
Decaying Moore’s law runs the risk of slowing down the economy—making economists confused again
As explained, over the last 70 years, semiconductor economics kept unfolding the reality of higher quality at less cost, a confusing fact to comprehend. The rapid pace of creation out of destruction, causing firms to rise and fall and migrating innovations across the boundaries of firms and nations, has kept confusion in constant flux. But that long journey of making irrelevant trade-offs is ending—ceasing Moore’s Law. Hence, the future of transforming products, markets, firms, industries, and economies will keep deviating from the past. But as the economists develop their comprehension models by processing past data, they will find their models failing again. Hence, we need to understand semiconductor economics better to comprehend the unfolding dynamics and predict the future.
Yes, by leveraging semiconductors Japan and Taiwan have experienced phenomenal growth. There is no denying that the USA has lost the silicon edge and epicenters of many inventions through the race. But as Moore’s law has been weakening, the scope of creating Wealth out of ideas will likely be getting narrower too. Hence, gaining supremacy in semiconductors will probably not repeat past successes like reviving Silicon Valley. Similarly, by preventing access to the latest semiconductor technologies, the USA will not get many edges to slow down the growth of China. On the other hand, by trading off quality for cost reduction, Chiplet technology will not likely repeat the past success of the integrated device makers.
Stacking Transistor atop each other–possibility of giving another life to Moore’s Law
As transistors cannot get smaller, what about going up build them atop each other as a three-dimensional multistoried structure? Hence, engineers are after exploring further possibilities for 3D fin-shaped devices. Other candidates are the gate-all-around (GAA) structure and RibbonFET. Experts are of the opinion that 3D-stacking of complementary metal-oxide semiconductor (CMOS), or CFET (complementary field-effect transistor), may extend Moore’s Law into the next decade. But turning each of the possibilities will be demanding decade-long R&D and facing the challenge of transferring the R&D outputs into profitable revenue.
So far, semiconductor economics has succeeded due to the evolution of transistors. It resembles climbing one mountain after another. Only when engineers get to the top of one they could see the vista beyond and map a route to climb the next; but it happens that successive mountains are getting taller and steeper, demanding growing investment. Hence, in addition to overcoming the technological challenge, there has been a growing hurdle to transferring the exponentially increasing R&D investment into a profitable return. Furthermore, there has been competition in climbing the same mountain. Hence, to exploit the possibilities at the summit, you need to climb there earlier than your competitors. Furthermore, you would not get much time for cultivating and harvesting, as competitors will start climbing the next one to cultivate better transistors at less cost.
...welcome to join us. We are on a mission to develop an enlightened community for making the world a better place. If you like the article, you may encourage us by sharing it through social media to enlighten others.
- Chip War
- Chiplet–for IDMs’ winning, survival or death?
- Chiplet Technology — a weak reinvention core?
- ASML–growing pearl gets caught in Chip War
- China’s Semiconductor Independence–prematurely caught?
- India’s Semiconductor Dream–pushed in the slow lane?
- Semiconductor Value Chain–globally distributed ecosystem
- Semiconductor IndustryWaves
- Intel Falling Due to PC and Mobile Waves
- ASML Lithography Monopoly from Sustaining Innovation
- Taiwan’s Semiconductor Monopoly – How did it arise?
- ASML TSMC Nexus Fuels Semiconductor Monopoly
- ASML Monopoly in Semiconductor — where is magic?
- SEMICONDUCTOR MONOPOLY DUE TO WINNING RACE OF IDEAS
- Semiconductor Industry Growth–personalities, new waves, and specialization underpin
- Transistor–technology core shaping global trade and power