US Semiconductor–from invention, supremacy to despair

The US semiconductor industry started with the invention of its Bell Labs. Subsequently, it attained supremacy in semiconductor production due to the success of making computers better and cheaper. Notably, the rise of the PC wave made Intel and Silicon Valley seemingly unsinkable technology superpowers. But during the first two decades of the 21^st century, America has lost it. The USA now relies on Asia to import the most advanced chips. Its iconic Intel is now a couple of technology generation behind Asia’s TSMC and Samsung.

Furthermore, China’s aggressive move has added momentum to America’s despair, triggering a chip war. But why has America lost the edge? Why does it rely on TSMC and Samsung to supply the most advanced chips to power iPhones, Data centers, and Weapons? Is it due to Asian Governments’ subsidies? Or is it due to America’s failure to understand dynamics, make prudent decisions and manage technology and Innovation?      

Invention and rise and fall of US semiconductor supremacy

In 1947, Bell Labs of the USA invented a semiconductor device—Transistor. Although American companies developed prototypes of Transistor radios and other consumer electronic products, they did not immediately pursue them. But American firms were very fast in using the transistor to reinvent computers—by changing the vacuum tube technology core. Due to weight advantage, US Airforce and NASA found transistors suitable for onboard computers. Besides, the invention of integrated circuits by Fairchild and Texas instruments accelerated the weight and size reduction of digital logic circuits. Consequentially, the use of semiconductors in building onboard computers kept exponentially growing. Hence, by the end of the 1960s, the US had become a powerhouse in logic circuit semiconductors. But America remained 2^nd to Japan in global production, as Japanese companies were winning the race of consumer electronics by using transistors.

But America started gaining traction in global semiconductor production due to the selection of Intel’s processor in making the IBM PC. The profitable opportunity of increasing microprocessors’ transistor density fueling the PC diffusion made Intel the global leader.  In addition to Intel, Motorola, AMD, and TI kept pursuing pushing the envelope of microprocessor performances and process nodes.  

At the dawn of the 21^st century, America started losing Semiconductor shares to Taiwan and Samsung. There have been three significant reasons. First, America’s firm did not find memory chips doing a profitable business. The next one is the growth of fabless companies asking for 3^rd party foundry services. In retrospect, American integrated device makers (IDMs) like Intel or TI did not find delivering foundry services to small fabless companies lucrative. The 3^rd most crucial reason is the fact that American firms overlooked the opportunity to make smartphone possessors. Hence, America’s loss of semiconductor supremacy is due to a lack of insights into unfolding dynamics and decisions.

Chip Act spurs foundry investment on US soil

As high as $76 billion in grants, tax credits, and other subsidies have been encouraging US domestic chip production. To benefit from subsidies and tax incentives offered by the Chip Act, semiconductor companies have pledged nearly $200 billion in investments in plants. This $200 billion investment is for 23 new chip factories, the expansion of nine plants, and investments from companies supplying equipment and materials to the industry. Money will go to 16 states, including Texas, Arizona, and New York.

In September 2022, Intel gathered officials to break ground at a patch of land near Columbus, Ohio, for building two foundries at the cost of $20 billion. Micron Technology followed Intel by unveiling the plan to invest $10 billion in a memory chip plant in Syracuse, New York. Soon after these two announcements, in December 2023, Taiwan Semiconductor Manufacturing Company (TSMC) expressed the intention to triple the Phoenix, Arizona investment to $40 billion. In addition to these three significant investments, many other companies have declared assets to expand chip production on US soil. So far, as of Jan 2023, the total announced investments by 35 companies to set up semiconductor production plants on US soil has reached almost $200 billion.  

Experts express that the US Federal Government’s massive subsidies constitute the most significant U.S. investment to spur domestic manufacturing since World War II. All these efforts are to reverse the sliding of the USA’s share of global semiconductor production capacity, which has gradually slid to around 12 percent in 2022 from about 37 percent in 1990. Furthermore, USA’s share in high-end chip production is virtually nil. According to some estimates, Chip Act’s subsidy-led corporate investment will increase US shares of global semiconductor production to 14% by 2030. 

Regulatory measures to prevent China from making high-end chips

To regain semiconductor supremacy, in addition to offering massive subsidies to spur private investment, the USA has also taken regulatory measures. It has applied sanctions to the export of high-end chip-making equipment to China. It has also restricted Chinese companies from getting foundry services from high-end chip-making plants like TSMC’s 5nm foundry.

The USA has pursued Dutch Government to prevent exporting ASML’s EUV lithography machine to Chinese microchip makers like SIMC. Similarly, the USA has prevented TSMC from offering 5nm foundry services to process Huawei 5G chips. The USA’s trade restrictions have been geared to make sure that Chinese semiconductor makers remain a few process nodes behind. For example, while TSMC has been operating a 4nm fab to produce Apple’s A16 Bionic chip, the USA has been desperate to ensure that Chinese SMIC cannot have fab equipment to make chips using better than a 14nm process node.

To apply harsh trade restrictions, the USA has pursued its allies, including the Netherlands, Japan, South Korea, and Taiwan. In favor of these restrictions, the USA has cited national security and technology’s origin. As ASML’s EUV lithography machine has America’s Cymer’s light source technology, the USA has compelled Dutch Government to restrict the export of this critical chip-making technology to China. In October 2022, the U.S. introduced rules restricting chips made using American tools from being exported to China.  It also includes semiconductors designed for artificial intelligence applications.

The underlying cause of the loss of US shares in global chip production

In the US policy-making circle, it’s widely believed that the underlying reason for the sliding of the US position is the incentives of countries in Asia to move manufacturing to those shores. One of the notable examples of taking US shares and silicon edge has been Taiwan. This island nation’s share of global semiconductor output has grown from nothing in the 1980s to 22 percent in 2022. Furthermore, according to industry analysts and the Semiconductor Industry Association, as high as 90 percent of the most advanced chips are made in Taiwan. Has Taiwan achieved this remarkable success by providing subsidies to domestic and foreign semiconductor firms like the USA has been doing?

Another example of Asia is South Korea. In 2021, this Asian country’s market share stood at 19.9%, which has grown from almost nil in the 1970s to this level. Is South Korea’s success due to alluring foreign firms with subsidies?

On the contrary, what for Japan’s share in global semiconductor manufacturing has been falling? By 2022, Japan’s share of the global market fell below 10%. But in the 1980s, Japan was the world’s largest semiconductor producer. Over 50% of global semiconductor production used to come from Japan.

There is no denying that governments of Asian countries have been patronizing the local semiconductor industry. But is the subsidy the sole reason for the growth of the market share of Taiwan and South Korea? What kind of subsidy would make TSMC a global leader? TSMC has become a global leader due to its ability to deliver high-end chips, which American companies have been failing to deliver. Such capability development demands insights about unfolding dynamics, management decisions in the midst of uncertainty, and meticulous management of R&D for idea production and commercialization.

Fall of US semiconductor shares due to smartphone wave

Despite America’s success in producing computer microchips, Japan remained the largest semiconductor producer for over 30 years, still 1986. This is because Japanese companies were unleashing Disruptive innovation effects on American firms through the Reinvention of consumer electronic products. Although under the umbrella of SEMATECH, the American government channeled subsidies, there had been no notable success.  Instead, the US semiconductor industry started gaining supremacy due to the rise of the personal computer (PC) wave.

While American IDMs were making more than 60% gross profit in making chips for PCs, processors for mobile handsets kept growing. Unfortunately, American firms like Intel kept overlooking it due to its small volume and margin. Furthermore, fabless companies in making mobile processors started going to South Korea and Taiwan to get their designs printed on the wafer, as US firms refused to cater to their needs. Hence, along with the growth of the sophistication of mobile handsets, foundries in Taiwan and South Korea kept improving the process node maturity, volume of production, and margin. Consequentially, the US semiconductor industry kept losing both edge and global shares of semiconductor outputs.

The subsidy could not sharpen US’s semiconductor edge

In the early 1980s, the Regan administration got concerned with the continued dominance of Japanese semiconductors firms like Toshiba and NEC. Hence, the White House devised several punitive measures to create barriers to Japanese semiconductor firms. Furthermore, it formed SEMATECH for unfolding subsidy programs. Several US lithography equipment makers got a pie of this subsidy package. Unfortunately, none of the subsidy recipients succeeded in regaining their competitive edge. Eventually, all of the American lithography equipment makers got foiled, leaving the market to the Japanese and Dutch firms. 

Loss of America’s vital chip-making tool to ASML—is it due to Government subsidies?

In the chip-making process, lithography is the most vital and expensive machine. In the 1960s, American firms started making this machine. Within a decade or so, Japanese firms began to enter. Unfortunately, despite being pioneers, American firms kept falling behind due to the superior performance of Japanese steppers. To give life support, American Government came up with a subsidy package. But upon receiving millions of dollars, none of the American firms succeeded in sustaining due to poor innovation performance. On the other hand, late entrant Dutch firm ASML kept growing. Notably, ASML’s success in EUV lithography is remarkable.

Yes, the ASML EUV machine has some American components. But American firms failed to create a market for those components by pursuing the whole machine innovation.  Besides, ASML’s success in EUV is not due to subsidy. It has been due to ASML’s strength in technology, innovation management, and decision-making. Due to weakness in this vital area, American firms decided not to pursue it. Hence, pointing to a subsidy of other countries as the case of weakening US semiconductor firms appear to be untrue.        

Rise of Shares of Taiwan and South Korea in Semiconductor—not due to subsidy  

In the discourse of sliding US semiconductor market share, the rise of China is cited. But China’s top chip maker SMIC has only a 5% share of the foundry service market. Instead, Taiwan’s TSMC and UMC and South Korea’s Samsung have 54%, 7%, and 17% market share in the foundry segment. Furthermore, Taiwan and South Korea have a significant presence in OSAT and fabless semiconductor industry segments. In almost $600 billion output in 2021, Taiwan and South Korea have respectively 21% and 19.9% market shares.

Furthermore, in the 5nm process node, Taiwan’s TSMC and South Korea’s Samsung are a duopoly. In the 4nm node, TSMC has the monopoly status. Consequentially, TSMC produces 90% outputs of high-end chips. Now the question: is it due to subsidy and trade restrictions Taiwan and South Korea have reached this status? For sure, the answer is NO.

Yes, Taiwan’s semiconductor industry started the journey with Government funded technology transfer agreement between Taiwan’s ITRI and America’s RCA. First of all, RCA transferred one entire generation of older technology. And during the learning process, over 11 years, till the birth of TSMC, Taiwan fell behind by two and a half generations. For sure, by operating the transferred technology, Taiwan did not attain the pinnacle of success. Besides, Government investment in technology transfer should not be treated as a direct subsidy, like the way USA’s Chips Act has been providing. Instead, Taiwan’s success is attributed to prudent management practices in understanding dynamics, taking decisions, conducting R&D, and transferring R&D outputs in the market at a profit. It has been the success of creating a virtuous cycle. The success of South Korea, to a lesser extent though, appears similar. Hence, perhaps, it’s unfair to attribute their success to subsidies.        

Rise of china’s domestic semiconductor production—subsidy alone will not create the success

Unlike Taiwan and South Korea, China has been pursuing an import substitution strategy to develop its semiconductor industry. Some estimates suggest that China has been providing $150 billion in subsidies for acquiring foreign and established domestic semiconductor firms. With the subsidy in capital expenditure, domestic firms like foundry service provider SMIC and memory chip maker YMTC have been importing capital machinery to grow as competent semiconductor makers. Yes, there has been some success; but that is insignificant. Historical lessons suggest that subsidy-driven capital machinery alone will not create China’s semiconductor edge. Why have America’s Intel and Global Foundries lost their edge if that were the case?

Due to the growing role of knowledge and ideas in the form of IPs, the semiconductor industry has a natural tendency to monopoly. If the subsidy-assisted import and demonstration of domestic production do not lead to sufficient self-learning capacity, the endeavor will likely fail.  For example, in the 1990s, there were more or less 20 NAND flash memory chip makers, which has come down to 06 (as of 2023). It has happened due to the race to gain market power through superior performance. Hence, subsidies have not created Asia’s success, and China’s subsidy approach alone will not likely succeed.    

Regaining the edge of the US semiconductor industry—will hopelessness sustain?   

Yes, America’s loss in semiconductor edge is a cause of concern. Hence, there is a reason behind the despair. But what is the solution? Are subsidies for setting up additional plants good enough? Yes, they will expand chip production capacity on US soil. But will it take America to the edge and sustain that position? Perhaps, NO. As we all know, chip-making competence is not static. It keeps moving. Hence, plants that are being announced will be a generation behind before they even start production. Thus, America runs the risk of getting more hopeless upon spending so much money on subsidies.

Furthermore, what will America be trading from those plants? Most of the semiconductor industry’s knowledge and ideas are traded as chips and chip-making services. Unfortunately, how America will progress in this vital area is missing in its measures of subsidies and restrictions.  Hence, the focus should be on winning the idea race, instead of giving subsidies for expanding production using existing technologies.

Related articles

• Chip War
• Semiconductor Lithography Economics–fuelling Moore’s law and market power
• Microchips–invention, evolution, transformation and chip war
• Moore’s Law–fuelling monopolization, reinvention, and migration
• Moore’s Law is Dead–Chiplet redefines semiconductor industry
• Semiconductor–illusive technology core changing world order
• Winning Chip War–fuelling reinvention waves for changing world order
• Chiplet Technology — a weak reinvention core?
• Semiconductor Economics–will Chiplet era slow down the growth?
• 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
• Loss of America’s Inventions–blame semiconductor economics?
• Soviet Computer Failure—reasons and lessons?
• Semiconductor Industry in India–innovation failure
• Taiwan Semiconductor Industry–rise of global monopoly from a virtuous cycle