From high-end smartphones to fighter jets, edge in all significant state-of-the-art innovations demands next-generation silicon chips. If you are not allowed to have it, you are in no way capable of pursuing your dream of acquiring an innovation superiority, whether in civilian or defence markets. Hence, one easy way to ensure that the incumbent’s technology edge is not under threat is to prevent aspiring new entrants from getting the latest chips through import or domestic production. For this critical reason, to prevent China from rising as a global technology leader, the USA has zoomed in on a crucial plan-to prevent China from accessing high-end silicon chips produced with the latest technology node. The USA would also like to ensure that the next generation chip production takes place on its soil so that China cannot risk supply by threatening smaller nations producing following generation chips.
Furthermore, for meeting this plan, the USA is after confining semiconductor value chain activities, starting from needed chemical production and fabrication to chip packaging in allies, forming the Blue supply chain. Hence, after a long discourse with academia, industry, think tanks, and politicians, the USA has come up with the Chips and Science Act. After WWII, this Act is perceived to be the most substantial government response to consolidate the industry’s supply chain to maintain superiority and prevent potential threat.
HISTORY OF SEMICONDUCTOR VALUE CHAIN AND ITS RISE AS A POWERFUL TECHNOLOGY CORE: In 1947, Bell Labs invented a solid-state switch–called Transistor–to make telephone switches more reliable. Like many other technologies, it emerged in primitive form. In addition to replacing bulky electro-mechanical telephone switches, American companies started reinventing on-board signal processing and computing modules of fighter jets, naval vessels, and space equipment like satellites by replacing energy-hungry, bulky vacuumed tubes with transistors. On the other hand, Japanese companies led by Sony embarked on reinventing Radio, TV, and an array of consumer electronics products. For leveraging the latent potential, both Japanese and American companies embarked on refining it, leading to growing specialisation in all the inputs needed, from chemicals and design automation software to testing equipment, to make increasingly better transistors at decreasing cost.
Subsequently, European firms joined the race to produce transistors to keep improving their electronic products. For sourcing labour-centric services for testing, bonding, and packaging, Singapore, South Korea, and Malaysia were brought into the value chain. By setting up industrial technology research institute (ITRI) and licensing a fabrication plant from American RCA in the 1970s, Taiwan started gathering momentum for making an entry-leading to the formation of a semiconductor cluster, seeded by the spinoffs of united microelectronics corporation (UMC) and Taiwan Semiconductor Manufacturing Company (TSMC). Lately, China has become aggressive in developing semiconductor production capacity–through the acquisition of foreign firms, and investing in developing domestic capacities by giving massive subsidies. Unfortunately, China has not succeeded in attaining any unbeatable edge.
Although semiconductors started the journey to reinvent and incrementally advance standalone electronic products like television or computer, over the decades, a growing number of industrial products found this technology helpful in making them increasingly better performing. For example, once thought to be pure mechanical product, each modern car has dozens of microcontrollers, making the automobile sector consumer of almost 20 per cent of global outputs of semiconductors. The race to make industrial products smart demands using high-end chips to add intelligence to machines. Hence, winning the innovation race of most industrial products depends on silicon sharpness.
RELEVANCE OF HIGH-END SILICON CHIPS FOR INNOVATION EDGE: The innovation edge of industrial products, from smartphones to missiles, has migrated from excellence in material, energy conversion, and aesthetics to computationally intensive human-like intelligence-artificial intelligence. As a result, the innovation race has been asking for silicon chips with a growing number of transistors, with decreasing power dissipation-leading to more than 10 billion transistors on a single chip. For example, Apple’s bionic chip used in iPhone has 11.8 billion transistors. As chip makers succeed in the next technology node tom package, more transistor performance increases, cost decreases, and power consumption decreases. For example, as reported by Samsung, migration from 5nm to 3nm node leads to 45 per cent more energy efficiency and 23 per cent better performance. Hence, having access to increasingly better chips or chip production capacity is at the core of pursuing an innovation edge in making dumb machines into intelligent ones.
MAKING IS NOT GOOD ENOUGH DUE TO DYNAMIC INNOVATION: Products have been evolving through reinvention and incremental advancement-turning the next version increasingly better and often less costly. Furthermore, such evolution has a natural tendency to reduce labour providing roles of humans in both production and usage. Hence, having the capability of making is no longer good enough, as the next version erodes making based competitive advantage further. For example, through protection and a sizeable domestic market, India acquired the capability of making more or less all parts of automobiles. But due to failure in advancement, India could not make the domestic automobile industry globally competitive. Furthermore, the emergence of reinvention of automobiles as electric vehicles poses a grave threat of turning India’s making capability irrelevant.
Hence, success depends on improving whatever we produce through a flow of ideas. It happens to be that those additional ideas of strengthening most of the products have been demanding next-generation chips for implementation. Hence, access to semiconductors’ edge is going to determine the competitive advantage of the industrial economy, whether for service or goods.
WINNER TAKES ALL OUT OF THE RACE OF REFINEMENT: The race of software-centric innovation, running on silicon chips, has a natural tendency to let the winner acquire price-setting capability. Zero cost of copying software and a long runway of improving the quality through software-centric ideas empowers the best performer to offer the highest quality at the lowest cost-resulting in monopolising the market out of specialisation. It has been happening in all major industries due to the growing role of software. Furthermore, all significant layers of the semiconductor value chain have been monopolised due to it.
Over the last 70 years, the race to make transistors better and cheaper and fabricate a growing number of them on the same-sized silicon die has monopolised all layers of the semiconductor value chain. For example, Japan has a monopolistic situation in wafers, chemicals, and equipment. Dutch ASML has attained a monopoly in lithography. US firms have a monopoly in software and designs of chips. Taiwan has achieved a monopolistic edge in fabrication. It happens to be that these monopolies are either in the USA or US allies-making it easier for the USA to form the Blue supply chain of high-end semiconductors-preventing China from accessing silicon edge. Hence, through the Chips Act, the USA is well positioned to limit high-end semiconductor production and consumption within the US-led network. As a result, China runs the risk of not only failing to produce high-end chips but also, most importantly, China’s dream of pursuing global innovation race will likely be seriously constrained. Therefore, aspiring developing countries should consider these unfolding geopolitics to find entry routes in new segments, expand existing capacity, and retain whatever they have already achieved.
M. Rokonuzzaman, Ph.D is academic and researcher on technology, innovation, and policy.
