The data in Figure 1 is derived from a worldwide survey of 70 CEOs conducted by EY in April 2024. The optimistic outlook of CEOs in Japan's technology, media & entertainment, telecommunications (TMT) sector may be based on near-term market stability but, from a global perspective, the possibility of geopolitical risk remains elevated. More specifically, the semiconductor industry faces a number of uncertainties, including global tensions and trade frictions, and vulnerability due to the clustering of supply chains. Any revised strategies for the semiconductor supply chain must take concrete steps to mitigate risks instead of relying on optimism.
Hamajima believes that addressing certain risks would lead to changes in the silicon cycle due to the capital-intensive nature of the semiconductor industry. He emphasizes that the yield rate is a crucial factor that differentiating the semiconductor industry, while also noting that at the start of a plant operations, the yield rate is typically between 20 to 30% whereas industry competition can raise this to nearer 90%. He underlines also that output can vary greatly, even when using identical equipment with the same level of investment. He also mentions that the risk associated with introducing new production equipment could double if the equipment is located within a specific region. Similarly, if a globally optimized process needs to be completed in the region, this would also impact the silicon cycle.
How can businesses ensure the resilience of the semiconductor supply chain and respond to uncertainty in the face of escalating geopolitical tensions?
To address vulnerabilities in the semiconductor supply chain, consideration should be given to the following measures:
1. Diversification of manufacturing operations: Mitigate geopolitical risks by spreading semiconductor production across multiple countries.
(a) The US is providing tens of billions of dollars in incentives to bolster domestic semiconductor manufacturing operations. In addition, leading semiconductor companies such as Intel and TSMC are pushing to expand their manufacturing facilities in the US.
(b) India has introduced a range of initiatives, such as granting incentives and tax credits to foreign-affiliated semiconductor manufacturers, with the aim of strengthening domestic operations. While R&D in semiconductor manufacturing is also gaining momentum, Professor Wakabayashi stresses the importance of making strategic decisions about relocating manufacturing to India with the example of Foxconn which is exploring shifting part of its manufacturing capacity from China to new facilities in India. If India is able to leverage its highly skilled talent pool, he believes that it has the potential to become a significant player in semiconductor manufacturing. However, he also recognizes that India faces several infrastructure challenges, including its road network, power supply and logistics systems. He cautions that becoming over dependent on India could result in extended supply chains for certain processes and standard chips, further increasing geopolitical risks which include shipping through the Straits of Malacca and air freight from India.
(c) The Japanese government also earmarked JPY1.98 trillion in its supplementary fiscal 2023 budget to boost domestic capacity to make and secure semiconductors.*
Shozo Saito, president and chairman of the Japan Electronic Device Industries Association, adopts a geopolitical perspective in his views on the semiconductor supply chain. He suggests that the supply chain currently crosses too many borders, raising concerns about international dependence. He emphasizes the importance of security and affirms that globalization is acceptable as long as there is a firm guarantee of product availability e.g. if a product is manufactured in the US but there is an assurance of interrupted supply. However, as such guarantees seldom reflect reality, this leads him to believe that it would be wiser to reduce international dependence.
2. Securing stocks: It is crucial to maintain an ample supply of semiconductors to respond to fluctuations in market demand and potential disruptions in the supply chain.
3. Leveraging predictive modeling: Predictive modeling facilitates the identification of geopolitical shifts and fast responses to emerging risks.
Models driven by data analytics, artificial intelligence (AI) and machine learning are able to analyze factors like market movements, political statements and social unrest to forecast potential interruptions in the supply chain. This allows companies to make strategic decisions which mitigate risks and enhance the resilience of their supply chain.
4. Greater transparency in supply chains: Sharing information and ensuring transparency are essential for assessing and mitigating risks throughout the supply chain.
Saito comments that while it is not feasible to shift from a decentralized supply chain and return to a model dominated by integrated device manufacturers (IDMs), potential solutions for closer alignment could include cooperation between companies, the formation of consortiums or company mergers. He acknowledges that this complex issue cannot be resolved by a single entity and accepts that the industry in Japan is reorganizing under the guidance of the Ministry of Economy, Trade and Industry. He believes that these collaborative efforts are an effective way to tackle the challenges posed by the diversification of the semiconductor supply chain, semiconductor products and final products containing semiconductors.
5. Investment in technological innovation: Investment is crucial to sustaining a leading position in technology and raising the competitiveness of industries through continued R&D and innovation
Saito has also commented on the subject of technological innovation within the semiconductor industry, highlighting the two primary types of technological advancements being observed. Firstly, he notes the significant role of AI in semiconductor manufacturing, where it is utilized to manage the vast amounts of data produced during various stages of production. AI can now determine the most efficient processes and integrate them into the manufacturing workflow. While the previous objective was to expedite wafer delivery, the focus has now shifted to minimizing energy consumption and machines at idle by using AI-driven process design. He emphasizes the importance of minimizing waste, including temporary pausing of equipment, by leveraging computer automation over manual control for programming robots to transport wafers via the shortest possible route to improve efficiency. This practice involves creating digital twins, simulating processes and employing the most energy-efficient methods. Saito’s second observation relating to designing semiconductors capable of handling increasing hardware demands, particularly with the expanding use of AI, also raises the question of how the semiconductor industry will respond to the expanding AI market.
To strengthen the economy, government and business must address the vulnerability of the semiconductor supply chain in order enhance its robustness. The semiconductor industry must embrace continuous evolution and be confident about new challenges to ensure that its innovation and development are sustainable. Exploring technology, developing human resources and creating a culture of risk taking are key elements for the industry's long-term success. Our expectation is that the industry will continue to overcome these challenges and foster continued prosperity.