The environment surrounding the digital industry and the semiconductors that form the basis of the industry is changing drastically. Semiconductors are a fundamental technology that sits at the intersection of four important policy axes: digital, green, economic security, and local economic spillover effects. This intersection represents the starting point for the development of digital hardware and software industry. What is required for Japan to regain its global competitiveness in the digital industry and what are Japan’s strengths regarding the digital industry? Where should policy resources be invested? In this discussion, Mr. Kanazashi Hisashi, Director of the IT Industry Division of the METI, introduced the latest trends in the digital industry and semiconductors, Japan’s strengths, and policy challenges, and RIETI’s Senior Fellow Dr. Ikeuchi Kenta talked about challenges regarding academic analysis in this sector and the roles that academia can play.
What is Happening Now in the World of the Information Industry?
Fukuda:
First of all, could you tell us what fields the IT Industry Division is responsible for, what notable industrial and technological changes and impacts are occurring in those fields, and what are the information industry policy trends in major countries?
Kanazashi:
The IT Industry Division is in charge of a very wide range of fields regarding electronics. While Japan’s competitiveness has peaked in the 1980s in many fields of electronics, we are developing policies based on four axes: digitalization, greening, economic security, and local economic spillover effects, with computing capabilities at the core. With a belief that semiconductors are a fundamental technology that is positioned at the intersection of these four axes, we are promoting integrated industrial policy for three areas: hardware, communications hardware, and software. While it is important to strengthen each area, we are pursuing another goal of restoring the competitiveness of semiconductor-using industries.
The most notable initiatives related to semiconductors include the construction of a semiconductor plant in Kumamoto Prefecture by a joint venture between TSMC, Sony, Denso, and Toyota, and that of a Rapidus semiconductor plant in Hokkaido. While the minimum linewidth of logic semiconductors for production in Japan is currently 40 nanometers, the Kumamoto plant is designed to produce 12- to 28-nanometer semiconductors at its first building, 6- to 12-nanometer semiconductors at its second building, and 40-nanometer semiconductors as well. These semiconductors are expected to not only meet the needs associated with the electrification of automobiles, but also contribute to the improvement of data processing capacity for mobile phones and other edge devices, and the extension of the service life of storage batteries, among other items.
The Rapidus Hokkaido plant aims to mass produce 2-nanometer and even thinner-linewidth logic semiconductors. While there is currently no need for such thinner-linewidth semiconductors in Japan, it is important to ensure accessibility for users of cutting-edge semiconductor technology with an eye on the future. By developing an ecosystem that facilitates communications between semiconductor suppliers and users, we aim to allow both semiconductor suppliers and users to improve their competitiveness.
While South Korea, China, Taiwan, and other Asian economies continued massive industrial policy spending, Japan shied away from aggressive industrial policy by separating the private sector from the public sector to some extent, allowing the electronics industry to stagnate. When the competitiveness of the Japanese electronics industry became weak, the COVID-19 pandemic and Russia’s aggression against Ukraine occurred, disrupting supply chains. As semiconductor shortages were identified as a factor behind various important supply disruptions three years ago, we began to seriously consider how to secure our own semiconductor technology base, making a major policy change.
It is important for Japan not only to manufacture semiconductors on its own, but also to have them used by domestic industries. So, design is the key word here. For example, software is the basis for various autonomous driving systems; however, in order to perform high-level digital functions while satisfying the other purpose of reducing power consumption, it is necessary to optimize the entire system, including hardware. In fact, both Google and Apple have started to design semiconductors on their own, indicating that design ability is very important in the manufacturing of custom semiconductors.” So, we would like to increase policy measures and funding for such areas.
Japan’s Position and Strengths
Kanazashi:
Japan’s strengths in the semiconductor sector has long been supported by manufacturing equipment and materials. Japan probably has a global market share of about 50% for semiconductor materials and about one-third for semiconductor manufacturing equipment. We now need to work on how to maintain and expand Japan’s shares in these fields. If we focus only on semiconductor manufacturing equipment and materials, which represent Japan’s strengths, and neglect the development and manufacturing of advanced semiconductors, however, we may end up in a situation where cutting-edge research activities for semiconductor manufacturing equipment and materials exit Japan. Therefore, I believe that building a robust domestic industrial ecosystem will ultimately lead to further enhancing Japan’s strengths in semiconductor manufacturing equipment and materials.
In the semiconductor materials sector, there are many Japanese material manufacturers that have top global market shares. If these companies fail to invest at a speed that matches the growing demand, however, it would be quite possible for overseas manufacturers to fill supply-demand gaps to reduce the market share of Japanese companies. I think that we must provide support to allow Japanese companies to continue their steady investment.
In response to the advancement of artificial intelligence in recent years, foundation models have been attracting attention. For example, in the robotics industry, a key point is how to pool common data and integrate foundational models into it. So, we are always aware of the importance of having sound ecosystems including semiconductor and users in our policy development relating to AI.
Fukuda:
In addition to semiconductor-related materials and equipment, I think Japan’s strengths lie in semiconductor users. Japan may be able to take advantage of this point in competing with Europe and the United States.
Kanazashi:
In terms of Japan’s relationships with other countries, it is necessary for Japan to distinguish itself from other countries. We often say that Japan represents the world of edge devices and IoT while the United States represents the IT world. For instance, TSMC is building a plant in Arizona, the United States, where it will manufacture 2- and 3-nanometer semiconductors. The production in this TSMC plant will respond to demand from U.S. companies such as NVIDIA and GAFAM (Google, Amazon, Facebook, Apple, and Microsoft), indicating that the U.S. IT industry is a customer for the plant. On the other hand, unlike the United States, Japan has an IoT-based industrial base and structure. How Japan can utilize AI in areas such as automobiles, robots, medical equipment, and space is a key question. Japan’s greatest strength is its large variety of industries, so if it can deliver cutting-edge custom semiconductors to individual industries, such user industries in Japan will once again be able to compete on the world stage.
Ikeuchi:
So, it’s about having a foundation in manufacturing, isn’t it?
Kanazashi:
That’s right. However, it is necessary for various industries, and especially those on the user side of edge devices, to incorporate AI functionality. I think that if this is combined with hardware optimization, it will lead to industrial competitiveness.
Fukuda:
How do you view the growing interest in the information industry and the approach of academic analysis in this field from an academic perspective?
Ikeuchi:
Academic interest has largely been focusing on IT market structure and competition, particularly regarding cloud services. Although there has been groundbreaking analysis dealing with the impact of IT technologies on productivity, or the relationship with market monopolies in platforms by players like GAFAM, I have the impression that the impact of information technology has not yet been sufficiently analyzed. When human labor is replaced by AI robots, it’s not just simple IT services; the hardware of machines also come into effect. I feel that there has not been sufficient research in this area.
Kanazashi:
The economic effects of increased labor productivity through human labor substitution are starting to appear in the data. It would be helpful if the academic community, including RIETI, could report on these quantitative effects. It would support the policies that we are working on and would be appreciated.
Ikeuchi:
Experiments with automated robots and autonomous driving are difficult because of relevant regulations.
Kanazashi:
As you can see from “digital twin” technology, the use of simulation technology has expanded considerably. The remaining challenge is how to conduct field experiments once simulations have been refined.
Ikeuchi:
Even if the simulations progress, I think that if regulations prevent final experiments from being conducted, it will be difficult for the industry to grow.
Policy Responses to Date and Future Challenges
Fukuda:
What policies have you focused on so far, and what challenges will you need to address in the future?
Kanazashi:
So far, we have been supporting large-scale investments with large economic ripple effects. In the future, we would like to strengthen our support for design, particularly in terms of interactions with users. One example is an initiative that we started last year in the automotive industry. When it comes to semiconductor usage, each company has cooperative areas and competitive areas. In terms of the cooperative areas, we started to provide automaker technical research association “Advanced SoC Research for Automotive (ASRA)” with public financial support for its design project. We aim to expand such initiatives in the coming years.
As the industrial base for advanced chips is established in Japan, there will be more opportunities for communication with chip suppliers in Japan, generating incentives for semiconductor manufacturing equipment and material companies to promote cutting-edge R&D domestically. To accelerate this, we will focus on creating experimental facilities and hubs for open innovation. Additionally, we will work with academia to foster human resources development to create a deep industrial structure.
Ikeuchi:
The main support has been provided to large companies, but are venture businesses and startups also being considered?
Kanazashi:
We have high expectations for venture businesses, especially regarding new ways of using semiconductors. In fact, we are in discussions with domestic ventures regarding providing support.
Also, theoretical analysis in an academic manner can only be undertaken by academia, and our cooperation with universities and technical colleges is indispensable for the development of next-generation human resources. Therefore, communication with academia is essential.
As the use of AI in the design field spreads in the future, it will be important to democratize such AI use, to allow many people to take on that role. This will become an important source of added value.
Fukuda:
The IT Industry Division also deals with computing infrastructure. Do you have any policies to encourage the use of computational resources for various fields such as developing materials and drugs?
Kanazashi:
Naturally, there are areas where AI is integrated in edge devices and areas where AI services are housed within the cloud. To that end, it is important to have a computing infrastructure within Japan and to ensure that industries on the user side can utilize it. In a macro sense, this will help to eliminate the digital trade deficit while fostering competitiveness in industries, so, we are also supporting cloud service providers.
Fukuda:
The ability to access and use data centers when necessary is crucial; however this requires a certain level of literacy, so support should be provided to facilitate this.
Kanazashi:
Ultimately, I think it is important to eventually democratize datacenter use. We need to form a community where the datacenter users and service providers can evolve rapidly within a supportive ecosystem in Japan.
Fukuda:
The information industry is vast, spanning from upstream to downstream sectors and even to users, and it seems that no single country or company can cope alone. How have you analyzed and understood complex supply chains with various players? How much remains unclear?
Ikeuchi:
In terms of global business relations, international organizations such as the OECD are developing platforms that are used to analyze the flow of goods and trade relations at the industry level by collecting statistics from each country. There is also progress in developing micro data, including databases being built to include business relationships at the company level, but there is still a lack of detailed data covering the international division of labor.
Fukuda:
The 9-digit Harmonized System codes sometimes cover both high-end and low-end products. Could it be that there is no option but to collect micro-level data in each industry or company?
Ikeuchi:
It depends on how comprehensive we want the data to be and for each company it is a critical issue.
Kanazashi:
That kind of awareness is gradually getting stronger. From the perspective of building reliable supply chains, for instance, Europe is considering regulations to measure carbon dioxide emissions across entire supply chains, especially in areas like storage batteries. I would like to take this as an opportunity to consider how each industry as a whole can promote information sharing beyond individual companies, while considering confidentiality of information from individual companies.
Hopes Placed on Industry and Academia
Fukuda:
Lastly, could you provide a message regarding your hopes for industry and academia.
Kanazashi:
We often speak of industry, academia, and government collaboration, but I hope that these sectors can develop a perspective on how to better leverage each other’s strengths. For example, the U.S. government and industry stakeholders often speak freely, but when it comes to critical issues, they recognize their respective roles and collaborate effectively. In Japan, people tend to draw lines between the public and the private sector. Public servants are responsible for providing services to everyone, so I hope that industry will adopt the mindset that they should take advantage of what the government has to offer. I also hope that academia will develop a more advanced system of industry-academia-government cooperation whereby academic institutions increasingly pitch ideas to domestic businesses rather than focusing on overseas opportunities in hopes of commercializing their research.
Ikeuchi:
I think it would also be beneficial for the government to make use of any promising technologies that academia develops to create value. This would also lead to a better industry-academia-government cooperation system. We in academia would like to support such developments.
(Honorifics omitted)
