The Japanese semiconductor industry dominated the global semiconductor throne for seven years (1985-1991) due to the rise of DRAM in the late 1980s to early 1990s, significantly aided by the VLSI Technology Research Association. The VLSI Technology Research Association, which existed from 1976 to 1980, included seven companies: Fujitsu, Hitachi, Mitsubishi Electric, Tokyo Shibaura Electric (now Toshiba), NEC, NTT Data Toshiba Information Systems, and the Computer Comprehensive Research Institute. They collaborated to develop manufacturing equipment and technologies for producing large-diameter, high-quality wafers aimed at VLSI. Over four years, the VLSI Technology Research Association applied for more than 1,000 patents, which laid the foundation for Japan’s future success over the United States. On September 22, 1985, finance ministers and central bank governors from the United States, Japan, West Germany, France, and the United Kingdom (referred to as G5) held a meeting at the Plaza Hotel in New York, reaching an agreement for a coordinated intervention in the foreign exchange market to induce an orderly depreciation of the dollar against major currencies to address the massive U.S. trade deficit (this agreement is known as the Plaza Accord). In September 1986, the U.S. pressured Japan to sign the first U.S.-Japan Semiconductor Agreement to combat the Japanese semiconductor industry; in 1991, dissatisfied with the agreement’s effects, the U.S. forced Japan to renew the second U.S.-Japan Semiconductor Agreement. The Plaza Accord and the U.S.-Japan Semiconductor Agreements had long-term and profound impacts on Japan’s semiconductor industry, electronics industry, and overall economy. Subsequently, the U.S. semiconductor industry shifted its strategy from DRAM to CPU and GPU, but the Japanese semiconductor industry failed to keep pace, leading to a decline and entering a “lost 30 years”. The period from 1970 to 1995 was dominated by home appliances, from 1985 to 2010 by personal computers, and from 2000 to 2025 by smartphones. Japan capitalized on the first wave but missed the second and third waves. Home appliances made physical spaces more convenient through analog technology, personal computers created virtual spaces through digital technology, and smartphones made virtual spaces portable through wireless network technology. The upcoming fourth wave will integrate network and physical spaces through the use of sensors, AI, and motors to promote economic development and solve social issues. In other words, it aims to create a human-centered society using “digital twin” technology, referred to as “Society 5.0”. How will the Japanese semiconductor industry develop in the fourth wave? Intel and AMD’s CPUs, as general-purpose serial (Von Neumann architecture) computing chips, dominate the world; in the era of artificial intelligence, NVIDIA’s GPUs, as general-purpose parallel computing chips, occupy the global market. However, history tends to repeat itself. From 1985 to 2000, Application-Specific Integrated Circuits (ASICs) created a large market. In the 1980s, the rise of CAD, FPGA, and CPLD improved chip design efficiency by three orders of magnitude. By around 2000, due to Moore’s Law, integration levels increased by three orders of magnitude, requiring more manpower and time than before, even with the use of computers. As a result, the ASIC business became unprofitable and came to an end. However, around 2010, the game changed again. Giant IT companies like Google, Apple, Facebook, and Amazon gradually realized that continuing to source general-purpose chips from semiconductor manufacturers like Intel, AMD, Qualcomm, and NVIDIA would not allow them to win in competition. Therefore, these industry giants began to develop their own dedicated chips. To regain the glory of 30 years ago, the Japanese semiconductor industry is once again focusing on dedicated chips. On December 21, 2022, the Japan Technology Research Consortium’s Leading-Edge Semiconductor Technology Center (LSTC) was established. The mission of this center is to create an open research and development platform to enable the mass production of next-generation semiconductors with a process of 2 nanometers or smaller within a short turnaround time, conducting research and formulating technology strategies covering design, devices, manufacturing, equipment, and materials. Additionally, to ensure the long-term prosperity of the semiconductor industry, LSTC will focus on cultivating talent in the semiconductor field. Members of LSTC include the National Institute of Advanced Industrial Science and Technology, RIKEN, the National Institute for Materials Science (NIMS), Tohoku University, Tsukuba University, the University of Tokyo, Tokyo Institute of Technology, Osaka University, the High Energy Accelerator Research Organization, and Rapidus, which is responsible for mass production. Rapidus was established in August 2022 as a new company tasked with Japan’s advanced manufacturing mission, starting with 2-nanometer processes to compete with the advanced process technologies of Intel, TSMC, and Samsung Electronics (notably, Rapidus products are not allowed to be sold to China). In addition to Rapidus and LSTC, the University of Tokyo established a publicly cooperative center on campus in October 2019—the Systems Design Lab (d.lab) of the Faculty of Engineering at the University of Tokyo, which also utilizes TSMC’s Open Innovation Platform Virtual Design Environment (VDE) for IC design. The “d” in d.lab stands for digital inclusion, where everyone can shine using digital technology, starting from data, integrating software and devices to achieve domain-specific system design. In August 2020, the University of Tokyo established a technology research organization for industry-government-academia collaboration based on strict information management—RaaS (Research Association for Advanced Systems). RaaS stands for Research as a Service, which is also its goal: to provide research as a service. In 2020, Toppan Printing, Panasonic, Hitachi, and Marubeni joined RaaS. In April 2023, Advantest and RIKEN joined RaaS. In 2021, Japan established a 3D integration technology research and development organization, NEDO (New Energy and Industrial Technology Development Organization), composed of a strong technical team. Companies like Skulink Group, Panasonic Connect, Daikin Industries, and Fujifilm joined this organization, focusing on developing 3D integration technology for chips. TSMC established a subsidiary for 3D IC research and development in Tsukuba City, Ibaraki Prefecture, Japan, in March 2021, and the 3D IC R&D center officially opened in June 2022. Thus, the development direction of the Japanese semiconductor industry is very clear: 1. Miniaturization. Institutions and companies represented by LSTC and Rapidus aim for research and production of miniaturized chip manufacturing technology; 2. Rapid development. Organizations and institutions represented by the University of Tokyo’s d.lab and RaaS aim to improve chip energy efficiency and shorten chip design cycles (including dedicated chips); 3. 3D integration. Organizations and institutions represented by NEDO and TSMC’s Japan 3D IC R&D center aim for research and production of 3D integration technology. Furthermore, Japan actively collaborates with companies/institutions like IBM, TSMC, and IMEC (Belgian Microelectronics Research Center) to cultivate a “forest” for the semiconductor industry, promoting the symbiosis and co-evolution of the industrial ecosystem. Interestingly, both the VLSI Technology Research Association and LSTC were born on the threshold of entering the dedicated semiconductor era. At this moment, it is reminiscent of that time. To quickly customize dedicated chips, Japan may once again ignite a wave of FPGA design. I find it strange that, in an era where chip manufacturing processes have reached the nanometer level and chip integration scales have exceeded tens of billions of transistors, especially with CPU core counts surpassing 128, FPGAs still use lookup tables (LUTs) as computing units, without the emergence of FPCA (Field Programmable Computing Array) using cores as computing units. For individuals, whether for job seeking or developing a technical side business, learning FPGA design remains a high-value knowledge path.