Japan bets on ‘physical AI’ with flurry of tech deals

Key Concepts

• Generative AI: AI capable of creating new content such as text, images, and videos.
• Physical AI: Applying AI into machines to enable them to think and act autonomously, particularly in industrial settings.
• Optical Devices: Components that manipulate light for data transmission, crucial for high-speed, low-power data systems.
• Supercomputer: A highly powerful computer designed for complex calculations and simulations.
• CPU (Central Processing Unit): The primary component of a computer that performs most of the processing.
• GPU (Graphics Processing Unit): Specialized electronic circuit designed to rapidly manipulate and alter memory to accelerate the creation of images in a frame buffer intended for output to a display device.
• AI Chip: A specialized semiconductor designed to accelerate artificial intelligence workloads.
• Robotics Business: The segment of a company focused on the design, manufacturing, and application of robots.
• Industrial Software: Software designed for use in industrial settings, such as manufacturing and infrastructure management.
• Foundation Models: Large AI models trained on vast amounts of data, capable of performing a wide range of tasks.
• Applied AI: The practical application of AI technologies to solve specific problems in various industries.
• Precision Engineering: A field of engineering that focuses on the design and manufacture of highly accurate and precise components and systems.
• Factory Automation: The use of technology to control and monitor industrial machinery and processes, reducing human intervention.
• Data Centers: Facilities that house computer systems and associated components, such as telecommunications and storage systems.
• Telco Networks: Telecommunications networks.

Japanese Companies' Surge in AI: Recent Deals and Strategic Positioning

This episode of Tech Latest from Nikkei Asia delves into the recent surge of activity by Japanese companies in the artificial intelligence (AI) sector, marked by a series of significant deals within a short period. The discussion features Shoaratani and Tbasa Suruga, a new Japan tech reporter.

Recent AI Sector Deals in Japan

The past week has seen a flurry of significant announcements in Japan's AI landscape:

• Hitachi and OpenAI Partnership:

  • Details: Hitachi, a major Japanese conglomerate, has signed an agreement with OpenAI.
  • Objective: To integrate generative AI into Hitachi's industrial software and infrastructure. This aims to leverage OpenAI's cutting-edge technology for practical applications in factories and industrial settings.

• Fujitsu and Nvidia Collaboration:

  • Context: This partnership builds upon a previous announcement in August where Nvidia joined a Fujitsu-led supercomputer project.
  • Objective: To develop an energy-efficient AI chip by 2030. This involves connecting Fujitsu's CPU chips with Nvidia's GPUs.
  • Significance: Nvidia CEO Jensen Huang was present in Tokyo for the announcement, highlighting the importance of this collaboration. Fujitsu executives noted Nvidia's appreciation for Japan's expertise in energy-efficient technologies, developed through projects like Fugaku.

• NTT and Broadcom Deal (October 7th):

  • Details: Japan's telecommunications giant, NTT, has signed a deal with US chipmaker Broadcom.
  • Objective: To produce a key optical device for NTT's "IOWN" (Innovative Optical and Wireless Network) initiative. IOWN is a next-generation, ultra-fast, low-power data system that uses light instead of electricity for data transmission.
  • Application: The deal focuses on developing a new switch for servers, aiming to enhance speed and energy efficiency in response to soaring AI workloads. This technology is expected to be mass-produced from next year, capitalizing on the global data center boom.

• SoftBank's Acquisition of ABB Robotics (October 8th):

  • Details: SoftBank announced the acquisition of the robotics business of Swiss-headquartered ABB for approximately $5.4 billion.
  • Significance: This is considered the largest deal among those discussed and a major step in SoftBank founder Masayoshi Son's vision of "physical AI."

Understanding "Physical AI"

• Definition: Physical AI is distinct from the digital AI focused on text, image, and video generation. It involves applying AI into machines to enable them to think and act autonomously.
• Applications: A key use case is in factories, where robotic arms can move beyond repetitive tasks to perform actions more quickly and articulately based on AI-driven decision-making.

Factors Driving Simultaneous AI Activity

The convergence of these deals within a short timeframe is attributed to a combination of factors:

• Business Strategy and Timing: Companies are strategically aligning their development and commercialization efforts.
• Japanese Government Policy: Government backing and investment in AI infrastructure are playing a crucial role.
• Geopolitics: The growing tech rivalry between the US and China has made Japan an attractive hub for global tech firms seeking partnerships and a more flexible regulatory environment. Japan is actively positioning itself as an AI hub by easing regulations and investing heavily.
• Commercialization of Existing Projects: Projects like NTT's IOWN, initiated in 2019, are now reaching commercialization stages.
• Japan's Supercomputing Legacy: Fujitsu's partnership with Nvidia builds on Japan's long-standing efforts in supercomputing, exemplified by the Fugaku supercomputer.

Most Significant Deal: SoftBank's Acquisition of ABB Robotics

While all deals are noteworthy, SoftBank's acquisition of ABB Robotics stands out due to:

• Masayoshi Son's Vision: It represents a significant statement about the future of AI extending beyond digital applications to physical machines.
• ABB's Expertise: ABB is a leading industrial robot maker known for its data-driven factory systems.
• SoftBank's Strategic Integration: SoftBank aims to integrate this robotics business with its existing investments in semiconductors, data centers, and power, positioning itself for the "artificial super intelligence era." Son's ambition is to apply new AI technology across various industries, not just for content generation.

Japan's Position in the Global AI Ecosystem

• Strengths: Japan is unlikely to compete directly with the US and China in building massive foundation models. However, it can excel in applied and industrial AI, particularly in robotics, manufacturing, and semiconductors, leveraging its strengths in precision engineering, factory automation, and hardware.
• Collaboration: Japan has opportunities to collaborate with US and other Western companies.
• Challenges: The Japanese AI ecosystem currently lacks the presence of many young, fast-moving, high-growth companies driving innovation, unlike in the US. The dominance of large players like Hitachi, Fujitsu, and NTT, while beneficial for infrastructure, can sometimes hinder rapid movement and risk-taking.
• Future Outlook: The success of these recent partnerships will depend on their ability to translate collaborations into scalable, real-world products that solve business problems, moving beyond basic text generation or search functionalities. The focus is on how Gen AI and data center investments will be practically applied on the ground.

The episode concludes by directing listeners to Nikkei Asia's website for further reading and to subscribe to their weekly Tech Asia newsletter.