Author: Wall Street View
NVIDIA is positioning Japan as a central pillar of its global physical AI strategy.
NVIDIA CEO Jensen Huang appeared in Tokyo this week, engaging in a series of partnership signings and business meetings. It was reported that on Thursday, NVIDIA announced a collaboration with Japanese robotics giants Fanuc and Yaskawa Electric to jointly advance the development of robotics and AI technologies.
Meanwhile, NVIDIA also announced an expansion of its collaboration with Toyota, covering areas such as autonomous driving, factory simulation, and urban intelligence. At a media event in Tokyo, Jensen Huang said, "With AI, robots will become intelligent, adaptable, and accessible."
The strategic intent of this trip to Japan is clear: NVIDIA is systematically integrating Japan’s manufacturing base and semiconductor supply chain with its full-stack AI technology.
Jensen Huang dismissed concerns about an AI investment bubble in an interview, stating, "We are still far from an AI bubble; demand is extremely strong," and emphasized that "we need to build infrastructure for at least a decade." He also indicated that NVIDIA will announce collaboration plans related to Japan’s sovereign AI later this week.
Additionally, noteworthy is that this visit also featured a historically significant reunion—Jensen Huang reunited on stage in Akihabara with Shoichiro Irimajiri, former president of Sega, who came to NVIDIA’s aid when the company was on the brink of bankruptcy 30 years ago, adding emotional depth to this business trip.
(Video screenshot: Jensen Huang embracing the former president of Sega)
Reunion in Akihabara: A Lifesaving Favor from 30 Years Ago
Beyond business meetings, the most emotionally resonant moment of Jensen Huang’s trip occurred at the former site of the Sega arcade in Akihabara, Tokyo.
According to Sega, Huang Renxun attended an event hosted by Sega on July 15 local time, appearing on stage alongside Shoichiro Irimajiri, former president of Sega, after many years. At the event, Huang Renxun remarked emotionally: "Without everything Sega has done, and without everything Shoichiro Irimajiri has done, NVIDIA would not have survived to this day."
This history dates back to around 1996. At the time, NVIDIA, which had just been founded, suffered a complete project failure due to a misjudged technical approach while developing a graphics chip for Sega’s next-generation console, bringing the company to the brink of bankruptcy.
Huang Renxun openly admitted failure to Shoichi Iriya, then vice president of Sega, and instead of holding him accountable, Iriya pushed Sega to invest approximately $50 million in the struggling startup. Huang Renxun once recalled:
I told Uchikoshi that if they invested this money in us, they would likely lose everything; but if they didn’t invest, we would go bankrupt... After considering it for a few days, he told me: We’re in.
With this funding, NVIDIA restarted after laying off 60% of its staff, launching the RIVA 128 in 1997 and subsequently establishing its market position in the GPU industry with products such as the RIVA TNT and GeForce 256.
Ironically, after NVIDIA went public in 1999 with a valuation of approximately $300 million, Sega quickly chose to cash out, receiving about $15 million from the sale of its shares. Today, NVIDIA’s market capitalization exceeds $5 trillion.
During this reunion, both parties announced the continuation of their partnership—Sega’s future titles will support NVIDIA’s newly launched RTX Spark platform, including the upcoming VIRTUA FIGHTER CROSSROADS. NVIDIA and Sega’s collaboration began 30 years ago, when NVIDIA’s NV1 chip provided graphics support for the PC version of the original Virtua Fighter, one of the world’s first 3D fighting games.
(Image source: NVIDIA official website)
Izakaya dinner: Cultivating the core of Japan’s semiconductor supply chain
The night before the official partnership announcement, Huang Renxun completed a quiet yet meaningful dinner to collectively engage with key nodes in Japan’s AI supply chain.
According to Singapore's Straits Times, on the evening of July 15 local time, Jensen Huang appeared at a izakaya in Tokyo's Kanda district, dining with senior executives from key Japanese supply chain companies for approximately two hours.
(Source: The Straits Times, Singapore)
Attendees include:
The CEO of Kioxia, a leading manufacturer of advanced flash memory chips; the head of Shin-Etsu Chemical, a global leader in silicon wafers; the leader of Tokyo Electron, a chip equipment provider; executives from Ajinomoto, the exclusive supplier of advanced chip packaging films; and the heads of Sumitomo Electric Industries, a fiber optic cable manufacturer, and Taiyo Yuden, a producer of advanced capacitors. Yuki Kusumi of Panasonic Holdings was also present.
It is reported that Huang Renxun and his party enjoyed grilled skewers and beef offal hot pot, and drank Japanese whisky.
The lineup of this gathering nearly captures the full picture of Japan’s hardware supply chain upon which NVIDIA’s next-generation AI system relies. According to attendees, conversations arose such as “Let’s work together to promote the prosperity of industries like semiconductors and drive stock prices higher.”
Outside the izakaya, a crowd of people holding smartphones gathered, hoping to catch a glimpse of this AI-era idol known on social media as "Kawajan-san." A 57-year-old tourist from Taiwan, Chang Hui-Yu, said outside the Sega event, "I think he is the most influential person on Earth."
Betting on Physical AI: Japan’s manufacturing sector is a "natural ally"
The core strategic narrative of Huang Renxun's visit is to position Japan as a key hub for global physical AI development.
According to Japan's Kyodo News, Huang Renxun said in an interview in Tokyo, "This is a historic moment for Japan, as Japan has traditionally excelled in precision manufacturing and large-scale production."
He believes that AI can help Japan address the severe labor shortage, "by automating tasks and leveraging AI and robotics to enhance the capabilities of the existing workforce and boost the nation's overall productivity."
At the specific collaboration level, NVIDIA’s partnerships with FANUC and Yaskawa Electric directly target the intelligent upgrading of industrial robots. Its collaboration with Toyota is more comprehensive:
Toyota is developing next-generation vehicles with L2++ capabilities based on NVIDIA's DRIVE AGX platform and DriveOS operating system; utilizing NVIDIA Megatron-LM to train an AI code assistant compliant with MISRA standards to accelerate vehicle software engineering; and leveraging NVIDIA Omniverse and Isaac Sim frameworks to advance digital twin and robotics simulation in manufacturing.
Woven by Toyota, a subsidiary of Toyota, has also developed a multimodal vision-language model for urban mobility intelligence based on NVIDIA H100 GPUs.
NVIDIA Vice President Rishi Dhall said, "Physical AI will bring intelligence to every mobile machine, from cars, robots, and trucks to the cities and factories in which they operate."
Full-stack layout: from healthcare and finance to quantum computing
Beyond robots and automobiles, NVIDIA's partnership footprint in Japan extends across multiple key industries, demonstrating a systematic, end-to-end penetration.
In healthcare and life sciences, leading Japanese pharmaceutical companies including Eisai, Astellas, Daiichi Sankyo, and Ono Pharmaceuticals are accelerating AI-driven drug discovery using NVIDIA’s BioNeMo platform. Canon has launched Japan’s first NVIDIA-accelerated photon-counting CT system, while Fujifilm has commercialized Japan’s first full-body CT system powered by NVIDIA Blackwell. Kawasaki Heavy Industries plans to develop surgical assistance robots and hospital transport robots using NVIDIA’s Holoscan IGX, Isaac GR00T, and Cosmos platforms. In finance, Mizuho Bank intends to build Japan’s largest on-premises AI factory, starting with NVIDIA DGX B200 systems; JRI, part of the Sumitomo Mitsui Financial Group, has already deployed an AI factory leveraging NVIDIA Nemotron open models to transform financial data into intelligence; and Rakuten Bank will use NVIDIA’s Agent Toolkit to develop transactional foundation models. In quantum computing, two supercomputers powered by NVIDIA GB200 have begun operations at RIKEN: RIKYU is equipped with 1,600 NVIDIA Blackwell GPUs to support open foundation model development, while the quantum-HPC system ROQUO integrates 540 Blackwell GPUs tightly connected to quantum computers within the RIKEN campus. Mitsubishi Chemical, Mizuho Bank, Keio University, AIST, and other institutions have partnered with NVIDIA to achieve a 13.4x acceleration in molecular spectroscopy workflows compared to pure CPU nodes.
Additionally, it has been reported that market speculation suggests NVIDIA may announce a partnership with Japan’s “national team” for physical AI models, Noetra. Noetra, led by SoftBank, brings together 44 Japanese companies including Honda and NEC, and has received a ¥1 trillion government subsidy.