NVIDIA’s Vera Rubin AI Factory Debuts in Japan as Jensen Huang Says AI Roadmap Is on Track
NVIDIA CEO Jensen Huang’s visit to Japan has brought a major announcement for the country’s artificial intelligence ambitions: Japan is set to receive its first Vera Rubin AI factory. The move reinforces NVIDIA’s confidence in its next-generation AI roadmap and signals that the company sees Japan as a key player in the future of AI infrastructure, robotics, and physical AI.
During his trip, Huang addressed recent speculation about potential delays affecting NVIDIA’s upcoming AI chips and rack-scale systems. His message was clear: the rumors are not accurate. According to Huang, the Vera Rubin platform is already moving through production, with “giant amounts” of output expected as NVIDIA ramps up its next wave of AI hardware.
The confirmation is important because Vera Rubin represents one of NVIDIA’s most anticipated platforms for future AI data centers. The system combines Vera CPUs with Rubin GPUs, designed to power large-scale AI workloads, advanced robotics, and next-generation industrial applications.
In recent weeks, market discussions have focused on whether NVIDIA’s upcoming rack solutions could face setbacks, including systems tied to the Oberon NVL144 and Kyber NVL576 platforms. Some speculation also centered on the design direction for Rubin Ultra chips, particularly whether NVIDIA would move forward with a larger multi-reticle design or adjust to a smaller configuration during the production ramp.
Huang’s latest comments suggest NVIDIA remains confident in its ability to execute. The company has repeatedly shown that it can address design and manufacturing challenges quickly by working closely with its extensive supply chain and ecosystem partners. That approach helped NVIDIA push forward with Blackwell and Blackwell Ultra, and the company appears to be taking the same aggressive path with Vera Rubin.
Japan’s first Vera Rubin AI factory will be developed as part of the FRONTia Project, an initiative focused on multimodal foundation models, AI robotics, and physical AI. NVIDIA is partnering with Noetra Corp. on the project, which is expected to become a major part of Japan’s AI infrastructure strategy.
The planned AI factory will include 13,750 Vera CPUs and 27,500 Rubin GPUs. It is expected to deliver 140MW of data center capacity and will be built on NVIDIA’s DSX platform. The facility will use NVIDIA Vera Rubin NVL72 racks and scale through Spectrum-X networking, giving Japan access to powerful infrastructure for training, deploying, and advancing AI systems.
Huang framed the announcement as a continuation of Japan’s long history of industrial leadership.
“Japan invented modern manufacturing. Now, it is building the AI factories that will power the next industrial revolution,” Huang said. He added that NVIDIA is honored to work with Japan and its industrial leaders to build the AI infrastructure that will support the country’s economy, industries, and next generation of innovation.
The project also highlights Japan’s importance in the development of physical AI. Japan has long been a global leader in robotics, automation, and humanoid machines, making it a natural market for NVIDIA’s growing AI robotics ecosystem. Several Japanese companies are already working with NVIDIA and are expected to use platforms such as Jetson Thor to accelerate robotics development and industrial AI applications.
Huang also shared a broader view of where he believes artificial intelligence is headed. While many technology cycles tend to mature within 10 to 15 years, he argued that AI is still in its early stages. In his view, the world is only at the beginning of the AI era, with agentic AI, robotics, and physical AI still far from reaching their full potential.
He said that, looking a decade ahead, the real question may be whether AI growth continues to accelerate or begins to flatten. However, Huang emphasized that he does not believe AI will decline. He also pushed back against the idea that the industry is already in an AI bubble, saying the world is still a long way from that point.
According to Huang, the buildout of AI infrastructure will continue for at least the next decade. That includes data centers, AI factories, networking systems, advanced chips, and software ecosystems capable of supporting national and enterprise-scale intelligence platforms.
He also stressed the importance of countries and companies controlling their own AI infrastructure. Open models, he said, make it possible for governments, enterprises, and researchers to inspect, improve, secure, adapt, and deploy AI according to their own needs. Huang described this as essential for ensuring that societies can participate in and benefit from the AI revolution.
Another topic raised during Huang’s Japan visit was the future role of Rapidus, Japan’s advanced semiconductor manufacturer. When asked whether Rapidus could become a future NVIDIA partner, Huang did not confirm any collaboration. However, he noted that rising AI demand will benefit the entire semiconductor industry and said NVIDIA looks forward to watching Rapidus make progress.
NVIDIA’s Japan announcement follows Huang’s broader push across Asia, where the company continues to deepen partnerships with governments, manufacturers, cloud providers, and robotics firms. With Vera Rubin entering production and Japan preparing to build its first AI factory around the platform, NVIDIA is positioning itself at the center of the next major phase of global AI infrastructure.
The message from Huang’s visit is clear: NVIDIA believes AI demand is not slowing down, Vera Rubin remains on schedule, and Japan is set to play a major role in the next industrial revolution powered by artificial intelligence.
