In Japan, a remarkable partnership is taking shape between RIKEN, the nation’s leading research institute, and NVIDIA, a global pioneer in computing technology. Together, they are building two next-generation supercomputers—systems that promise to transform how we understand the world and solve its most challenging problems.
A Dual Approach: AI and Quantum Computing
At the heart of this collaboration are two powerful supercomputers, expected to launch by spring 2026. Each is tailored for a specific purpose. One will be dedicated to artificial intelligence (AI) research, using the latest advancements in machine learning and data processing. The other will focus on quantum computing, a new frontier with the potential to revolutionize science and technology.
Together, these systems will feature 2,140 of NVIDIA’s cutting-edge Blackwell graphics processing units (GPUs). Of these, 1,600 GPUs will power the AI supercomputer, while 540 GPUs will specialize in quantum research. This combination ensures flexibility, strength, and speed—qualities essential for addressing problems once thought impossible to solve.
Technology at the Forefront
Both supercomputers will rely on NVIDIA’s GB200 NVL4 platform. This is a sophisticated, liquid-cooled system built to handle the massive demands of high-performance computing and AI workloads. For the quantum-focused supercomputer, NVIDIA’s advanced Quantum-X800 InfiniBand networking will allow researchers to process and share information at extraordinary speeds.
These technological advances are not simply about faster calculation. They provide the tools scientists need to explore life sciences, discover new materials, improve weather forecasting, and revolutionize factory automation. The possibilities stretch across fields, inviting new discoveries that could shape our future.
A Unified Platform for the Future
What makes this partnership truly special is its vision—a unified platform where AI, quantum computing, and high-performance computing come together. This marks a significant step in Japan’s strategy to become a leader in artificial intelligence, all while strengthening the nation’s independence in technology.
Central to this is the deep integration of NVIDIA’s CUDA-X software stack, which brings hundreds of specialized libraries and tools to researchers’ fingertips. This software foundation accelerates discovery in chemistry, physics, medicine, and beyond, making the supercomputers not just powerful, but also broadly accessible to scientists.
Laying the Groundwork for the Next Era
The significance of this project extends beyond its immediate impact. The experience and technologies developed here will shape future projects, such as the planned FugakuNEXT supercomputer, scheduled for 2030. By building on these advancements, Japan aims to achieve breakthroughs that will redefine computing as we know it.
In this way, the RIKEN-NVIDIA initiative is more than an investment in machines—it is a testament to Japan’s enduring commitment to scientific excellence. By merging AI and quantum computing, these supercomputers will support discoveries that touch every aspect of modern life, from health and industry to climate science and education.
A Future Shaped by Innovation
Every era in science is marked by a leap forward. For Japan, these new supercomputers represent such a leap—a moment when ambition, technology, and vision come together to chart bold new directions. As these systems go online, they will empower researchers, unlock new knowledge, and spark innovations that move humanity forward. The journey begins now, opening doors to discoveries we have only begun to imagine.
Leave a Reply