In a landmark collaboration, NVIDIA, Fujitsu, and RIKEN are launching groundbreaking supercomputers, including a hybrid quantum-GPU platform. This move is set to redefine Japan’s technological landscape, accelerate its “AI for Science” goals, and bolster its semiconductor self-sufficiency.
TOKYO, JAPAN – Japan is making a seismic leap in the global technology race, launching a new generation of supercomputing infrastructure powered by strategic partnerships with U.S. tech giant NVIDIA. The initiatives, led by industry leader Fujitsu and the renowned RIKEN research institute, are set to create one of the world’s most advanced platforms for artificial intelligence (AI), high-performance computing (HPC), and quantum computing.
At the center of this transformation are two flagship systems: the ABCI-Q supercomputer and the successor to the world-famous Fugaku, tentatively known as FugakuNEXT. These projects signal Japan’s determined push for AI sovereignty and a resilient domestic semiconductor ecosystem.
A New Era of Hybrid Computing: ABCI-Q and FugakuNEXT
The collaboration between NVIDIA and Fujitsu is pioneering a new frontier of hybrid computing that seamlessly blends classical and quantum processing power.
-
ABCI-Q: A Quantum-AI Hybrid: Set to be a revolutionary system, ABCI-Q will be powered by over 2,000 of NVIDIA’s cutting-edge H100 Tensor Core GPUs. What makes it unique is its integration with multiple quantum processors, including Fujitsu’s 256-qubit superconducting system. Orchestrated by NVIDIA’s CUDA-Q software platform, ABCI-Q will tackle complex problems in materials science, drug discovery, and financial modeling that are beyond the reach of classical computers alone.
-
FugakuNEXT: Power and Efficiency Redefined: Building on the legacy of the Fugaku supercomputer, FugakuNEXT will feature Fujitsu’s next-generation MONAKA-X CPU. This domestically developed chip, built on advanced 2-nanometer technology, is optimized for extreme energy efficiency and includes specialized AI enhancements. Paired with NVIDIA’s powerful GPU accelerators, FugakuNEXT is poised to deliver unprecedented performance for large-scale scientific simulations, from climate modeling to pandemic preparedness.
“These projects are not just about building faster machines; they are about creating a resilient, scalable infrastructure to meet Japan’s ‘AI for Science’ goals,” said an industry expert familiar with the national strategy.
Bolstering National Ambition and Semiconductor Sovereignty
This technological push is deeply intertwined with Japan’s national strategy to achieve semiconductor self-sufficiency. The government-backed Rapidus initiative, which aims to mass-produce 2nm chips by 2027, is a direct response to global supply chain vulnerabilities.
While Rapidus focuses on logic chips, Fujitsu’s development of the MONAKA-X CPU and NVIDIA’s advanced interconnect technologies, like the upcoming NVLink Fusion, create a powerful and complementary domestic ecosystem. This synergy reduces Japan’s reliance on foreign suppliers and accelerates its journey toward becoming a self-sufficient AI leader.
NVIDIA’s role is critical. Its GPUs and networking hardware form the backbone of Japan’s new AI infrastructure, while its CUDA-Q software ensures that the nation’s quantum research aligns with global standards, attracting international collaboration.
Why This Matters: Fueling a New Wave of Innovation
Japan’s massive investment—including $740 million in AI cloud infrastructure and substantial funding for sovereign AI initiatives—underscores a global trend: advanced computing infrastructure is now a strategic national asset.
The insatiable demand for computational power is evident in projects like RIKEN’s RAIDEN supercomputer, which was upgraded using NVIDIA V100 GPUs to achieve 54 petaflops for AI research. As AI models grow more complex, the need for next-generation hardware like NVIDIA’s upcoming Blackwell platform will only intensify.
Furthermore, Japan’s focus on “AI for Science” will create a virtuous cycle of innovation. By using AI, digital twins (powered by platforms like NVIDIA Omniverse), and hybrid simulations, researchers can accelerate discoveries in physics, chemistry, and biology, further fueling demand for specialized, high-performance semiconductors.
This strategic partnership between Japanese ingenuity and NVIDIA’s ecosystem represents a win-win, creating a blueprint for how nations can leverage cross-border collaboration to achieve technological leadership in the AI era.
