In a move that could redefine the future of computation, Japan’s RIKEN and IBM have announced a groundbreaking collaboration to integrate quantum systems directly into FugakuNEXT, the successor to Japan’s world-leading Fugaku supercomputer. This isn’t just an upgrade — it’s a complete rethinking of what computing can achieve. IBM and Japan’s RIKEN are teaming up to integrate quantum systems into FugakuNEXT — the zetta-scale AI quantum HPC platform. Learn how this partnership could redefine global computing, from climate modeling to quantum-AI research, by 2026.
The collaboration marks a crucial step in Japan’s national quantum strategy, which aims to merge the power of classical high-performance computing (HPC) with quantum technologies. The objective: achieve quantum-accelerated artificial intelligence (AI) and climate modelling breakthroughs by 2026.
By late 2025, the first hybrid workflows between FugakuNEXT and IBM Quantum systems are expected to go live — setting the stage for one of the world’s first operational quantum-HPC simulation environments.
From Fugaku to FugakuNEXT: The Evolution of Japan’s Supercomputing Legacy
The original Fugaku supercomputer, jointly developed by RIKEN and Fujitsu, made global headlines when it became the world’s fastest system in 2020. At its peak, it achieved a staggering 442 petaflops of performance and was used in critical applications ranging from COVID-19 drug discovery to disaster simulation and AI language modeling.
But the next chapter — FugakuNEXT — is designed to go beyond speed. It’s envisioned as a “Zetta-scale AI-HPC platform”, capable of integrating quantum processors, AI accelerators, and next-generation semiconductors into a single ecosystem.
FugakuNEXT represents a paradigm shift — moving from raw computational power to hybrid intelligence, where classical, AI, and quantum systems collaborate dynamically depending on the problem type.
IBM’s Quantum HPC Role: A Marriage of Two Titans
IBM, a long-standing partner in Japan’s computing ecosystem, is bringing its Quantum System One and future Heron-class processors into the FugakuNEXT environment.
The integration will allow RIKEN scientists to offload certain computational workloads — such as molecular simulations, optimization problems, and climate projections — onto quantum systems through hybrid orchestration layers.
IBM’s Qiskit Runtime and quantum serverless infrastructure will be key to managing this interaction. Essentially, quantum processors will act as “co-processors” to FugakuNEXT’s exascale CPUs and GPUs, enhancing calculations that benefit from quantum parallelism and entanglement-based correlations.
This design aligns with IBM’s broader Quantum System Roadmap, which envisions modular, scalable architectures that can interconnect classical supercomputers with quantum devices over cloud-based networks.
Why Quantum HPC Integration Matters
Traditional HPC systems are reaching physical and economic limits. As transistor scaling slows and energy consumption soars, adding more nodes or GPUs isn’t always feasible. Many real-world problems — especially in chemistry, materials science, and optimization — exhibit combinatorial complexity that classical systems struggle to handle efficiently.
Quantum computing, meanwhile, offers a fundamentally different approach. By representing information in quantum states (qubits), it can explore multiple solutions simultaneously. When combined with HPC, quantum can become an accelerator — a catalyst for exponential efficiency.
The hybrid FugakuNEXT-IBM environment aims to bring three key advantages:
- Precision Simulation: Quantum systems can simulate atomic-scale phenomena that are prohibitively expensive for classical machines.
- Faster AI Training: Quantum kernels may speed up certain machine learning workflows, especially in high-dimensional feature spaces.
- Climate and Energy Insights: The system could help simulate materials for carbon capture or optimize global energy grids — applications directly tied to Japan’s green innovation goals.
By 2026, Japan hopes to demonstrate quantum-accelerated AI models that outperform purely classical equivalents — a milestone that could redefine how national labs and enterprises approach computing infrastructure.
Japan’s 2026 Vision: Quantum for National Competitiveness
This collaboration isn’t an isolated experiment. It’s part of Japan’s national roadmap for quantum innovation, guided by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) and the Cabinet Office’s Quantum Technology Innovation Strategy.
The plan targets three major milestones by 2026:
- Establish quantum-classical integration testbeds (of which FugakuNEXT is the flagship).
- Achieve quantum-accelerated simulation capabilities for key industries like pharmaceuticals, energy, and materials.
- Train over 10,000 quantum-literate professionals across research and industry.
RIKEN’s collaboration with IBM positions Japan at the forefront of this global race — alongside similar efforts by the U.S. Department of Energy, European HPC consortiums, and China’s National Quantum Computing Lab.
As RIKEN President Hiroshi Matsumoto noted in a recent statement:
“FugakuNEXT represents not just faster computing, but smarter collaboration between paradigms. With IBM, we are building the foundation of Japan’s next-generation innovation ecosystem.”
Technical Outlook: What the Hybrid Workflows Might Look Like
The integrated FugakuNEXT–IBM system will operate through quantum-HPC hybrid simulation environments. These environments use middleware orchestration to decide which parts of a problem are best handled by quantum processors and which remain on classical nodes.
For example:
- A quantum module might simulate electron-level interactions in a molecule.
- A classical HPC cluster could handle macro-scale thermodynamic modeling.
- An AI layer might interpret the data, train models, and adjust inputs dynamically.
This orchestration will be managed through containerized quantum services, running on IBM’s cloud or potentially co-located infrastructure. It’s a distributed but deeply integrated workflow — enabling seamless cooperation between distinct computing architectures.
The resulting ecosystem will not only serve Japanese research labs but could eventually support open-access frameworks for universities and enterprises worldwide, much like how Fugaku currently provides shared computational resources.
Global Implications: The Rise of Quantum-Enhanced Supercomputing
FugakuNEXT’s quantum integration signals a global inflection point. Across Europe, North America, and Asia, nations are racing to build hybrid quantum-HPC systems as part of their strategic computing roadmaps.
In Europe, the EuroHPC JU program has already announced plans for a quantum-integrated supercomputer by 2025. The U.S. Department of Energy’s Argonne National Lab is testing similar hybrid workflows with quantum simulators and AI accelerators.
What makes RIKEN’s effort distinct is scale and direction — FugakuNEXT isn’t just adding quantum as an experiment; it’s embedding it into the core architecture of Japan’s zetta-scale platform. This systemic integration could create one of the world’s first true hybrid infrastructures, where quantum computing becomes part of national-scale research operations, not a niche add-on.
The AI Connection: Quantum Meets Machine Learning
Another major goal of the IBM–RIKEN collaboration is AI acceleration. Training modern AI models requires immense computing resources. GPT-style language models, for instance, consume petaflops of compute power and terawatt-hours of energy.
Quantum computing could offer shortcuts in AI model training and feature extraction. Quantum kernels can process data in higher-dimensional spaces more efficiently, potentially improving learning rates and reducing compute costs.
By integrating these capabilities within FugakuNEXT, Japan aims to pioneer quantum-accelerated AI research — developing new algorithms that blend quantum physics with neural architectures. This could have profound implications for everything from drug discovery to robotics.
Environmental and Economic Significance
Another critical aspect of the FugakuNEXT vision is sustainability. Traditional supercomputing consumes massive energy — Fugaku itself draws around 30 megawatts, enough to power a small town.
Quantum systems, when scaled and stabilised, promise dramatically higher energy efficiency per computation. IBM’s roadmap emphasizes modular cryogenic systems designed for sustainability and scalability.
From an economic perspective, integrating quantum into HPC could extend the life and ROI of massive infrastructure investments. Instead of replacing supercomputers every few years, institutions can augment existing systems with quantum nodes, achieving exponential improvement without exponential cost.
Conclusion:
The IBM–RIKEN collaboration for FugakuNEXT symbolizes the next great transition in computing — from siloed architectures to integrated intelligence systems. By merging quantum processors with AI-HPC infrastructure, Japan is not just future-proofing its technology stack; it’s shaping a new global model for research, innovation, and competitiveness.
As Mattias Knutsson, Strategic Leader in Global Procurement and Business Development, aptly observes:
“Breakthroughs don’t come from choosing between technologies — they come from integrating them. The future belongs to those who connect the dots early.”
By 2026, when the first quantum HPC hybrid workflows go live on FugakuNEXT, we may look back at this moment as the dawn of a new computational era — one where the boundaries between physics, data, and intelligence finally begin to blur.
