Fugaku NEXT: Japan–US Alliance with NVIDIA and Fujitsu Targets World’s Fastest Supercomputer

BUSINESS

Supervisor: Naohiro Takemura

Started his career based on his experience at an accounting firm.
For approximately 30 years, he has been engaged in database marketing, finance, entrepreneurship, BPO operations, and new business planning.
He is well-versed in fundraising and financial management, and currently manages his own company while also undertaking outsourced work in management and new business development.

The development of Fugaku NEXT, the successor to Japan’s world-class supercomputer “Fugaku,” has officially begun. With Fujitsu leading the system design and NVIDIA contributing cutting-edge GPU technology, this project brings together Japan–US innovation to achieve unmatched computing speed and energy efficiency. Fugaku NEXT is not only a national project, but a symbol of global technological cooperation.

What Is Fugaku NEXT? A Future-Defining National Infrastructure

Fugaku NEXT is positioned as the successor to Japan’s current flagship supercomputer, Fugaku, which achieved global top performance rankings in 2020. The new machine will not only surpass its predecessor in terms of raw speed but will also offer greater versatility, efficiency, and adaptability across a wide range of computational fields.

Here is a summary of Fugaku NEXT’s core development principles:

PrincipleObjective
High PerformanceDramatically enhanced computing power
Energy EfficiencyReduced power consumption for sustainable operation
VersatilityApplicable across diverse scientific and industrial domains

RIKEN is taking a holistic approach, placing emphasis not only on speed, but also on power efficiency, system flexibility, and real-world applicability. The project is notable for incorporating international cooperation, marking a significant shift from previous Japan-only efforts.

What NVIDIA Brings to the Table: Revolutionary AI Processing

NVIDIA, the global leader in GPU-based computing, is contributing to the Fugaku NEXT project by co-developing the system’s AI processing components. With its state-of-the-art GPU architecture, NVIDIA’s involvement is expected to significantly enhance the system’s ability to manage massive datasets and complex machine learning algorithms with speed and efficiency.

FeatureContribution
GPU PowerIndustry-leading capabilities in AI and parallel computing
Massive ParallelismSupports complex workloads simultaneously
Energy EfficiencyMaximizes performance with reduced power draw

This marks the first time a foreign company has played a core role in developing Japan’s national supercomputing infrastructure. NVIDIA is not merely a technology vendor, but a true design partner working alongside RIKEN and Fujitsu from the early stages.

Fujitsu’s Role: Core System Design and CPU Engineering

Fujitsu has a proven track record in supercomputing, having developed the A64FX processor that powers the current Fugaku. In the Fugaku NEXT project, the company is once again leading the development of the central processing unit (CPU), which serves as the “brain” of the system—executing instructions, controlling workflows, and managing resources.

Here is a breakdown of Fujitsu’s design priorities:

Design FocusDetails
High IntegrationMore cores for higher computational throughput
Thermal OptimizationAdvanced cooling solutions for stability
ScalabilityArchitecture designed for future expansion and flexibility

Fujitsu’s design philosophy emphasizes not just peak performance, but also operational stability and energy-conscious innovation, qualities that set Japanese engineering apart in the global market.

Real-World Applications and Societal Benefits

The potential applications of Fugaku NEXT span every sector—from academic research to industrial operations. This supercomputer is designed to handle tasks that were previously considered too computationally expensive or time-consuming.

FieldApplications
Artificial IntelligenceLarge-scale model training and inference
Medical ResearchMolecular dynamics, personalized medicine
Climate ScienceHigh-resolution forecasting, disaster prediction
EnergyNuclear fusion simulation, renewable energy optimization
Materials ScienceNano-level structural analysis, new material discovery

In essence, Fugaku NEXT is not just a technological breakthrough—it is a solution platform for global challenges.

Striving for World-Class Performance and Sustainability

Fugaku NEXT aims to lead international performance benchmarks such as the TOP500 and Green500. The system is designed to be among the best in terms of both raw computational power and energy efficiency.

A key metric will be performance-per-watt, where the combination of NVIDIA’s GPUs and Fujitsu’s CPU architecture promises unmatched efficiency. This aligns with a broader shift in global computing—from focusing solely on speed to achieving sustainable, scalable performance.

Driving Education, Industry, and Human Capital Development

The impact of Fugaku NEXT will extend far beyond research labs. In education, the system will help foster scientific curiosity, boost programming literacy, and support simulation-based learning. In industry, it will support next-generation manufacturing technologies like digital twins and smart factories.

Moreover, the project will play a pivotal role in cultivating Japan’s next generation of technical experts. By exposing engineers, researchers, and students to cutting-edge systems, it will create ripple effects throughout the Japanese innovation ecosystem.

Conclusion

Fugaku NEXT represents a new era in international collaboration and supercomputing excellence. With RIKEN’s strategic oversight, Fujitsu’s intricate design skills, and NVIDIA’s state-of-the-art AI processing capabilities, the project is well-positioned to deliver unprecedented computational power with real-world impact.

More than just a machine, Fugaku NEXT will serve as a backbone for future society—advancing science, protecting lives, and enabling innovations that will define the decades to come.