The Signal
Seven Japanese Groups Start Work on a Distributed Quantum Platform
Seven organizations in Japan have begun developing a NEDO-backed platform that links quantum computers, Tsurugi database clusters and optical networks. The project targets a scale of hundreds of thousands of qubits and plans to test applica
Quantum.world Team · Written with AI · fact-checked ·
A target measured in hundreds of thousands of qubits
Seven organizations in Japan began developing a distributed computing platform on September 29, with a goal of linking multiple quantum computers at a scale of hundreds of thousands of qubits. Waseda University announced the project on October 26, saying the platform will connect quantum-computer clusters to clusters running Tsurugi, a high-speed relational database, over an ultra-low-latency optical network.
The figure is a development target, not a report of an operating system at that scale. The announcement describes an effort to combine quantum processors, data processing and networking so their resources can work as one environment and connect with existing business systems.
Seven partners divide the technical work
The group comprises Nautilus Technologies, Quanmatic, GHELIA, 1Finity, Internet Initiative Japan (IIJ), Waseda and Nagoya University. Nautilus will lead the project and build the distributed data platform around Tsurugi; Quanmatic will develop distributed quantum algorithms and implement the quantum-computing system. GHELIA is responsible for system integration between quantum computers and the data platform.
1Finity, a Fujitsu company, will provide optical network interface cards and networking between resources within and across clusters. IIJ will procure, build and operate the server environment at its Shiroi Data Center Campus in Chiba Prefecture. Waseda will work on distributed quantum algorithms, while Nagoya University will develop distributed indexing technology.
Optical links connect computation to business data
The proposed architecture couples several technical layers. The optical network is intended to link geographically separated quantum computers and databases, while Tsurugi handles data extraction, conversion and updates needed by quantum workloads. A separate integration layer is meant to carry quantum-computing results into existing operational systems.
The partners frame the work around problems that involve many interacting choices, including electricity supply and demand, production planning and financial portfolios. Their first named application area is “Watt-Bit coordination”: allocating electricity, computing and communications resources in response to data such as renewable-power output, electricity prices, network use and GPU activity.
Energy allocation and AI workloads lead the use cases
The project also plans a local version of Watt-Bit coordination that would dynamically assign GPU resources among users and services. In other sectors, the group cites energy management for automobiles, portfolio optimization in finance, production planning in manufacturing, and workforce allocation or structural design in construction. Those are intended applications, not reported demonstrations.
Kazuhiro Nabeta, an executive vice president of Chubu Electric Power, said: “The practical application of distributed quantum computing will require not only advances in computing technology itself, but also the establishment of a comprehensive technological foundation encompassing data processing and system integration.”
The roadmap runs through 2030
The partners plan to develop the platform through fiscal 2028 and pursue use cases from fiscal 2029. For the wide-area energy-and-computing application, they aim to offer a large-scale platform as software as a service around 2030, with the stated goal of stabilizing AI inference costs and improving electricity use. Tokyo Electric Power Company Holdings, Chubu Electric Power and Kyushu Electric Power are participating in or considering participation in that initiative.
The announcement sets out the scale ambition and application areas but does not report a demonstrated qubit count, network-latency figure or benchmark showing an advantage over conventional optimization methods. Those measurements will determine how much of the proposed system can move from an integration plan into operational use.