Seven organizations in Japan have begun developing a computing platform designed to connect quantum computers with high-speed databases over optical networks to optimize AI infrastructure and power resources.

Announced October 26, the project brings together Fujitsu's networking subsidiary 1Finity with Nautilus Technologies, Quanmatic, Ghelia, Internet Initiative Japan (IIJ), Waseda University, and Nagoya University. Work began September 29 with support from Japan’s New Energy and Industrial Technology Development Organization (NEDO).

The partners plan to complete development by fiscal 2028 and pursue applications from fiscal 2029. The announcement says a large-scale software-as-a-service offering for allocating power, computing, and communications resources is envisaged around 2030.

Connecting quantum computing with business data

The proposed architecture connects a quantum computing cluster with a classical database cluster running the open-source Tsurugi relational database over an ultra-low-latency optical network.

Tsurugi would handle the extraction, transformation, and updating of data needed for quantum calculations. An integration layer would feed results back into existing business systems.

1Finity will provide optical network interface cards and networking technology within and between clusters. The partners intend to support real-time data exchange between geographically separated data centers.

IIJ will procure, build, and operate the project’s server environment at its Shiroi Data Center Campus in Chiba Prefecture.

Nautilus will lead the project and develop the distributed data infrastructure. Quanmatic will develop distributed quantum algorithms and implement the quantum computing system, while Ghelia will connect the computing and data platforms. Waseda will research distributed quantum algorithms, and Nagoya will develop distributed indexing technology.

The partners describe a goal of expanding computational scale to hundreds of thousands of qubits through distributed quantum computing. That is a development objective; the announcement does not report a system operating at that scale or specify the quantum hardware and inter-processor connections.

Allocating GPUs alongside power and network capacity

One planned application would combine renewable energy generation data, electricity prices, network utilization, and GPU operating data to inform resource allocation across multiple locations.

The partners call this “watt-bit collaboration,” linking decisions about electricity and digital infrastructure. The aim is to improve power utilization and stabilize AI inference costs.

A second application would dynamically allocate GPU capacity among users and services, selecting resources according to demand. The partners intend to connect that local allocation capability with the wider infrastructure optimization platform.

Tokyo Electric Power, Chubu Electric Power, and Kyushu Electric Power are examining the broader concept, according to the release. Tohoku Electric Power also contributed a statement describing potential applications to its GPU cloud and data center services.

The project joins other efforts to integrate quantum computing with existing infrastructure. Earlier this month, NTT won support for international discussions on quantum network architecture and interoperability, including connections with AI infrastructure and data centers.

Beyond AI infrastructure, the seven partners plan to explore optimization applications in automotive energy management, financial portfolios, manufacturing schedules, and construction.