Summary
- The consortium brings together data centre operators, power suppliers, telecommunications providers, and property developers.
- It will work on electricity, communications, development incentives, and local economic integration.
- Ishikari already hosts operating facilities and a separate 15MW project expected to launch during 2026.
NTT East has joined a group of data centre operators, utilities, telecommunications providers, and property businesses seeking to develop Ishikari City into a larger Japanese data centre cluster.
The Ishikari Data Center Consortium will focus on the power and communications infrastructure required by new facilities in the Ishikari Bay New Port area of Hokkaido. It also intends to work with local authorities on development incentives and ways for residents and local businesses to benefit from the cluster.
Members include Sakura Internet, Kyocera, Tokyu Land Corporation, Ishikari Renewable Energy Data Center No. 1, Flower Communications, Broadband Tower, NTT ME, Ishikari Regional Energy, Liene, and Hokkaido Integrated Communications Network, known as HotNet.
The location is being promoted around access to renewable energy and a comparatively low natural-disaster risk. Hokkaido’s cooler climate can also reduce the energy required for heat rejection during parts of the year, although individual facilities still require engineered cooling systems and protection against winter conditions.
Ishikari is not a greenfield concept without existing infrastructure. Sakura Internet has operated a data centre there since 2011 and has been expanding the site, including the installation of GPU systems.
Tokyu Land, Flower Communications, and Broadband Tower are separately involved in a 15MW renewable-powered facility expected to enter operation during 2026.
A cluster requires shared infrastructure
The consortium model acknowledges that individual developers cannot create a large regional market by building isolated facilities. Data centres need diverse fibre, substations, transmission capacity, roads, skilled labour, emergency services, and a planning system able to coordinate several projects.
Shared work on power and communications can reduce duplication and identify where public or regulated infrastructure must be strengthened before further campuses are approved.
The arrangement can also make renewable generation easier to use. A group of facilities with different load profiles and commissioning dates may support larger generation and network investments than a single project could justify.
Renewable availability does not remove the need for firm supply. Wind and solar output vary, while data centre loads generally operate continuously. Contracts, storage, grid balancing, and conventional generation remain part of the reliability equation.
Hokkaido’s location creates a further trade-off. Land, energy, and climate may be favourable, but the island is distant from Japan’s largest population and interconnection centres around Tokyo and Osaka.
AI training, batch computing, and some cloud workloads can tolerate that distance. Services requiring extremely low latency or dense proximity to customers and network exchanges may remain closer to the main metropolitan markets.
The cluster’s success will therefore depend on both physical infrastructure and workload selection. High-capacity fibre routes must connect Ishikari to customers and other cloud regions without creating a single point of failure.
Similar development strategies are visible in the Nordic data centre markets, where cooler climates and renewable generation have attracted large compute projects outside the traditional European hubs. Those markets show that low-carbon electricity is an advantage, but not a substitute for transmission certainty, connectivity, construction capability, and long-term customer commitments.
Local integration is another stated objective. Large data centres create construction activity and tax revenue but may employ relatively small permanent teams compared with their power and land requirements. Training, supplier development, heat use, telecommunications improvement, and access to computing services can broaden the local return.
The Ishikari consortium gives the participating organisations a forum for coordinating those questions before the cluster expands further. It does not itself commit a defined volume of new capacity or a construction timetable.
The practical measure will be whether the group can convert renewable resources and existing facilities into additional powered sites, diverse network connections, and projects with identified customers. Cluster branding is useful only when the underlying infrastructure can be delivered.
