AI data centers, from development and operations to computing capacity sales.

With KAMUI immersion cooling racks at the core, we support AI infrastructure from micro data centers to large-scale facilities. We support both the infrastructure and the business.

Build the infrastructure. Plan for the operations that follow.

Every customer has different sites, GPUs, and uses for computing resources. Quantum Mesh designs configurations around your use case and scale, bringing post-opening monitoring and maintenance, computing capacity sales, and revenue management into one plan.

From 1 rack

Configurations suited to your business scale

Immersion cooling

Design with the latest GPUs in mind

Development + operations

Support that continues after opening

Scale, performance, costs, schedule, and service scope are proposed individually after reviewing equipment and site conditions.

Start small. Expand as your plans develop.

From a single-rack micro data center to a large-scale AI facility.We propose immersion cooling racks and facility configurations to suit your use case, deployment scale, and site conditions.

Rack counts are a guide to product selection.Large-scale configurations are individually designed by combining multiple units after reviewing power, cooling, and site conditions.Product images are illustrative.

MICRO

KAMUI γ

KAMUI γ immersion cooling rack

From 1 rack

For micro data centers and on-premises AI environments.Supports the deployment of small-scale, in-house computing infrastructure.

SMALL SCALE

KAMUI α

KAMUI α immersion cooling rack

4 or 8 racks

For AI and GPU workloads at enterprises and research institutions.We build a consolidated computing infrastructure tailored to your use case.

LARGE SCALE

KAMUI β

KAMUI β container interior

Recommended for 12 racks or more

Plan expansion based on the standard 36-rack configuration.Large-scale AI facilities are custom-designed for site conditions.

Capture GPU heat in liquid.

1.3or below

Energy-efficient design with PUE in mind

We select configurations after checking the latest GPUs for immersion compatibility, power, heat output, weight, and maintenance requirements.

Server heat is collected in oil and transferred through a heat exchanger to heat sinks such as chillers or groundwater.Chiller-based example.

Immersion cooling design can target PUE of 1.3 or below.
PUE depends on design conditions including load, ambient air, cooling method, and measurement boundary; it is not guaranteed across all configurations or seasons.

Apply our own data center experience to your facility plans.

Equipment room and electrical panels at Takahama Drip during construction, September 2024

TAKAHAMA, FUKUI

Takahama Drip 1

Our data center in Takahama, Fukui, began operating in December 2024.It uses our proprietary immersion cooling technology.

We consider monitoring, maintenance, renewal, and sales conditions from the development stage.We apply our own construction experience to proposals for our customers.

Takahama Drip equipment during construction (September 2024). It does not indicate the current completion status or operating capacity.

Plan servers, cooling, power, and connectivity together.

Example configuration: 36 racks / 6 containers

An illustrative KAMUI β configuration.

Estimated server load: 1,260 kW (35 kW × 36 racks).

Facility container illustration
C1

Server container

12 racks

C2

Server container

12 racks

C3

Server container

12 racks

C4

Facility container

Heat exchange & circulationControl & fire suppression

C5

Power & connectivity container

C6

Power & connectivity container

C1–C3: 36 racks in totalC5 & C6: 3 UPS systems in total, plus networking

The diagram summarizes functions; it is not a finalized layout.35 kW per rack is a reference operating load, not a product maximum capacity or a guarantee of equipment support.Each facility system’s capacity is designed individually.

Develop the right facility configuration for your site and business.

We define the development scope by matching the site, GPU configuration, and required operating conditions.

  1. Site & buildings

    Access and lifting, ground and foundations, maintenance access, and room for future expansion

  2. Power & outage response

    IT load, auxiliary loads and losses, incoming power conditions, and UPS and generator coordination

  3. Cooling & heat rejection

    Chillers / groundwater, ambient air, water temperature and quality, and heat rejection and drainage conditions

  4. Connectivity, safety & maintenance

    Connectivity, monitoring, access control, safety equipment, and access for equipment replacement

Area, required power, cooling method, backup capacity, schedule, and costs are finalized based on site conditions and formal design.

From defining requirements to opening.

Quantum Mesh coordinates design, KAMUI installation, and work with construction companies and other partners. The schedule varies with incoming power coordination, equipment procurement, permits, and site construction conditions.

  1. 01

    Survey & planning

    Clarify use cases, GPUs, candidate sites, and operating conditions.

    Deliverables to review: Requirements document & feasibility

  2. 02

    Design & estimate

    Define layout, power, cooling, and responsibilities.

    Deliverables to review: Basic configuration, costs & schedule

  3. 03

    Procurement & construction

    Proceed with equipment procurement and foundation, cabling, and piping work.

    Deliverables to review: Design drawings & construction records

  4. 04

    Testing & handover

    Verify load testing, outage switchover, and operational readiness.

    Deliverables to review: Acceptance, procedures & operational handover

Opening is the start of operations.

From tracking facility conditions and computing resources to maintenance, recovery, and renewal, we also support data center operations after construction.

Continuously manage facilities and computing resources
  1. Monitoring
  2. Maintenance
  3. Reporting
  4. Improvement

Back to monitoring

Operate the facility

Track temperature, flow, and power. Plan inspections, consumables, replacement parts, and renewal.

Respond to incidents

Define responsibility for detection, notification, initial response, and recovery, along with shutdown and restart procedures.

Manage computing resources

Manage GPU usage, available capacity, and contracts, coordinating maintenance and provisioning schedules.

Report to customers

Share facility, operating, and maintenance status to support decisions on improvements and expansion.

Service scope, monitoring hours, on-site response, SLA, equipment warranties, and replacement terms are defined by contract.

From computing capacity sales to revenue management.

We assess available capacity and contracts, combining direct sales with supply to compatible overseas markets. We clarify responsibilities and compensation from facility operations through sales, billing, and settlement.

  1. Capacity & contract management

  2. Sales channel coordination

  3. Usage & revenue management

DIRECT

KAMUI VM

Direct provision to customers with long-term requirements.

MARKETPLACE

Overseas computing capacity markets

Sales channels for initial operations and available capacity.

Sales conditions, settlement methods, and Quantum Mesh’s compensation are finalized by contract. Market approval, demand, and other conditions are prerequisites; utilization and revenue are not guaranteed.

Clarify roles and costs from the outset.

Customer

Define business requirements, make investment decisions, and fund data center development. Decide facility ownership and operating policies.

Quantum Mesh

Design the data center, sell and install KAMUI, and coordinate partners. Support data center operations, computing capacity sales, and revenue management.

Construction, equipment & other partners

Carry out specialist foundation, building, electrical, and piping work. Provide construction, inspection, and maintenance within the contracted scope.

Estimate development and operations separately

We organize the plan to give visibility into capital investment, post-opening costs, and computing capacity sales conditions.

The responsibilities shown are a proposal-stage outline. Facility ownership, revenue sharing, billing and settlement, warranties, and end-of-contract arrangements are agreed individually.

01Initial development

Land, site preparation and foundations, incoming power and connectivity, KAMUI, cooling and power systems, GPUs and IT equipment, design, construction, testing, and handover.

02Ongoing operations

Electricity, connectivity, consumables, maintenance and inspection, outsourced operations, incident response, and equipment renewal and replacement.

03Sales & revenue management

Sales conditions, market fees, billing and settlement, and Quantum Mesh’s service scope and compensation.

Prices, returns, and cost-reduction rates are not yet determined. GPU funding and ownership, equipment renewal, taxes, insurance, and other conditions are also considered individually.

Let’s turn your data center plans into a concrete proposal.

Even if some details are undecided, we can start with your current conditions. Please share the following four points when you contact us.

  1. 01

    Use case & scale

    Internal use / computing capacity sales, GPU models and quantities, and expansion plans

  2. 02

    Site

    Candidate sites, building drawings, incoming power, connectivity, and cooling and heat rejection conditions

  3. 03

    Operations & revenue management

    Tasks you would like to outsource, services to be offered, and required operating conditions

  4. 04

    Investment & timing

    Budget, preferred opening date, and GPU procurement and ownership policies

A formal proposal for PUE, scale, budget, schedule, and operating scope will follow a review of equipment and site conditions.

Scroll to Top