Long-Term AI Infrastructure Planning

Power-Backed Capacity for AI and Large-Scale Compute

Island Roadhouse is developing a long-term data center capacity platform for organizations planning AI, hyperscale, and enterprise compute deployments. Explore the capacity offering, understand the engagement path, or begin a capacity discussion.

On-site generation behind the meter, with advanced atomic power in the completed campus

Closed-loop warm-water liquid cooling for high-density data halls

Proprietary thermal management designed for optimal energy use

Customers bring and operate their own compute

Campus at a glance

The Missouri campus, in eight facts

Capacity

Up to 154 MW

Sellable IT load planned in Phase 1

Data halls

12 MW

Repeatable GPU and HPC halls, alongside 5 MW general-purpose halls

Electrical

2N

Electrical distribution to the rack in GPU and HPC halls

Cooling

N+1

Cooling equipment at block level

Liquid cooling

Direct-to-chip

Closed-loop warm-water liquid cooling for high-density racks

Thermal

Low-water

Proprietary thermal management designed for optimal energy use

Connectivity

Dual

Meet-me rooms, carrier-neutral, with diverse fiber pathways planned

Power

On-site

Generation behind the meter carries the compute load, with advanced atomic power in the completed campus

Open farmland under a clear sky

Location

Why Missouri

Missouri provides a strategic foundation for next-generation infrastructure: industrial land, central U.S. access, engineering talent, and long-term development potential.

Industrial Scale

Large-scale infrastructure requires land, logistics, and room to expand.

Engineering Talent

Missouri offers access to technical institutions, skilled trades, and a growing infrastructure workforce.

Long-Term Development

IRDC is being planned as a phased campus designed to support infrastructure growth over time.

The platform

How the campus fits together

The campus is planned as one system: power generated on site, proprietary thermal management, repeatable data halls and carrier-neutral connectivity, designed together rather than assembled from separate projects.

A concept illustration of the campus as four layers planned together, from the ground up: on-site power generation; proprietary thermal management; repeatable data halls, with 12 MW GPU and HPC halls, 5 MW general-purpose halls and room for more in phases; and carrier-neutral connectivity, with dual meet-me rooms and diverse fiber pathways. Beside the platform, heat reuse where a real use exists. Not a site plan and not to scale. 12 MW GPU and HPC hall 5 MW general-purpose hall Later phases 01 02 03 04 05
Concept illustration. Not a site plan and not to scale.

On-site power generation

Generation behind the meter carries the compute load, with advanced atomic power in the completed campus.

Proprietary thermal management

Designed for optimal energy use, in a low-water campus.

Repeatable data halls

12 MW GPU and HPC halls and 5 MW general-purpose halls, repeated in phases.

Carrier-neutral connectivity

Dual meet-me rooms and diverse fiber pathways planned, so customers keep their choice of carrier.

Heat reuse

Recovered heat made available for beneficial use, where a real use exists.

Development status

Where the project stands

The campus is in design review, and capacity discussions are open now. Early engagement shapes hall requirements, density and deployment timing.

Announced

Design approach

The integrated campus design has been announced publicly.

In progress

Engineering and design review

Power, thermal and data hall design are being developed together.

Open now

Capacity discussions

Initial conversations, with a mutual NDA where detailed planning calls for one.

Ahead

Development and construction

Approvals, procurement and phased construction follow engineering.

Ahead

Commercial operation

We will give a date when we can stand behind it.

Engage with Island Roadhouse

Start the conversation that fits your program

Each path opens the contact form with your inquiry type already selected. Hyperscaler and government programs also have a page written for them.

Enterprise and AI

Discuss future capacity

Plan long-term, power-backed capacity for AI, HPC or enterprise deployments.

Hyperscalers and cloud

Hyperscaler co-development

Plan dedicated halls and fit-out requirements within the campus’s repeatable 12 MW blocks.

Read about the program

Government and defense

Government and secure compute

Plan a sponsored secure program around a hall planned as an ICD 705 accreditation-ready shell.

Read about secure compute

Technology and supply chain

Technology and infrastructure suppliers

Equipment, systems and services for power, cooling, construction and connectivity.

State and local

Economic development and public sector

Officials and development organizations working on infrastructure, jobs and growth in Missouri.

Everyone else

Media and general inquiries

Press questions and anything that does not fit the paths above.

Next step

Start a Capacity Discussion

Organizations evaluating long-term AI, hyperscale, or enterprise compute capacity can begin with a capacity discussion to align deployment needs, timing, and the appropriate capacity profile.