AI & HPC Infrastructure

Clusters

GPU clusters

IRIS has been purpose-built to support GPU-intensive AI training, inference, and hyperscale cloud workloads through a scalable high-density architecture designed for next-generation compute environments.IRIS has been purpose-built to support GPU-intensive AI training, inference, and hyperscale cloud workloads through a scalable high-density architecture designed for next-generation compute environments.

IRIS has been purpose-built to support GPU-intensive AI training, inference, and hyperscale cloud workloads through a scalable high-density architecture designed for next-generation compute environments.IRIS has been purpose-built to support GPU-intensive AI training, inference, and hyperscale cloud workloads through a scalable high-density architecture designed for next-generation compute environments.

Optimized for next-generation AI infrastructure operations

KONNECT is a modular, vendoragnostic high density data center platform designed for 50 MW+ net ITE loads and AI/HPC workloads

5 MW KONNECT base system

Approximately 4.2 MW IT capacity

56 racks and Up to 112 kW per AI rack

Optimized for next-generation AI infrastructure operations

KONNECT is a modular, vendoragnostic high density data center platform designed for 50 MW+ net ITE loads and AI/HPC workloads

5 MW KONNECT base system

Approximately 4.2 MW IT capacity

56 racks and Up to 112 kW per AI rack

Get Started

DLC

Direct liquid cooling

DLC

Direct liquid cooling

The campus integrates direct liquid cooling systems designed to maximize rack density, improve thermal efficiency, and support high-density AI and HPC rack environments.

Improve thermal efficiency

Optimize operational performance across hyperscale compute environments

Hybrid liquid cooling using DLC + RDHx, with aircooled variant also available

The campus integrates direct liquid cooling systems designed to maximize rack density, improve thermal efficiency, and support high-density AI and HPC rack environments.

Improve thermal efficiency

Optimize operational performance across hyperscale compute environments

Hybrid liquid cooling using DLC + RDHx, with aircooled variant also available

Hyperscale

Hyperscale readiness

A campus built at hyperscale

IRIS combines scalable power infrastructure, vendor-agnostic architecture, modular-by-design deployment, renewable energy integration, and hyperscale-grade operational design to support long-term AI and cloud infrastructure expansion across Europe. Designed for maximum flexibility, the platform can scale efficiently from single modules to large multi-block campuses as demand grows.

AI-Ready Infrastructure

modular HPC deployment

A campus built at hyperscale

IRIS is engineered to support high-performance computing environments through factory-built modular infrastructure deployment, AI-ready rack density architecture, 4N/3 distributed power topology, modular UPS with hot-swap modules and thermal buffer vessels, flexible power configurations, and advanced cooling systems optimized for large-scale compute operations.

AI-ready Rack Density Architecture

Flexible Power Configuration

Advanced Cooling Systems

features

Safety & Reliability Features

features

Safety & Reliability Features

N2Gen Suppression

ESD Flooring

RC2 security doors

IEC/UL variants

Vendor-Agnostic Procurement Flexibility

N2Gen Suppression

ESD Flooring

RC2 security doors

IEC/UL variants

Vendor-Agnostic Procurement Flexibility

Capabilities

AI compute capabilities

Capabilities

AI compute capabilities

The infrastructure supports:

The campus integrates modular high-density rack architecture, flexible deployment infrastructure, and AI-ready cooling systems engineered for next-generation compute operations.

AI training workloads

AI inference clusters

high performance compute

hyperscale operators

enterprise HPC workloads