Intelligent scheduler
An AI tool for predictive workload orchestration.
NextGenDC
A data centre that carries its own research inside it.
- Role in the group
- Data centre campus with a dedicated research and development block
- Stage
- Investment proposal · pre-construction
- Location
- Rudniki
- 900MW
- Target capacity of the campus
- 746€m
- Investment sought
- 15MW
- IT load of the first module
- below 1.25
- Target PUE of the first module
Context
A research block inside the data centre
Unlike conventional data centres, NextGenDC integrates a dedicated research and development block.
The strength of NextGenDC lies in its ability to connect leading-edge research with actual deployment. Each pillar addresses a fundamental problem in operating a modern data centre, and each is validated on the campus itself.
Buildings and facilities
The first module of the campus
| Building | Function | Parameters |
|---|---|---|
| BO 01 | Server halls | 4,000 m², IT load of 15 MW, density of 40–50 kW per rack |
| BO 02 | Energy core | 2,500 m², HV/LV switchgear, UPS, back-up generation (diesel/gas transitioning to hydrogen/biogas); total output of 50 MW |
| BO 03 | Cooling centre | 1,800 m², free cooling and immersion cooling, COP above 6.5, target PUE below 1.25 |
| BO 04 | Research block | 2,900 m², laboratories for algorithm development, AI testing and digital twin simulation |
| BO 05 | Administration | 1,200 m², offices, helpdesk and customer facilities |
These figures describe the first module. The campus at Rudniki is designed to be phased from 100 MW up to a total capacity of 900 MW.
Technology
What the facility is built from
IT infrastructure
480 blade servers and 72 GPU nodes; NVMe SAN storage, hybrid arrays and LTO-9 tape libraries; 400 GbE backbone switching, SDN and full A/B redundancy; next-generation firewalls, IDS/IPS and AI-assisted SIEM.
Energy infrastructure
22/110 kV transformers, modular UPS and back-up generation with 48 hours of autonomy; smart-grid integration with the ability to export power back to the grid; a transition strategy towards renewable fuels.
Cooling
Six 1.5 MW water-cooled units and four 1 MW air-cooled units on standby; CRAH and in-row liquid cooling for high-density racks; more than 600 sensors for dynamic AI-driven temperature control.
Security
Multi-layer protection: biometrics, RFID, security airlocks and an AI-assisted camera system; redundant fibre connections through separate providers; fire protection with VESDA and inert gases.
The research component
What is being developed on the campus
Algorithm development
Classical machine learning combined with neuromorphic methods and approaches inspired by quantum computing.
Integration laboratory
Interoperability testing across IoT, edge and cloud systems.
Ontological AI orchestration
A higher-level orchestration system that understands infrastructure, workload, energy and security processes in context and manages them.
Pillars of the solution
Three pillars, one operating system
- 1
Intelligent scheduler
Traditional data centres struggle with unpredictable workloads and demand peaks, and most scheduling is still manual — which wastes resources and raises costs. The intelligent scheduler introduces AI-driven predictive workload management that allocates resources according to current and anticipated demand, and integrates the CI/CD pipeline so that new services deploy without downtime and without manual intervention.
- 2
Algorithm development
At the heart of an autonomous data centre are algorithms that learn, adapt and respond. This pillar combines machine learning, brain-inspired neuromorphic models and quantum computing techniques in order to optimise energy consumption, predict failures before they occur and counter cyber threats without human intervention — a shift from reactive administration to self-healing infrastructure.
- 3
Integration laboratory
The third pillar is the framework that joins everything into a single ecosystem and ensures that IoT devices, edge nodes, cloud layers and the hardware beneath them work together. It is a test environment in which new technologies are validated under real conditions before deployment, so the infrastructure stays scalable and ready for what comes next.
SWOT analysis
Where the project is strong and where it is exposed
Strengths
A distinctive combination of infrastructure and research; a high degree of autonomy and AI orchestration; an ESG-aligned design and low PUE; a strategic location in central Europe.
Weaknesses
High initial capital expenditure; the demands of running a data centre and a research operation in parallel; dependence on long-term partnerships; rapid technological change may overtake individual components.
Opportunities
Double-digit growth in the AI and edge markets; a shortage of local capacity in central Europe; EU programmes supporting green and digital infrastructure; partnerships with telecommunications operators and academia.
Threats
Entry of global players; volatility in energy prices; geopolitical and regulatory risk; rapid technological obsolescence.
What will actually decide this
Of everything above, two items decide the outcome and the rest is secondary.
The grid: a signed connection agreement turns land into land with power, and is the single largest step up in value.
Zoning: without an amendment to the municipal plan, roughly half of the area cannot be built on at all.
Global competition and energy prices matter, but they change the return; these two items change whether the project exists at all.
Financial plan
The investment proposal
- Investment sought
- €746 million
- Use of proceeds
- Construction, IT infrastructure, research and development, expansion
- Exit options
- IPO, strategic sale or a long-term income-producing asset
Summary of the financial model
Financial indicators for the period 2025–2035 and the valuation outputs.
| Indicator | Value |
|---|---|
| Revenue | €4.81 billion |
| EBITDA | €4.56 billion |
| Net cash flow | €3.54 billion |
| CAPEX | €979 million |
| Payback | From 2028, repaid within six years |
| IRR | Significantly above a WACC of 8–9% |
| EVA | €8.71 billion cumulative |
| DCF equity value | €16 billion, risk-adjusted to €13.6 billion |
Financial model outputs from feasibility study v1.1.
Overview
Key details
- Campus
- Rudniki; phased from 100 MW to a total capacity of 900 MW
- First module
- Five buildings, BO 01–BO 05, IT load of 15 MW, total energy core output of 50 MW
- Operating parameters
- Density of 40–50 kW per rack, target PUE below 1.25, cooling COP above 6.5
- Research
- A dedicated 2,900 m² block: algorithm development, AI testing and digital twin simulation
- Investment sought
- €746 million for construction, IT infrastructure, research and development, and expansion
- Exit options
- IPO, strategic sale or a long-term income-producing asset
The financial model outputs derive from feasibility study v1.1. They are modelled values, not commitments or forecasts of achieved performance.
The detail sits in the data room
Investment studies, contractual documentation and the data room are made available to qualified investors after an initial meeting and the signing of an NDA.