AI Data Centers and mining site conversion
We help site owners, investors and companies bring GPU compute infrastructure online. We work with existing data centers and mining farms as well as with projects built from the ground up.
We take on the technical and organisational side of the project: from assessing the site and its available electrical capacity to designing, equipping and commissioning the compute infrastructure.
Discuss your projectWorking with existing sites
If you already run a mining farm or a data center, we can assess whether it can be repurposed for AI workloads.
We review:
- available electrical capacity;
- condition of the engineering infrastructure;
- the cooling system;
- equipment layout;
- network infrastructure;
- location of the facility;
- compute hardware already on site.
Depending on the condition of the site, several routes are possible: acquiring the facility, partial modernisation, or a full conversion to GPU infrastructure.
From mining to AI
Mining sites can become the basis for new compute capacity, but the existing infrastructure is not always designed for modern AI loads.
During a retrofit we determine which parts of the facility can be kept and which have to be replaced or upgraded.
A project may include:
- a new cooling system;
- power distribution upgrades;
- a revised rack layout;
- high-speed networking;
- server infrastructure;
- GPUs and other compute hardware.
The goal is a site that meets the requirements of modern compute clusters.
Building a new AI data center
For projects where the existing infrastructure is not enough, we organise the construction of a new facility.
We consider projects from 5 MW upwards. Depending on the task we can take part in site selection, power capacity assessment, engineering design, permitting, construction and fit-out.
For large projects we plan infrastructure that leaves room for further expansion of compute capacity.
Hardware for compute infrastructure
We supply the components required to build AI clusters and purpose-built data centers.
Depending on the project we can specify and source:
- GPUs and servers;
- liquid cooling systems;
- electrical equipment;
- network components;
- rack hardware;
- modular solutions.
We work with new, used and refurbished hardware. The configuration is selected against the required performance, the density of the deployment and the power available at the site.
Technical and investment analysis
Before committing significant capital, the project can be assessed up front.
We analyse the site, the power infrastructure, the intended hardware configuration and the economics of the project. That establishes technical feasibility and allows several development scenarios to be compared.
Advisory work can be delivered as a standalone service or as part of a full project.
How the work is phased
Work usually starts by reviewing the inputs and defining the requirements for the future infrastructure.
- 1
Project analysis
We study the site, the power supply, the budget and the target compute capacity.
- 2
Site and power verification
We establish whether the existing location is suitable, or find an alternative with the required power availability.
- 3
Permitting
We review zoning, environmental requirements and construction permits.
- 4
Engineering design
We develop the power, cooling, structural and equipment layout solutions.
- 5
Construction and procurement
We coordinate the build and the hardware supply as a single project.
- 6
Commissioning
We test the engineering systems, install and verify the hardware, after which the facility is prepared for operation.
For greenfield projects, first power is typically targeted around nine months after the required permits and the power capacity are confirmed. Retrofitting an existing site can shorten that timeline.
Engagement models
We can join a project in several formats:
Build to order
We build the infrastructure to the client's requirements and hand the finished facility over to the owner.
Partnership project
We co-invest in the facility and its development, sharing the investment risk and the future returns.
Developing an existing facility
We identify promising sites or facilities with unresolved infrastructure issues, assess the risks and prepare them for further use.
Hardware supply
We arrange the supply of the required hardware, whether or not the full scope of work is needed.
Advisory support
We join at a specific stage or accompany the project from the initial assessment through to commissioning.
Why this approach
The main advantage of an end-to-end engagement is that the parts of the project are coordinated within one system.
Power supply, cooling, equipment layout and construction are considered together rather than as independent tasks. Technical constraints surface early, and the bulk of the capital is not committed until feasibility is confirmed.
Discuss your project
Get in touch if you are considering:
- selling a mining farm or a ready site;
- converting an existing facility into AI infrastructure;
- building a new compute center;
- investing in infrastructure for AI;
- purchasing GPUs, servers or engineering equipment;
- a technical review of an existing project.
Tell us about your site or your objective - after an initial assessment we can determine the most suitable way forward.