Skip to content
SOLUTIONS

Build. Convert. Equip. Fill.

Four services in the order a project needs them. One team from the first site visit to the first tenant.

THE MODEL

One partner for the whole chain, or for the step you need

A 5–100 MW owner usually manages a consultant, a contractor, several equipment vendors and a leasing broker. Each has its own schedule. The hand-offs between them are where the months go.

We do all of it. Engage us for one service, such as equipment only or tenants only, or for the whole chain. The same team carries the project either way.

Typical engagement
Construction consultant alongside a local contractor, or turnkey delivery. Role agreed per project.
What you get
A facility energised on a schedule driven by equipment lead times we control, sized for the tenants you will sign.
01 · BUILD

Data Centre Construction Consulting

Construction consulting and turnkey delivery for 5–100 MW data centres. We start before land is bought: grid capacity, connection path, cooling water, fibre and permits. We stay through energisation, commissioning and handover. Substations, cooling plant, generator yards and fibre routes are in our scope.

What we deliver

  • Site, power and grid-connection due diligence
  • Load schedule, single-line diagram, redundancy topology (N+1 / 2N)
  • Design management and specification: electrical, mechanical, structural
  • EPC management or turnkey delivery, including ancillary works
  • Procurement of the long-lead power chain through our own supply lines
  • Commissioning, integrated systems testing, handover documentation
Discuss a build
Typical engagement
Feasibility study first, then phased design and delivery. Mining usually keeps running until the last hall.
What you get
AI/HPC-ready capacity months faster than a new build, on power you already pay for.
02 · CONVERT

Mining-Site Conversion

Crypto-mining sites already have grid power, transformers, buildings and often cooling water. We assess what stays, then rebuild the rest for AI and HPC: redundancy, cooling, distribution, network and security. Conversions can run hall by hall so mining revenue continues while HPC capacity comes online.

What we deliver

  • Feasibility: power reuse, structural and cooling assessment
  • Redundancy upgrade, usually N to N+1, with UPS and generator integration
  • Cooling retrofit to liquid-ready or immersion-ready systems
  • Distribution rebuilt for 40–130 kW racks: switchgear, busway, PDU
  • Network, fibre and physical security to colocation standard
  • Phased cut-over plan and commissioning of each hall
Assess a site
Typical engagement
A coordinated equipment package for a whole facility, or one urgent unit from stock.
What you get
One packing list, one delivery plan, one contact for drawings, site readiness and warranty.
03 · EQUIP

Power Equipment Supply

The power chain has the longest lead times in any build, so we supply it ourselves, for our projects and for owners building with someone else. Transformers, MV/LV switchgear, SST, three-phase UPS, PDU, busway and cable from JST, Pearl Electric, China XD, Shunde TBEA, Dongfang Electric, CATL and Delta, delivered on one schedule, with spares and a field team in North America. Where the grid cannot deliver in time, we supply the generation as well: gas turbine + storage + SST as one system.

What we deliver

  • Oil-immersed transformers to 180 MVA / 230 kV; dry-type to 69 kV; ANSI pad-mounted
  • Medium-voltage switchgear 5–38 kV, compact substations, LV switchboards
  • Critical power: Delta three-phase UPS 10 kVA – 2.5 MW with CATL lithium batteries
  • Cabinet and rack PDU, cast-resin busway 250–6,400 A, remote power panels
  • On-site power: Dongfang Electric gas turbines + CATL storage + Delta SST as one package
  • Factory testing, approval drawings, North American listing confirmed per model
Send a specificationEquipment catalog and stock
Typical engagement
An owner mandate to fill a facility, or a compute reservation for an AI team. Both start with a short requirements call.
What you get
Leased halls and recurring revenue for owners. Blackwell-class capacity on demand for AI teams.
04 · FILL & COMPUTE

Capacity & Compute

A data centre earns once it is leased. For owners, we source and qualify tenants and structure colocation, wholesale or powered-shell leases. For AI teams, we lease NVIDIA B100 / B200 GPU compute by the node, rack or cluster, hosted in facilities in our project network.

What we deliver

  • Tenant sourcing and qualification against your capacity, density and term
  • Lease structuring: colocation, wholesale, powered shell
  • Technical fit review between tenant requirements and the facility
  • HGX B100 / B200 nodes, eight GPUs per node, by node, rack or cluster
  • High-bandwidth fabric for multi-node training; monthly to multi-year terms
  • Capacity planning that links tenant demand to what we build
Talk about capacityCapacity & compute in detail
HOW AN ENGAGEMENT RUNS

From first call to first tenant

Six steps. Enter at any one. The same team carries the project through the rest.

  1. 01

    Requirements call

    Site, power, capacity, timeline and budget in one conversation. We say early whether the project fits.

  2. 02

    Feasibility and site assessment

    Grid capacity, connection path, cooling, fibre, permits. For conversions, what existing infrastructure can stay.

  3. 03

    Design and specification

    Load schedule, single-line diagram, redundancy and equipment list, with lead times on every long-lead item.

  4. 04

    Procurement and build

    Equipment ordered the week the design is frozen. EPC management or turnkey delivery, including ancillary works.

  5. 05

    Commissioning and handover

    Factory and site acceptance, integrated systems testing, documentation and training. Spares positioned in North America.

  6. 06

    Fill and operate

    Tenants sourced and leases signed. GPU compute leased on the capacity. Field service and spares for the life of the equipment.

WHO WE WORK WITH

Built for the 5–100 MW segment

Too large to do without a delivery team. Too small for the hyperscale EPCs.

  • 01

    Small and mid-size developers

    Planning a 5–100 MW facility and want one partner for feasibility, design, delivery and equipment.

    Build · Equip · Fill

  • 02

    Mining operators

    Have power and buildings in place and want to move into AI/HPC hosting without starting over.

    Convert · Equip · Fill

  • 03

    Land and power holders

    Own a site with secured grid capacity and want a data centre built on it, with a tenant lined up early.

    Build · Fill

  • 04

    AI companies and enterprises

    Need Blackwell-class GPU capacity now, by the node or the cluster, without building anything.

    Compute

  • 05

    EPCs, utilities and industrial buyers

    Building their own facility and need long-lead power equipment on one schedule, with service in North America.

    Equip

Which step are you at?

Site, conversion, equipment or tenants. Tell us where the project stands and we come back with the fastest route to power and revenue.

Start a project