Skip to content
AI GPU Data Center

Capacity & offtake

Capacity should be planned before it becomes urgent.

Organisations with substantial future compute requirements rarely secure the infrastructure they need reactively. AI GPU Data Center works with prospective clients to define demand, configure infrastructure and structure the commercial framework that supports phased deployment.

What we structure

Commitment on both sides of the constraint.

01

Dedicated capacity reservations

Reserved infrastructure held against a defined workload and term, rather than contended for at the point of need.

02

Long-term client offtake

Structured commitments that underwrite phased infrastructure development and give clients a defined pathway to dedicated capacity.

03

Workload forecasting

Joint modelling of training, inference and research demand so reserved capacity reflects the real growth curve.

04

Client-specified hardware

GPU architecture, interconnect, storage and operational configuration selected against technical and commercial requirements.

05

Phased infrastructure deployment

Capacity commissioned in increments aligned to the agreement rather than speculative build-out.

06

Commercial alignment

Commercial structure, term and expansion triggers agreed before infrastructure is committed.

07

Managed operational capacity

Reserved capacity operated, monitored and maintained under an agreed service framework.

Capacity enquiry pathway

From first conversation to operating capacity.

  1. 01

    Workload assessment

    Model current and projected workloads, utilisation patterns and data requirements.

  2. 02

    Technical configuration

    Specify GPU architecture, interconnect, storage, security and operational needs.

  3. 03

    Capacity and term planning

    Size reserved capacity against the planning horizon and expected growth curve.

  4. 04

    Commercial framework

    Structure reservation or offtake terms that underwrite phased infrastructure development.

  5. 05

    Deployment

    Build and commission contracted capacity in line with the agreed phasing.

  6. 06

    Ongoing operations

    Operate, monitor and plan lifecycle activity under the agreed service framework.

Contract-led financing model

Capital follows contracts, not sentiment.

Infrastructure development is financed against contracts rather than sentiment. Approximately US$20m of committed debt capital forms the base, and capacity is expanded as offtake is signed.

Client offtake agreements are structured over three to five year terms, which is what makes each deployment financeable.

The capital cycle

  1. 01Committed capital
  2. 02GPU deployment
  3. 03Contracted offtake
  4. 04Financeable cash flow
  5. 05Additional capital
  6. 06Additional GPU capacity

Why offtake matters

  • Offtake converts planned capacity into contracted revenue, which is what lenders underwrite.

  • It sets the size and timing of each module, so capital is not committed ahead of demand.

  • It gives clients a defined pathway to dedicated capacity instead of competing for it later.

Phased development

Expansion follows contracted demand.

Phase 01

1MW planned initial deployment

Dedicated high-density infrastructure based on NVIDIA B300 systems, subject to final engineering and client requirements.

Expansion

Phased pathway towards 5MW

Additional capacity developed according to client uptake, power availability, technical requirements and contracted offtake.

Future Scale

Repeatable infrastructure growth

A modular operating model designed to support additional capacity and future infrastructure locations as demand develops.