IREN signs USD $9.7 billion AI cloud deal with Microsoft

IREN has signed a multi-year agreement with Microsoft worth about USD $9.7 billion to supply AI cloud infrastructure, centred on GPU deployments at IREN's site in Childress, Texas.

Under the deal, IREN will provide large-scale GPU clusters housed in liquid-cooled data centres being built at its 750MW campus. The infrastructure will use NVIDIA GB300 GPUs and will be rolled out in four phases through 2026, with the full deployment representing 200MW of critical IT load.

The five-year contract includes a 20 per cent prepayment and is expected to add roughly USD $1.94 billion in annualised run-rate revenue once the installation is fully commissioned.

Capacity demand

The agreement gives Microsoft access to a substantial tranche of AI computing capacity at a time when major technology groups are seeking more data centre infrastructure for model training and inference. For IREN, it marks a sharp expansion of its role as a supplier of GPU infrastructure to large customers.

IREN has also signed an agreement with Dell Technologies to buy the GPUs and related equipment needed for the project for about USD $5.8 billion. The purchase underpins the hardware side of a build-out that will be housed in facilities IREN controls directly.

IREN has presented that ownership structure as a key part of its approach, retaining control over land, grid connections, data centres and GPU hardware. It says the model helps it respond to rising demand for dense computing infrastructure as AI workloads grow in scale.

Texas build-out

Childress is central to that effort. The Texas campus has 750MW of capacity, and the Microsoft deployment will account for 200MW of critical IT load, indicating that a significant share of the site is being reserved for the project as each phase comes online.

The deal is one of the larger AI infrastructure agreements announced this year, reflecting the spending race among cloud providers and their suppliers as demand for advanced computing systems rises. It also shows how specialised operators are trying to secure long-term contracts before new data centre supply comes to market.

IREN has sought to position itself around large-scale GPU clusters for AI training and inference, supported by sites in the US and Canada. Its broader North American power portfolio totals 3GW of secured capacity.

Customer expansion

Daniel Roberts, described the agreement as a marker of the company's expansion into a different customer base.

"We're proud to announce this milestone partnership with Microsoft, highlighting the strength and scalability of our vertically integrated AI Cloud platform. This agreement not only validates IREN's position as a trusted provider of AI Cloud services, but also opens access to a new customer segment among global hyperscalers. It marks another major step forward for IREN as we continue to expand large-scale GPU deployments across our 3GW secured power portfolio in North America, reinforcing our position as a leading AI Cloud Service Provider," said Roberts.

Microsoft also explained why it chose IREN for the project.

"Together with IREN, Microsoft is delivering cutting-edge AI infrastructure for our customers. IREN's expertise in building and operating a fully integrated AI cloud from data centres to GPU stack, combined with their secured power capacity, makes them a strategic partner. This collaboration unlocks new growth opportunities for both companies and the customers we serve," said Jonathan Tinter.

Investment scale

The scale of the agreement underlines the financial intensity of AI data centre construction, where operators must secure land, electricity, cooling systems, servers and chips well before revenue is fully realised. With a USD $9.7 billion customer contract and a separate USD $5.8 billion equipment agreement, the Childress build-out is a large example of how infrastructure providers are tying together power access, hardware procurement and long-term customer commitments.