IREN Limited has completed delivery of its first AI data center module to Microsoft at its Childress, Texas campus, while simultaneously earning NVIDIA Exemplar Cloud status for its deployment of the NVIDIA GB300 NVL72 system, the company announced.

Horizon 1 Delivered and Accepted

The milestone marks the completion of Horizon 1, the first of four planned 50-megawatt direct-to-chip liquid-cooled AI Cloud deployments scheduled to go to Microsoft at the Childress facility during 2026. Microsoft has formally accepted the delivery.

The four deployments are part of a five-year, USD 9.7 billion cloud services contract that the two companies announced in November 2025. Daniel Roberts, Co-Founder and Co-CEO of IREN, credited the company's vertically integrated structure for enabling rapid execution.

"Delivering Horizon 1 demonstrates the strength of our vertically integrated model and our ability to execute complex AI infrastructure projects at speed and scale," Roberts said in the announcement.

He also acknowledged the workforce behind the effort, thanking more than 3,000 people across the site team for their expertise, commitment, and execution.

What Vertical Integration Means for IREN's Model

IREN described its vertically integrated approach as providing end-to-end control across the design, engineering, and construction of the data centers that support its GPU deployments.

The company characterized this control as central to the speed at which Horizon 1 was completed and handed over to Microsoft.

The company operates as a vertically integrated AI Cloud platform, delivering data centers, compute, and software for AI training and inference.

Its platform is supported by a portfolio of land- and grid-connected power in renewable-rich regions across North America, Europe, and Asia-Pacific.

NVIDIA Exemplar Cloud Designation

IREN's NVIDIA Exemplar Cloud status was granted following NVIDIA's own testing of the GB300 NVL72 deployment housed within Horizon 1.

According to IREN, the designation demonstrates the company's ability to support demanding AI workloads with what NVIDIA evaluated as meeting standards for performance, reliability, and scale.

The GB300 NVL72 is part of NVIDIA's current generation of high-performance computing hardware aimed at large-scale AI workloads.

Achieving Exemplar Cloud status on this specific platform places IREN within a designated tier of cloud providers recognized by NVIDIA for their deployment capabilities.

Three More Modules Scheduled for 2026

With Horizon 1 now accepted by Microsoft, IREN has turned its attention to the remaining three modules under the current agreement.

Roberts stated the company looks forward to building on this momentum with Microsoft as it works to deliver Horizons 2, 3, and 4 later this year.

Each of the four modules carries a 50-megawatt IT load capacity and uses direct-to-chip liquid cooling technology, a design approach suited to the thermal demands of dense GPU deployments.

The Childress, Texas campus serves as the central location for all four deliveries under the existing contract term.

The structure of the deal, spanning five years and totaling USD 9.7 billion, makes it one of the more substantial long-term cloud services arrangements disclosed in the sector in recent years.

Broader Capacity Targets for 2026 and 2027

Beyond the immediate Microsoft deliveries, IREN has outlined broader infrastructure growth targets. The company continues to target expansion to 480 megawatts of gross AI Cloud capacity during 2026, with a further target of 1.2 gigawatts of gross AI Cloud capacity in 2027.

These figures represent the company's stated ambitions for its total platform footprint, not solely the capacity being delivered to Microsoft under the current contract.

The 480-megawatt target for 2026 would encompass the four 50-megawatt Horizon deployments at Childress along with other capacity across IREN's portfolio.

The jump to 1.2 gigawatts in 2027 would represent a significant scaling of that base should the company execute on its plans.

Context Within the AI Infrastructure Buildout

The delivery of Horizon 1 comes as demand for high-density, liquid-cooled AI data center capacity has accelerated across the industry, driven by the compute requirements of large language models and other AI workloads.

Direct-to-chip liquid cooling, which IREN has deployed across the Horizon modules, has emerged as a preferred method for managing heat in environments running dense GPU clusters such as the NVL72. IREN's stock trades on the NASDAQ exchange under the ticker IREN.