Microsoft has finalized a $9.7 billion, five-year agreement with Australia’s IREN to secure additional AI cloud capacity. The deal grants Microsoft access to compute infrastructure powered by Nvidia GB300 GPUs, deployed in phases at IREN’s Childress, Texas, facility through 2026. The site is planned to support up to 750 megawatts of capacity.
IREN is separately investing $5.8 billion in GPUs and equipment from Dell to expand its AI-ready infrastructure. Microsoft recently launched its first production cluster using Nvidia GB300 NVL72 systems for Azure, designed to optimize reasoning models, agentic AI systems, and multi-modal generative AI applications.
This deal follows Microsoft’s recent agreement with Nscale for roughly 200,000 Nvidia GB300 GPUs across European and U.S. data centers. IREN, initially a bitcoin-mining company, has pivoted to AI workloads, with CEO Daniel Roberts estimating the Microsoft contract will consume about 10% of total capacity and generate $1.94 billion in annualized revenue.
