Microsoft is making every effort to expand its computing resources to satisfy the growing demand for its AI offerings.
On Monday, the technology leader from Redmond entered into a five-year agreement worth $9.7 billion with Australia’s IREN to boost its AI cloud capabilities. Through this partnership, Microsoft will gain access to computing infrastructure powered by Nvidia’s GB300 GPUs, which will be rolled out in stages at IREN’s Childress, Texas facility through 2026, with plans to support 750 megawatts of capacity.
IREN also announced a separate purchase of GPUs and hardware from Dell, totaling around $5.8 billion.
This agreement follows Microsoft’s recent launch of its inaugural production cluster featuring Nvidia’s GB300 NVL72 systems for Azure, which the company claims are tailored for reasoning models, agent-based AI, and multimodal generative AI applications.
In the previous month, Microsoft also reached an agreement with Nscale to supply roughly 200,000 Nvidia GB300 GPUs to three European data centers and one in the United States.
Much like rivals such as CoreWeave, IREN originally focused on bitcoin mining but soon realized its extensive GPU resources were better suited for AI processing. This strategic pivot has significantly benefited the company. According to Bloomberg, CEO Daniel Roberts anticipates that the Microsoft partnership will utilize just 10% of IREN’s total capacity and is expected to generate approximately $1.94 billion in annual revenue.
