Microsoft’s $9.7B AI Cloud Deal with Australia’s IREN
▼ Summary
– Microsoft signed a $9.7 billion, five-year contract with IREN to secure additional AI cloud capacity using Nvidia GB300 GPUs.
– The compute infrastructure will be deployed in phases through 2026 at IREN’s Texas facility, which is planned to support 750 megawatts of capacity.
– IREN is separately purchasing GPUs and equipment from Dell for approximately $5.8 billion to support this initiative.
– This follows Microsoft’s recent deals for Nvidia GB300 GPUs, including a production cluster on Azure and an agreement with Nscale for data centers in Europe and the U.S.
– IREN, originally a bitcoin-mining company, shifted to AI workloads and expects the Microsoft deal to use 10% of its capacity and generate about $1.94 billion in annual revenue.
Microsoft has solidified a massive $9.7 billion artificial intelligence cloud partnership with Australian firm IREN, a strategic five-year agreement designed to significantly expand its computational resources. This move directly addresses the soaring global demand for advanced AI services, ensuring Microsoft can deliver robust infrastructure to its enterprise clients.
The collaboration grants Microsoft access to cutting-edge computing infrastructure powered by Nvidia’s high-performance GB300 GPUs. Deployment will occur in stages, with full implementation expected by 2026 at IREN’s data center located in Childress, Texas. This facility is engineered to support an impressive 750 megawatts of power capacity, a critical requirement for running intensive AI workloads.
In a related development, IREN is independently procuring GPUs and associated equipment from Dell Technologies in a transaction valued at approximately $5.8 billion. This separate investment underscores the scale of the overall operation and the company’s commitment to building out its AI-focused data center capabilities.
This landmark deal follows closely on the heels of Microsoft’s recent launch of its first production cluster utilizing Nvidia’s GB300 NVL72 systems on its Azure cloud platform. Microsoft has highlighted that these systems are specifically fine-tuned for complex tasks, including running sophisticated reasoning models, developing agentic AI systems, and processing multi-modal generative AI. Just last month, the tech giant also entered into another substantial arrangement with Nscale, securing around 200,000 Nvidia GB300 GPUs for deployment across three European data centers and one in the United States.
IREN’s journey mirrors a broader industry trend. Originally a bitcoin-mining enterprise, the company astutely recognized that its vast array of GPUs could be far more profitably leveraged for the burgeoning field of artificial intelligence. This strategic pivot has proven enormously successful. According to a Bloomberg report, IREN’s CEO Daniel Roberts anticipates that the Microsoft agreement will utilize merely 10% of the company’s total available capacity. Despite this modest allocation, the deal is projected to generate roughly $1.94 billion in annualized revenue for IREN, highlighting the immense financial potential of the AI infrastructure market.
(Source: TechCrunch)
