Microsoft Partners with IREN for USD 9.7 billion AI Compute Boost

Microsoft has signed a monumental five-year, $9.7 billion deal with Australian data center operator IREN to secure much-needed compute power for its Azure AI services. The agreement marks a significant shift in IREN's business focus as it moves from bitcoin mining to providing infrastructure for AI workloads, capitalising on the increasing demand for AI cloud capacity.

Addressing AI Compute Bottlenecks

For Microsoft, the partnership with IREN represents a strategic manoeuvre to bypass the existing compute bottleneck that many AI companies are facing. The demand for powerful GPUs to handle complex workloads, like AI reasoning models and generative systems, has surged globally, pushing hardware providers to capacity limits.

Instead of building new data centres from scratch, Microsoft has turned to IREN, a company with large-scale GPU infrastructure, to secure Nvidia's GB300 GPUs at IREN’s Texas-based data centre. The site, which will support 750 megawatts of capacity, will roll out in phases through 2026, offering scalable AI compute resources to meet increasing demand.

Once focused primarily on bitcoin mining, IREN is shifting gears with this partnership. CEO Daniel Roberts of IREN acknowledged that this deal will significantly transform the company into a prominent player in the AI cloud services market. The deal is projected to bring IREN around $1.94 billion in annualised revenue while providing a substantial revenue stream from the growing AI cloud demand.

A Strategic Partnership with Industry Leaders

Microsoft’s partnership with IREN follows a trend where hyperscalers such as Amazon Web Services and Google Cloud are battling for control over the AI compute market. This deal highlights the increasingly critical need for advanced hardware to fuel the growing AI revolution, with companies like Microsoft securing strategic partnerships to scale AI infrastructure faster than ever.

The association with IREN is part of a trend in which hyperscalers like Amazon Web Services and Google Cloud are battling over the market of AI compute. The deal underscores the ever-increasing importance of advanced hardware to power the ever-increasing AI revolution, as companies such as Microsoft win strategic alliances to grow AI infrastructure at a rate previously unknown.

The agreement comes on the heels of Microsoft's deal with Nscale for 200,000 Nvidia GB300 GPUs and its launch of the first production cluster with Nvidia’s GB300 NVL72 systems last month. The company’s goal is clear: expand Azure’s AI services and keep up with AI-driven innovation.

The IREN-Dell Partnership

In order to supply the necessary infrastructure, IREN has separately inked a $5.8 billion deal with Dell Technologies to purchase the necessary Nvidia GPUs and hardware, cementing its role as a central player in the AI compute ecosystem.

This deal positions IREN to become a leader in the emerging “neocloud” market, a fast-evolving sector in which companies provide infrastructure tailored to next-gen technologies like AI and machine learning.

This multi-billion-dollar deal is just one example of Microsoft’s aggressive strategy to lock in high-performance AI infrastructure at a time when AI and machine learning models are consuming unprecedented amounts of compute resources. - Competition in the global chip shortage remains fierce, with the tech giants such as Google, Amazon, and Microsoft all competing to have a bigger piece of the AI infrastructure pie.

The future of AI on clouds such as Azure will greatly rely on the partnerships, as AI innovation is going to push the limits of computing power. It is evident that Microsoft is making certain that it has a robust compute pipeline to expand the latest generation of AI applications.