Pre-Market Surges More Than 9%: IREN Secures $2.8 Billion AI Contract, Microsoft and Nvidia Lead Elite Customer Lineup

As generative AI continues to develop rapidly, global AI developers' demand for GPU computing power and cloud computing resources remains robust, and IREN has welcomed a new round of order growth as a result.

At present, the company's client base already covers Microsoft ( MSFT ), Nvidia ( NVDA ), Perplexity, Figure AI, Together AI, Fluidstack, Fireworks AI, Fal AI, Hume AI, and an undisclosed major AI development enterprise, with businesses covering various areas such as bare-metal GPU server leasing and hosted AI cloud platforms.

Compared to one-off hardware sales, these multi-year cloud service contracts can bring more stable cash flows. IREN disclosed that the weighted average term of its current customer contracts is approximately four years, providing high predictability for future revenue growth.

Management stated that current demand from hyperscalers, large enterprises, AI model developers, and frontier AI labs remains significantly higher than the company's existing and planned computing power supply. The company has engaged in in-depth discussions with customers regarding all of its expansion plans for 2026 to 2027, indicating that its future order pipeline remains robust.

Compared to just over a year ago, IREN's business model has undergone a fundamental shift.

Originally starting as a Bitcoin mining operation, the company's management quickly pivoted its strategy as demand for AI compute exploded in recent years, progressively transitioning its existing data centers, power resources, and infrastructure toward AI GPU cluster deployment, officially entering the AI cloud computing market.

Leveraging its existing power and data center resources, IREN, as a relatively new entrant, enjoys faster construction speeds and lower deployment costs, thereby attracting ongoing collaborations with major clients such as Microsoft and Nvidia.

Meanwhile, IREN has recently continued to strengthen its management team, bringing in senior executives across product, infrastructure, and information security to support the rapid expansion of its AI business and prepare for larger-scale data center operations in the future.

Daniel Roberts, co-founder and co-CEO, stated that over the past year alone, the company's proprietary AI cloud compute capacity has rapidly expanded from approximately 3 megawatts (MW) to the 480 MW scheduled for delivery this year, with plans to scale this further to 1.2 gigawatts (GW) in the future, aiming to complete the deployment by 2027.

This content was translated using AI and reviewed for clarity. It is for informational purposes only.