Inside IREN's $3.6B GPU Financing: How a Former Bitcoin Miner Is Funding a Massive Pivot to AI Infrastructure

A critical feature of this deal is its financing component. Microsoft agreed to a 20% prepayment for each tranche of capacity, providing IREN with a substantial upfront cash infusion of about $1.9 billion . This structure helps de-risk the hardware purchases, which IREN arranged separately through a $5.8 billion procurement agreement with Dell Technologies .

The Financing Engine: The $3.6 Billion GPU-Backed Loan

To cover the remaining hardware costs, IREN closed the gap with a dedicated financing facility. In its Q2 FY26 results on February 5, 2026, the company announced it had secured $3.6 billion in GPU financing directly tied to the Microsoft contract .

Key disclosed terms include:

• Interest Rate: Below 6% per annum, with management indicating an average cost of about 3% on earnings calls .
• Capital Coverage: This $3.6 billion, combined with the $1.9 billion Microsoft prepayment, covers approximately 95% of all GPU-related capital expenditures for the contract .
• Mechanism: The financing is structured as an asset-backed or project-finance-style loan, using the future revenue stream and the high-value GPUs themselves as a foundation for the borrowing .

This aggressive capital structure means IREN needs very little of its own equity capital to fund the hardware for this deal. However, it also means the company must execute flawlessly and begin delivering revenue to service this debt.

The Bigger Picture: A $9.3 Billion Total Financing Package

The $3.6 billion facility is just one part of a broader $9.3 billion capital raise IREN assembled over roughly eight months. The company used a mix of financing instruments, including :

• $1.0 billion in zero-coupon convertible notes issued in October 2025 .
• $400 million in other incremental GPU financing .
• At-the-market (ATM) equity programs boosted to $6 billion with major banks .
• Customer prepayments and operating cash flow.

Management earmarked roughly $3.5 billion of this capital for additional GPU orders beyond the immediate Microsoft scope, aiming for a massive fleet expansion .

The Output: 140,000 GPUs and Billions in Revenue

All this financing serves a singular goal: rapid deployment. IREN is targeting a fleet of 140,000 GPUs by the end of calendar year 2026 . Management projects that this fleet will generate approximately $3.4 billion in annualized recurring revenue (ARR), with about $1.94 billion coming from the Microsoft contract alone . By Q3 FY26, the company had already contracted $3.1 billion in ARR and raised its internal target to $3.7 billion .

Rounding Out the Portfolio: The NVIDIA Partnership

IREN's ambitions extend beyond its work for Microsoft. On May 7, 2026, the company announced a separate $3.4 billion, five-year AI Cloud contract with NVIDIA to provide managed GPU cloud services using air-cooled Blackwell GPUs .

This deal came with a larger strategic partnership. IREN issued NVIDIA a five-year right to purchase up to 30 million IREN shares at an exercise price of $70 each, a potential equity investment worth about $2.1 billion at the time . The partnership aims to support the deployment of up to 5 gigawatts of AI infrastructure over time .

The Build-Now, Pay-Later Gamble

IREN's approach is an extreme version of a wider industry trend. Data center operators are increasingly using long-term customer contracts and the immense value of AI chips themselves as collateral to secure massive loans, allowing them to build infrastructure now and pay for it as revenue comes online .

For IREN, the benefits are clear: a rapid pivot from Bitcoin mining to AI cloud services backed by two of the world's most valuable technology companies. The risk, however, is substantial and entirely centered on execution. The company must procure and deploy over 100,000 cutting-edge GPUs on a tight timeline, bring data centers online, and manage a mountain of new debt and contractual obligations well before the full revenue stream begins to flow . A miss on the deployment schedule would put immediate pressure on a highly leveraged balance sheet.