Australian AI cloud provider IREN has secured USD 3.65 billion in investment-grade financing to support delivery of its multi-billion-dollar AI infrastructure agreement with Microsoft, marking what the company claims is the highest-rated GPU-backed financing announced to date. The financing supports deployment of Nvidia GB300 GPU infrastructure under IREN’s previously announced USD 9.7 billion contract with Microsoft, which will see dedicated AI cloud capacity delivered from the company’s Horizon campus in Childress, Texas.
The transaction comprises a USD 2.1 billion private placement and a USD 1.55 billion delayed draw term loan, carrying a blended cost of debt of 6 percent. Fitch and DBRS assigned investment-grade ratings of A and A (low) respectively to the facility, which is secured against the GPUs and the associated contracted cash flows.
The financing package is designed to support delivery of GPU infrastructure that will be deployed under IREN’s long-term agreement with Microsoft. Announced in November, the deal will see Microsoft access dedicated Nvidia GB300 GPU capacity over a five-year period from IREN’s Horizon data centre complex in Texas. The campus is planned to comprise four facilities with a combined 200MW of IT load and forms part of the company’s broader 750MW Childress development. The agreement represented a significant milestone in IREN’s transition from cryptocurrency mining into AI infrastructure and cloud services.
Under the latest financing arrangements, IREN said the debt facility, together with customer prepayments from Microsoft, will fund approximately 96 percent of the USD 5.81 billion GPU capital expenditure required under the contract. The company said the blended financing structure results in an average funding cost of 3.31 percent when Microsoft’s prepayments are taken into account.
Goldman Sachs and J.P. Morgan acted as joint lead managers and arrangers for the transaction, which attracted participation from global financial institutions, asset managers and insurance investors.
Data centre ownership supports financing strategy
IREN co-founder and co-chief executive Daniel Roberts said the transaction demonstrated the value of combining contracted AI workloads with ownership of the underlying infrastructure. “Securing investment-grade financing on these terms reflects both the quality of our customer contracts and the fact that we own the data centre infrastructure these GPUs run in,” he said. “That combination broadens our access to institutional capital and lowers our cost of capital as we scale.”
The financing comes as IREN continues to expand its AI cloud business, which is built around a portfolio of grid-connected sites across North America. The company has previously highlighted its access to approximately 3GW of secured power capacity as a key differentiator as demand for AI infrastructure accelerates.
According to IREN, the latest financing strengthens its capital structure as it works towards delivering 480MW of AI cloud capacity by the end of 2026.
AI infrastructure attracts new financing models
The transaction also reflects the growing scale of capital required to deploy AI infrastructure and the emergence of new financing structures linked directly to GPU assets and long-term customer contracts. As hyperscale and AI deployments grow, operators are increasingly seeking funding models that sit alongside traditional corporate debt and project finance structures.
For IREN, the Microsoft contract provided the foundation for securing investment-grade ratings and accessing the US private placement market, creating what the company described as the first GPU financing of its kind in that market.