IREN (NASDAQ: IREN) saw its shares edge lower in premarket trading Monday, slipping approximately 0.7% to $63.54, after the data-center operator announced the closure of a $3.65 billion investment-grade GPU financing facility. The funding is earmarked to support the company's delivery obligations under its high-profile Microsoft AI cloud contract, signaling a key step in bridging AI ambition with hardware execution.
The financing structure, detailed in an SEC filing, involves a subsidiary named IE US Hardware 3 LLC. It comprises a roughly $1.5 billion delayed-draw term loan—available to be drawn over time—and $2.1 billion in 5.96% senior notes due December 31, 2031. Borrowings under the loan carry an interest rate of term SOFR plus 2.25%, with SOFR serving as a short-term dollar benchmark. The debt is secured by graphics processing units (GPUs) and cash flows from the Microsoft contract, and it requires a minimum debt service coverage ratio of 1.05 to 1, ensuring a cash cushion for debt payments.
Investment-grade status is a critical distinction IREN is keen to highlight. The transaction received an 'A' rating from Fitch and an 'A(low)' from DBRS. Daniel Roberts, co-founder and co-CEO of IREN, emphasized that the favorable terms reflect the quality of the company's customer contracts and its ownership of the infrastructure where the GPUs will be deployed. This milestone comes as the market increasingly scrutinizes which AI infrastructure players can not only secure contracts but also fund and deliver the expensive, power-hungry hardware required.
The Microsoft contract serves as the anchor for this financing. Reuters reported in November that the software giant signed a five-year, $9.7 billion deal with IREN for access to Nvidia chips, providing Microsoft with capacity without building new data centers or securing additional power. The report also highlighted CoreWeave and Nebius as peers in the 'neocloud' space—companies that sell cloud computing built on Nvidia processors.
Last week's Dell order connects Monday's financing to physical deployment. On May 26, Reuters reported that IREN agreed to purchase Nvidia air-cooled Blackwell systems from Dell for approximately $1.6 billion. These systems are slated for installation at IREN's Childress, Texas campus and are expected to be operational by early 2027. Once commissioned, IREN projects its annualized run-rate revenue (ARR) will climb to $4.4 billion from $3.7 billion. Roberts noted that 'time-to-compute is everything' in this competitive landscape.
Execution precision is also a focus. BE Networks announced Monday that it is collaborating with IREN to use NVIDIA DSX Air to simulate and validate network architecture for over 50,000 Blackwell Ultra GPUs before physical deployment. Denis Skrinnikoff, IREN's chief technology officer, stressed that AI cloud infrastructure at such scale demands 'extreme precision.'
Peer performance was mixed in early trading: CoreWeave rose about 2.5%, Nebius gained roughly 2.0%, while Applied Digital fell around 4.8%. This divergence underscores that the AI-infrastructure trade is not monolithic; funding costs, contract backing, and delivery schedules are increasingly separating winners from laggards. Broader market conditions provided a supportive backdrop, with U.S. stock index futures climbing Monday as AI headlines around Nvidia and Microsoft helped offset concerns over Middle East tensions. Nasdaq 100 e-minis were up 0.32% at 06:48 ET.
However, risks remain. IREN's $4.4 billion ARR target hinges on assumptions about GPU models, utilization rates, and pricing. The target is not fully contracted and depends on on-time delivery and commissioning. Any delay in deliveries, weaker pricing for AI compute, or a mismatch between financing costs and utilization could undermine the narrative, even with Microsoft as the anchor customer. The market is now testing which companies can truly deliver on their AI promises.
