Key Takeaways
- IREN has raised its 2026 AI Cloud annualized run‑rate revenue (ARR) target from $3.7 billion to over $4 billion, driven by new multi‑year contracts with leading AI developers.
- Approximately 85 % of the projected $4 billion+ ARR is already under contract, representing roughly $2.8 billion in total contract value.
- The customer roster now includes hyperscalers (Microsoft), AI hardware leaders (NVIDIA), and a broad set of AI‑focused firms such as Perplexity, Figure AI, Together AI, Fluidstack, Fireworks AI, Fal AI, Hume AI, and a newly added leading AI developer.
- Recent contracts feature customer prepayments covering about 45 % of the associated GPU capital expenditures, lowering IREN’s net funding needs for those deployments.
- Across its portfolio, IREN’s contracts have a weighted‑average term of ~4 years, providing revenue visibility and stability.
- As of June 30 2026, IREN held approximately $7.6 billion in cash and cash equivalents, including $1.7 billion of restricted cash tied to the Microsoft GPU‑financing arrangement at Horizon 1‑4.
- The company’s self‑built AI Cloud capacity has expanded from ~3 MW a year ago to 480 MW slated for delivery in 2026, with a goal of 1.2 GW by 2027.
- IREN emphasizes a vertically integrated model—owning grid‑connected land and power in renewable‑rich regions across North America, Europe, and APAC—to deliver large‑scale data‑center compute for AI training and inference.
- Forward‑looking statements caution that achieving these targets depends on successful commissioning, customer acceptance, utilization, pricing trends, and continued execution of its growth strategy amid known risks and uncertainties.
IREN Limited (NASDAQ: IREN) announced on July 20, 2026 that it has revised its year‑end AI Cloud annualized run‑rate revenue (ARR) target upward from $3.7 billion to more than $4 billion. This increase reflects the company’s confidence in the scale and durability of its AI‑focused cloud services business. The new target is anchored in the expectation that roughly 480 MW (gross) of AI Cloud capacity will be commissioned by the end of 2026, based on internal assumptions about GPU models, contract terms, utilization rates, and pricing. ARR is calculated by multiplying the GPU/hour price for commissioned GPUs as of December 31, 2026 by the 8,760 hours in a year and then adding annualized revenue from storage and ancillary services. It is important to note that ARR is an operating metric, not a GAAP measure, and does not incorporate the timing or recognition rules that affect reported revenue.
A substantial portion of the projected ARR is already locked in. Following a series of new multi‑year cloud services agreements with top AI developers, approximately 85 % of the $4 billion+ ARR target is now under contract, amounting to roughly $2.8 billion in total contract value. These agreements span a diverse set of customers, including hyperscalers such as Microsoft, GPU architect NVIDIA, and a growing list of AI‑centric firms: Perplexity, Figure AI, Together AI, Fluidstack, Fireworks AI, Fal AI, Hume AI, and a newly added leading AI developer. IREN offers both bare‑metal and managed cloud services, allowing clients to choose the level of control and abstraction that best fits their workloads.
The company’s contracting strategy continues to emphasize selective capacity allocation ahead of commissioning, aiming to diversify its customer base and platform layers. Demand from hyperscalers, enterprises, AI developers, and frontier research labs remains in excess of IREN’s currently available and planned capacity, prompting ongoing discussions with customers across the full 2026‑2027 expansion pipeline. This strong demand environment has allowed IREN to negotiate more favorable pricing in recent contracts.
A notable feature of the newer agreements is the inclusion of customer prepayments that cover roughly 45 % of the estimated GPU capital expenditure for the associated deployments. These prepayments reduce IREN’s net funding requirement for those projects, improving cash flow efficiency and lowering the amount of external financing needed to build out the infrastructure. Prepayment terms vary by contract, and there is no guarantee that future agreements will mirror these conditions, but the current trend underscores the confidence that leading AI players have in IREN’s delivery capability.
Across its entire contract portfolio, the weighted‑average term is approximately four years, calculated by weighting each contract’s stated length by its contribution to ARR. This multi‑year horizon provides revenue visibility and helps smooth cash‑flow expectations as new capacity comes online. The longer‑term nature of the deals also aligns with the multi‑year nature of AI model training and inference workloads, which often require sustained access to high‑performance compute.
Financial strength underpins IREN’s aggressive expansion. As of June 30 2026, the company reported approximately $7.6 billion in cash and cash equivalents. This figure includes $1.7 billion of restricted cash earmarked for the GPU financing arrangement tied to the Microsoft contract at the Horizon 1‑4 sites. The robust liquidity position gives IREN flexibility to fund capital expenditures, pursue strategic acquisitions, or weather market volatility while continuing to roll out its data‑center footprint.
Leadership commentary highlighted the rapid scale‑up of IREN’s vertically integrated AI Cloud platform. Co‑Founder and Co‑CEO Daniel Roberts noted that the company’s self‑built AI Cloud capacity has grown from about 3 MW a year ago to 480 MW slated for delivery in 2026, with a target of 1.2 GW by 2027. This expansion is supported by IREN’s ownership of grid‑connected land and power assets in renewable‑rich regions across North America, Europe, and APAC, which enables the firm to secure low‑cost, sustainable energy for its high‑density compute facilities.
IREN’s business model centers on delivering large‑scale data‑center infrastructure purpose‑built for AI training and inference. By controlling the full stack—from land and power procurement to server rack design, direct‑to‑chip liquid cooling, and software‑defined orchestration—the company aims to offer performance, reliability, and cost advantages that appeal to AI developers needing massive, predictable compute resources.
Finally, the press release includes the customary forward‑looking statements disclaimer. IREN cautions that achieving its ARR and capacity targets depends on a range of factors, including successful commissioning and customer acceptance of GPUs, utilization and pricing trends, the ability to continue developing its data‑center sites, and broader market conditions. Known risks and uncertainties are detailed in IREN’s Form 10‑K and other SEC filings, and actual results could differ materially from the projections outlined herein.
Prepared for informational purposes only; does not constitute an offer or solicitation to buy or sell any securities.