IREN (NASDAQ:IREN) announced a significant upward revision to its 2026 AI Cloud annualized recurring revenue (ARR) target, now projecting the figure will exceed $4 billion. The company attributed the boost to the signing of new multi-year agreements valued at a total of $2.8 billion. The news sent shares climbing 8.7% to $36.54 during premarket trading on Monday, as reported by The Wall Street Journal.
The premarket activity began at 4:00 a.m. EDT, ahead of the U.S. cash market open at 9:30 a.m. EDT. The stock had been under pressure recently, declining 18.3% over the past week, mirroring broader weakness in the AI infrastructure space, with peers like CoreWeave (NASDAQ:CRWV) and Nebius (NASDAQ:NBIS) also sliding 17.6% and 19.1%, respectively.
The updated ARR target represents an increase of at least $300 million, or just over 8%, compared to the previous forecast of $3.7 billion disclosed in May. However, the headline $2.8 billion contract value is spread across a four-year portfolio term, translating to approximately $700 million per year—a figure that notably exceeds the $300 million ARR uplift. The discrepancy suggests that some of the new contracts may involve staggered launches or delayed client onboarding, with ARR being an operational measure rather than a GAAP revenue recognition metric.
IREN also provided an updated snapshot of its financial position. As of June 30, the company held preliminary cash and equivalents of $7.6 billion, including $1.7 billion in restricted cash. Excluding the restricted portion, the estimated cash balance stands at $5.9 billion, a dramatic increase from $2.6 billion at the end of April. The firm had previously outlined $3.5 billion in capital expenditures for the second half of 2026 to support its 150,000-GPU initiative, with roughly 45% of related GPU expenses now covered by recent prepayments.
While the company disclosed names such as Microsoft (NASDAQ:MSFT), Nvidia (NASDAQ:NVDA), Perplexity, and Figure AI among its clients, it did not specify which of these were party to the $2.8 billion deal. Co-CEO Daniel Roberts emphasized that the platform is “scaling at pace,” with IREN aiming to deliver 480 megawatts of AI Cloud capacity this year and targeting 1.2 gigawatts by 2027.
Investors are now closely watching the broader market context. U.S. futures climbed on Monday, but chip shares continued to face pressure, with the semiconductor index closing Friday more than 20% below its June high. Major tech earnings are expected to be scrutinized for signs of AI spending discipline. Key risks remain centered on construction timelines, power availability, and GPU acceptance rates. Prepayment conditions may vary, and future agreements might not cover the same proportion of equipment costs.
For IREN, the critical test ahead lies in converting contract values into commissioned capacity and recognized revenue. While demand appears to be locked in through signed agreements, the delivery phase has yet to materialize fully.
