IREN Q4 Earnings Call Highlights

AI cloud growth accelerated: IREN ended fiscal 2026 with $4 billion in contracted annualized recurring revenue, including $1 billion from Microsoft’s accepted Horizon 1 deployment. The company said 2026 capacity is largely sold out and is already negotiating substantial portions of its 2027 and 2028 capacity. Major capacity expansion is underway: IREN targets approximately 300 megawatts of IT load delivered in 2026 and another 500 megawatts in 2027, reaching about 1.2 gigawatts of platform capacity. Horizon 2 is nearing commissioning, while Horizons 3 and 4 are under late-stage construction. Expansion requires substantial financing: IREN secured about $19 billion in funding over the past year and expects fiscal 2027 capital expenditures of $25 billion to $30 billion, including roughly $8 billion of additional GPU financing and prepayments. June-quarter revenue was $137.2 million, while a $684 million net loss was driven largely by non-cash impairments tied to decommissioned mining hardware. Riot Platforms Re-Wires the Ledger for a $9B AI Power Play IREN NASDAQ: IREN said it ended fiscal 2026 with $4 billion in contracted annualized recurring revenue, or ARR, for its 2026 capacity, including $1 billion that was operating following Microsoft’s acceptance of the Horizon 1 deployment. Co-Founder and Co-CEO Daniel Roberts said the company’s AI cloud strategy is centered on owning the full infrastructure stack, including land, power, data centers, compute equipment and software services. He said AI demand has intensified while physical infrastructure remains difficult and time-consuming to develop. “Signing deals is not the bottleneck in this market,” Roberts said. “Bringing GPUs online is.” The company said its 2026 capacity is largely sold out and that it is in late-stage discussions with prospective customers for a significant portion of 2027 capacity, while discussions for 2028 are also underway. IREN disclosed new multicloud contracts with Cohere, Prometheus, Perplexity, Figure AI, Fal.ai and Higgsfield AI, as well as a separate multiyear agreement with an unnamed frontier AI lab. Roberts said existing customers Together AI and Fireworks AI had renewed and expanded their relationships with IREN. The company is seeking to broaden its customer base across hyperscalers, enterprises, AI developers and frontier labs, while also adding managed-services capabilities through Mirantis. IREN said Horizon 1, the first of four 50-megawatt liquid-cooled deployments at its Childress, Texas, site, was delivered to Microsoft during the month. The deployment achieved NVIDIA Exemplar Cloud status on GB200 NVL72 systems, according to Roberts. Horizon 2 was progressing toward commissioning, while Horizons 3 and 4 were in late construction. The company is targeting delivery of all three remaining phases during the December quarter. The company is targeting about 300 megawatts of IT load delivered in 2026 and another 500 megawatts in 2027, which it said would bring gross platform capacity to about 1.2 gigawatts in 2027. IREN is developing capacity across Texas, British Columbia, Oklahoma, South Australia and Spain. At its British Columbia sites, IREN said GPUs are being installed at Mackenzie, Prince George’s air-cooled fleet is fully commissioned and liquid-cooling installation is underway. The company also decided to convert the Canal Flats site to liquid cooling for GB300 systems. For 2027, Sweetwater 1 is under construction, while civil work has begun for Horizons 5 and 6 at Childress. IREN said.