IREN Orders 50,000+ Nvidia B300s to Build 150,000-GPU AI Cloud Fleet

Overview

• IREN says the added GPUs will be installed in stages in the second half of 2026 at its Mackenzie, British Columbia, and Childress, Texas, data centers.
• Management projects the expanded fleet could support more than $3.7 billion in annualized AI cloud revenue by year-end 2026, a figure based on internal assumptions and not fully contracted.
• The company reports securing $9.3 billion over the past eight months and plans about $3.5 billion of additional capex tied to these orders, with payments structured post-shipment and procurement sequenced to commercial milestones.
• Shares jumped about 12.8% on Wednesday following the expansion news, then fell roughly 5% in pre-market trading Thursday after the filing for a potential at-the-market equity sale.
• Co-CEO Daniel Roberts said scaling to 150,000 GPUs positions IREN among the largest specialized AI cloud providers and that early procurement reduces time-to-compute in a supply-constrained market.