IREN (IREN), a data center operator focusing on AI cloud infrastructure, said it agreed to purchase more than 50,000 specialized processing chips from Nvidia (NVDA) to expand production capacity by approximately 50%.
The B300 GPUs, or graphics processing units, will bring the Sydney-based company’s total AI computing fleet to about 150,000 GPUs. A GPU is a specialized chip used to perform massive parallel calculations, enabling high-speed training and running of artificial intelligence models.
The company has also filed for a potential market sale of up to $6 billion in shares as part of its broader capital management strategy. The stock fell 5% in premarket trading Thursday due to potential dilution.
The additional hardware is expected to be deployed in phases at the company’s air-cooled data centers in McKenzie, British Columbia, and Childress, Texas, through the second half of 2026. Once fully deployed, the expanded fleet is expected to support more than $3.7 billion in annual AI cloud revenue, placing IREN among the largest AI cloud infrastructure providers in the world.
IREN said it has secured approximately $9.3 billion in funding through customer advances, convertible notes, GPU leasing and financing arrangements in the past eight months, and expects to receive approximately $3.5 billion in additional capital expenditures for new GPU deployments in the second half of 2026.
