IREN has announced a major expansion of its AI Cloud infrastructure, doubling capacity to approximately 23,000 GPUs following the purchase of an additional 12,400 high-performance units. The investment, valued at around $674 million, positions the company to reach more than $500 million in annualized run-rate revenue from its AI Cloud services by the end of the first quarter of 2026.
The procurement includes 7,100 NVIDIA B300s, 4,200 NVIDIA B200s, and 1,100 AMD MI350Xs, reflecting the increasing market demand for both NVIDIA and AMD hardware.
Deliveries will be staged across the coming months at IREN’s Prince George campus in British Columbia, with the expanded fleet set to support ambitious revenue targets. The company’s GPU inventory now spans 1,900 NVIDIA H100s and H200s, 19,100 NVIDIA B200s and B300s, 1,200 NVIDIA GB300s, and 1,100 AMD MI350Xs.
By incorporating AMD’s latest chips alongside NVIDIA systems, IREN is broadening its portfolio and appealing to a wider base of enterprise and research clients.
IREN’s Canadian Campuses
This diversification comes as global demand for AI compute continues to outpace supply, leading more customers to contract capacity before systems are commissioned.
IREN’s Canadian campuses are designed to host more than 60,000 Blackwell GPUs, with additional expansion capacity in Texas.
The company is currently working on financing streams to support this scale-up. Although some Bitcoin mining capacity could be affected, IREN expects to mitigate the impact by redeploying ASIC miners to alternate sites.
The expansion underscores a broader industry shift toward pre-contracted compute, as businesses seek assurance of access to the infrastructure necessary to support AI workloads. With its vertically integrated platform and accelerated deployment capabilities, IREN is positioning itself as a key player in meeting this demand.
Discover more from WIREDGORILLA
Subscribe to get the latest posts sent to your email.
