The global surge in artificial intelligence (AI) adoption is fueling unprecedented demand for high-performance computing infrastructure, and IREN Limited (NASDAQ: IREN) is positioning itself at the forefront of this transformation. The company has announced a major expansion of its AI compute capabilities with the acquisition of 2,400 next-generation NVIDIA Blackwell B200 and B300 GPUs—a strategic move that underscores its commitment to meeting rising demand from AI-native enterprises, cloud operators, and hyperscalers.
This significant hardware investment will be deployed across IREN’s data center campus in Prince George, British Columbia, where it will join an existing fleet of 1,900 Hopper architecture GPUs. With this addition, IREN’s total NVIDIA GPU count reaches approximately 4,300 units, creating one of the most powerful AI-ready infrastructures in North America.
Strategic Expansion in a Supply-Constrained Market
The new order consists of 1,300 B200 and 1,100 B300 GPUs, which will be installed over the coming months. These advanced processors are designed for large-scale AI training and inference workloads, offering substantial improvements in performance and energy efficiency compared to previous generations. The deployment is fully funded through existing cash reserves, though IREN is also exploring additional financing options to support continued growth in its AI Cloud Services segment.
Prince George’s facility, backed by 50MW of dedicated power capacity, has been engineered for scalability. It can ultimately accommodate more than 20,000 air-cooled Blackwell GPUs, assuming a power usage effectiveness (PUE) of 1.1 and a per-GPU draw of 1.93kW—including ancillary systems like cooling and networking. This provides IREN with a clear runway for phased, demand-driven expansion.
Any legacy mining equipment displaced during the upgrade will be redeployed to other operational sites, ensuring optimal asset utilization across the company’s broader infrastructure network.
👉 Discover how next-gen GPU infrastructure is reshaping AI cloud services.
Next-Gen Performance Meets Real-World Demand
NVIDIA’s Blackwell architecture represents a leap forward in AI compute, delivering breakthrough capabilities in throughput, memory bandwidth, and power efficiency. For IREN, this means the ability to offer customers cutting-edge resources for training massive language models, running real-time inference pipelines, and scaling complex machine learning applications.
As AI workloads grow in complexity and volume, demand from enterprises seeking reliable, high-density compute environments continues to intensify. IREN’s early access to Blackwell GPUs places it among the first cloud providers capable of delivering next-generation AI infrastructure at scale—a competitive advantage in a market defined by scarcity and speed.
Daniel Roberts, Co-CEO of IREN, emphasized the dual value of the investment:
“This investment is compelling on two fronts: strong near-term economics and meaningful strategic positioning. Expanding our Blackwell footprint not only enables us to meet current customer demand, it also supports the broader infrastructure conversations driving our next phase of growth.”
Vertically Integrated Platform for End-to-End AI Solutions
What sets IREN apart is its vertically integrated model, supported by 2,910 MW of grid-connected power across strategically located sites in the U.S. and Canada. This foundation allows the company to deliver flexible, end-to-end solutions across the AI infrastructure stack, including:
- Powered shell deployments
- Build-to-suit data centers
- Turnkey colocation services
- Fully managed AI cloud computing
By controlling everything from land acquisition and power sourcing to facility design and operations, IREN reduces complexity for clients while accelerating deployment timelines—critical factors in today’s fast-moving AI landscape.
The company’s AI Data Center division is already scheduled to deliver up to 50MW of liquid-cooled capacity by 2025, further enhancing thermal efficiency and supporting even higher compute densities.
Core Capabilities Driving Long-Term Growth
IREN operates at the intersection of three high-growth verticals:
1. Bitcoin Mining
As one of the world’s largest and lowest-cost Bitcoin producers, IREN maintains an installed self-mining capacity of 50 exahashes per second (EH/s). This provides stable cash flow that helps fund innovation in its AI divisions.
2. AI Cloud Services
Leveraging state-of-the-art NVIDIA GPUs, IREN delivers high-performance cloud compute tailored for AI developers and enterprises. The addition of Blackwell units significantly boosts available processing power for both training and inference tasks.
3. AI Data Centers
From concept to operation, IREN designs and manages data centers purpose-built for AI workloads. With modular designs and scalable power delivery, these facilities are optimized for performance, sustainability, and rapid deployment.
👉 See how integrated data center platforms are powering the future of AI.
Keywords Driving Industry Relevance
To align with search intent and enhance discoverability, key terms naturally integrated throughout this article include:
AI computing, NVIDIA Blackwell GPUs, AI cloud services, data center expansion, GPU fleet upgrade, AI infrastructure, high-performance computing, and vertically integrated data centers.
These keywords reflect core industry trends and user queries related to scalable AI infrastructure, helping position this content for visibility among technical decision-makers, investors, and technology analysts.
Frequently Asked Questions
Q: What are NVIDIA Blackwell B200 and B300 GPUs used for?
A: These next-generation GPUs are optimized for large-scale AI training and inference, enabling faster processing of complex models such as generative AI and large language models.
Q: Where will the new GPUs be installed?
A: The 2,400 GPUs will be deployed at IREN’s Prince George, BC campus—a facility designed for high-density computing with 50MW of available power.
Q: Is IREN shifting away from Bitcoin mining?
A: No. While expanding into AI, IREN continues to operate one of the largest Bitcoin mining fleets globally. Legacy hardware is being redeployed, not retired.
Q: How does IREN support sustainable AI growth?
A: By locating facilities in renewable-rich regions and utilizing efficient cooling technologies—including upcoming liquid-cooled deployments—IREN minimizes environmental impact.
Q: Can enterprises lease GPU capacity from IREN?
A: Yes. Through its AI Cloud Services and colocation offerings, IREN provides flexible access to high-performance compute resources on a managed or self-service basis.
Q: What makes IREN’s approach different from traditional cloud providers?
A: Its vertically integrated model—controlling power, land, construction, and operations—allows for faster deployment, lower costs, and customized solutions tailored to AI workloads.
Future Outlook and Scalable Ambitions
With over 2 gigawatts of additional development pipeline, IREN is well-positioned to scale beyond its current footprint. The company’s focus on integrating renewable energy, advanced cooling systems like direct-to-chip liquid cooling, and modular data center designs ensures long-term adaptability in an evolving technological landscape.
As AI adoption accelerates across industries—from healthcare and finance to autonomous systems and content creation—the need for powerful, reliable infrastructure will only grow. IREN’s proactive expansion into next-gen GPU compute ensures it remains a key enabler of this transformation.
👉 Explore how scalable GPU infrastructure powers the future of artificial intelligence.
While forward-looking statements involve inherent risks—such as execution challenges, market shifts, or supply chain constraints—IREN’s combination of strategic assets, technical expertise, and financial discipline positions it to navigate uncertainty and capture long-term value in the AI era.