IREN has placed the first 50-megawatt phase of its $9.7 billion Microsoft AI cloud contract into operation, giving the former Bitcoin miner an early operating milestone in its planned expansion into data-center infrastructure for artificial intelligence workloads.
The deployment, known as Horizon 1, is running at IREN’s Childress, Texas, campus, the company said. It is the first of four planned installations under the five-year Microsoft agreement signed in November 2025. The remaining three Horizon phases are scheduled for delivery later in 2026.
Horizon 1 uses direct-to-chip liquid cooling, a system designed to remove heat directly from high-performance processors rather than relying primarily on air cooling. IREN said the site has also received Nvidia Exemplar Cloud status, a designation linked to Nvidia’s performance and reliability requirements for AI cloud infrastructure.
The initial delivery gives IREN a live reference site for a much larger program that is due to reach 200 megawatts of IT capacity across four phases. Each deployment is expected to use Nvidia GB300 GPUs, hardware aimed at training and operating advanced AI models.
Microsoft contract moves from financing to operation
The Childress launch marks the transition of the Microsoft deal from an announced contract and financing plan into an operating deployment. IREN has tied much of its near-term strategy to building capacity for large enterprise AI customers, using power infrastructure originally developed around its data-center operations.
In June, IREN announced a $3.65 billion debt package to fund GPUs associated with the Microsoft contract. The company said the financing covered 96% of the $5.81 billion in GPU spending connected to the agreement.
Fitch rated the debt A, while DBRS assigned an A(low) rating, according to IREN. Those ratings reflected the financing structure’s backing by the Microsoft contract rather than a broad assessment of the economics of every part of IREN’s business.
The scale of the debt package also illustrates the capital-intensive nature of the pivot. Securing land, power connections and data-center buildings can establish a foundation for AI infrastructure, but the GPU fleets required by customers represent a major upfront cost. The Microsoft agreement gave IREN a contracted framework around which it could arrange financing for those purchases.
The planned 200-megawatt rollout is substantial even before IREN reaches its longer-term capacity targets. The company is aiming for 480 megawatts of AI cloud capacity in 2026 and 1.2 gigawatts by 2027. Delivering Horizon 2 through Horizon 4 on schedule would be central to meeting the first of those goals.
Childress campus becomes a test of execution
The Horizon 1 deployment puts IREN’s Childress campus at the center of its AI cloud buildout. Data centers serving dense GPU clusters require reliable electricity, cooling systems and networking infrastructure capable of operating continuously under heavy workloads.
Liquid cooling has become increasingly relevant as AI processors consume more power and generate more heat than earlier generations of data-center equipment. Direct-to-chip systems circulate coolant close to the computing hardware, helping operators manage thermal loads that could be difficult or inefficient to handle with conventional air cooling alone.
IREN’s reported Nvidia Exemplar Cloud status may help demonstrate that the initial site meets hardware maker standards expected by customers running demanding AI applications. Nvidia’s technology has become the core computing platform for much of the commercial AI sector, so certification-related recognition can carry practical value for a cloud operator seeking to host Nvidia-based workloads.
The designation does not by itself guarantee that future deployments will be delivered on time or at the same operating profile. The next three phases will require IREN to repeat the construction, equipment installation and commissioning process at a far larger cumulative scale. Yet the completion of Horizon 1 reduces one immediate uncertainty: whether the company can bring the first contracted deployment online at its Texas campus.
Bitcoin mining remains part of IREN’s backdrop
IREN built its public profile through Bitcoin mining, an industry where operators compete for block rewards using specialized computing equipment and large volumes of electricity. The company’s strategy increasingly places greater emphasis on converting its power and data-center assets into AI cloud capacity.
That transition does not mean IREN’s mining origins are irrelevant. Its existing experience in managing high-power computing facilities, along with its access to energy infrastructure, provides a starting point for the AI buildout. The hardware, customer arrangements and revenue model are different, though.
Bitcoin mining depends on the economics of network rewards, transaction fees, Bitcoin’s market price and power costs. AI cloud contracts can provide longer-duration revenue visibility when they are secured with major customers, while requiring expensive GPU purchases and strict operational standards.
IREN’s Microsoft agreement therefore gives the company a clearer commercial anchor than a pure mining operation, but it also concentrates execution around a small number of large installations and a major customer relationship. The 50-megawatt Horizon 1 launch is the first practical measure of that strategy.
Shares rise modestly after deployment update
IREN shares rose 2.6% to $45.22 on Monday afternoon after trading as high as $46.50 earlier in the session. The stock was up less than 6% year to date, according to the market figures provided by the company update.
The muted year-to-date gain, despite the size of the Microsoft contract and associated financing, suggests traders remain focused on the cost and operational challenge of turning announced AI capacity into functioning, revenue-producing infrastructure. The company’s targets imply a rapid expansion from the 50 megawatts now operating at Horizon 1.
The next milestones will be tangible: delivery of the remaining 150 megawatts under the Microsoft program, installation of Nvidia GB300 systems, and progress toward IREN’s stated 480-megawatt AI cloud capacity target for 2026. Horizon 1 has established the first operating foothold; the investment case now rests heavily on whether IREN can replicate it three more times this year.
Explore how institutional demand shapes AI and crypto—discover future market impacts in our AI–blockchain insights guide.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
