Hut 8 has signed a second 15-year lease worth $9.8 billion for its Beacon Point campus in Nueces County, Texas, completing commercialization of the 1-gigawatt site and deepening the company’s shift from Bitcoin mining toward artificial intelligence data center hosting.
The latest agreement doubles Hut 8’s contracted capacity with the same high-credit tenant to 704 megawatts and lifts the total base-term contract value at Beacon Point to $19.6 billion, the company said. The deal follows a first lease at the Texas campus in May, also valued at $9.8 billion and covering 352 megawatts of IT capacity.
Across its AI data center portfolio, Hut 8 now has 949 megawatts of contracted capacity. That includes 704 megawatts at Beacon Point and 245 megawatts at River Bend. The company said those facilities carry a combined base-term contract value of $26.6 billion and are expected to generate average annual net operating income of more than $1.75 billion.
The announcement was followed by a sharp market reaction. Hut 8 shares rose more than 16% after the company disclosed the second Beacon Point lease, reflecting stronger trader interest in former cryptocurrency mining companies that control large power supplies and can redeploy them into AI computing infrastructure.
IREN, another major operator with roots in Bitcoin mining, also moved higher after raising its AI Cloud revenue outlook. The company increased its year-end 2026 AI Cloud annualized revenue target to more than $4 billion, up from $3.7 billion, after securing $2.8 billion in new multi-year contracts. IREN shares climbed more than 17% following the update.
Together, the announcements underline a broader change across the sector. Companies that once depended heavily on Bitcoin block rewards are increasingly marketing power access, land, cooling systems, and data center construction skills to large technology customers that need computing capacity for AI training and inference.
The shift is being closely watched by traders because it changes the revenue profile of these companies. Bitcoin mining revenue can fluctuate with network difficulty, energy prices, transaction fees, and token prices. Long-term AI hosting leases, by contrast, can provide contracted cash flow over many years, although they also require large upfront spending on power infrastructure, buildings, networking, cooling systems, and high-performance computing equipment.
Beacon Point reaches full commercialization
Hut 8’s second Beacon Point lease matches the structure of the first agreement signed in May. Each lease is valued at $9.8 billion over a 15-year base term and covers 352 megawatts of IT capacity. With both agreements in place, the Texas campus is now fully commercialized at 704 megawatts of contracted IT capacity.
The leases also include three five-year renewal options. If those options are exercised, Hut 8 said the total potential value of Beacon Point could rise to as much as $50.2 billion.
Beacon Point was developed using what Hut 8 describes as a power-first strategy. Before commercializing the site, the company secured 1,000 megawatts of utility capacity through an agreement with AEP Texas. That early control of power supply gave the company a foundation to approach major AI and cloud customers at a time when access to large-scale electricity has become one of the biggest bottlenecks in data center development.
In the current AI buildout cycle, power availability can be as important as land or capital. Large AI campuses require hundreds of megawatts of reliable electricity, and in many regions grid interconnection timelines can stretch for years. Companies that already hold utility agreements or can move projects through permitting and grid planning quickly may have an advantage when negotiating with technology customers.
Hut 8’s model at Beacon Point reflects that market reality. Rather than build only for its own computing operations, the company is leasing capacity under long-term agreements to a tenant with strong credit standing. That structure can reduce exposure to short-term swings in Bitcoin mining economics while positioning the company as a landlord and infrastructure partner for AI workloads.
A larger contracted AI portfolio
With the Beacon Point expansion, Hut 8’s AI-related contracted capacity now totals 949 megawatts across two campuses. The larger portion sits at Beacon Point, while River Bend accounts for 245 megawatts.
The company’s combined base-term contract value across the two campuses is now $26.6 billion. Hut 8 has projected average annual net operating income above $1.75 billion from the contracted facilities, a figure that highlights the scale of the company’s planned transition.
Net operating income is a key measure for infrastructure assets because it focuses on property-level earnings before corporate overhead, financing costs, taxes, and some non-cash items. For data center developers and operators, it can help show how much recurring income a site may generate once it is fully operational and leased.
The company is not alone in pursuing this model. The boom in generative AI has created intense demand for large data center campuses that can host graphics processing units, specialized AI chips, storage, and networking equipment. The result has been a rush among power-rich operators to reposition assets that were previously used for cryptocurrency mining or other high-energy computing tasks.
Bitcoin miners are especially active in this shift because their businesses already involve large power contracts, technical teams familiar with high-density equipment, and relationships with utilities. However, AI data centers often require more complex facilities than traditional mining sites, including higher reliability standards, advanced cooling, network redundancy, and tighter customer service requirements.
That difference means the transition can be lucrative but not simple. Operators must prove they can deliver facilities on time, meet uptime commitments, and manage demanding customers. They may also need to finance major construction programs before revenue fully begins.
IREN raises AI cloud target
IREN separately raised its year-end 2026 AI Cloud annualized revenue target to more than $4 billion after signing $2.8 billion in new multi-year contracts. In filings with the U.S. Securities and Exchange Commission, the company said about 85% of the revised target is already under contract.
The new agreements include prepayments covering around 45% of related GPU expenditures. That is important because GPUs are among the largest cost items in AI cloud deployments. Customer prepayments can help reduce how much funding a company needs to provide on its own, limiting pressure on cash balances and lowering the need for additional borrowing or equity issuance.
IREN reported $7.6 billion in cash and equivalents as of June 30. That liquidity, combined with customer prepayments, gives the company greater flexibility as it expands cloud capacity.
The company said its customer relationships now include Microsoft, NVIDIA, and several AI developers across both bare metal and managed cloud services. Bare metal services generally give customers direct access to dedicated hardware, while managed cloud services include a broader range of operational support and software-layer management.
IREN’s co-chief executive said the firm has expanded from roughly 3 megawatts of in-house AI Cloud capacity to 480 megawatts this year and is targeting 1.2 gigawatts by 2027. That growth plan places the company among the most aggressive former mining-linked operators moving into AI infrastructure.
The pace of expansion also reflects how quickly the AI cloud market is evolving. Demand for GPU clusters has remained elevated as software companies, cloud platforms, research labs, and enterprise customers compete for access to computing power. Some customers are signing multi-year agreements to secure capacity in advance, especially when deployments involve the newest hardware.
Shares move as traders reprice the sector
The latest announcements triggered strong moves in both stocks. Hut 8 rose more than 16% after confirming the second Beacon Point lease, while IREN gained more than 17% after lifting its AI Cloud revenue target.
Stock trading data available on July 20, 2026, showed a much larger longer-term move in the sector. Hut 8 shares were up about 323% over the prior year, while IREN had risen about 87% over the same period. Hut 8’s equity market value reached roughly $10.3 billion after the facility announcements, according to the data cited.
Those gains indicate that traders have been assigning higher value to companies able to convert power portfolios into contracted AI infrastructure revenue. The market response also suggests that traders are distinguishing between traditional cryptocurrency mining exposure and businesses with long-duration hosting contracts.
Still, the transition carries execution risks. Large data center campuses require complex construction schedules, major power delivery work, specialized cooling systems, and reliable supply chains for electrical equipment. Any delay in transformers, switchgear, substations, or utility interconnections can affect project timelines. For AI cloud operators, access to GPUs can also influence how quickly revenue is recognized.
There are commercial risks as well. Long-term leases and cloud contracts depend on customer credit quality, delivery milestones, service levels, and the durability of AI demand. While current demand for AI compute remains strong, the sector is capital intensive and competitive. Technology standards can change quickly, and customers may seek lower prices over time as more capacity comes online.
Power access becomes the main asset
The common factor in both Hut 8’s and IREN’s announcements is control of large-scale power. AI infrastructure growth is increasingly constrained by access to electricity, not only by access to chips or land.
For operators with existing power agreements, the opportunity is clear. A site that once supported Bitcoin mining can potentially be upgraded or redeveloped for AI hosting, producing more predictable contracted revenue. The economics are not identical, and the technical requirements are higher, but the underlying asset — large amounts of power at scale — has become more valuable.
Hut 8’s Beacon Point campus shows how that strategy can work when power is secured before customer commercialization. The company locked in 1,000 megawatts of utility capacity with AEP Texas, then signed long-term leases for most of the site’s IT load with a high-credit tenant.
IREN’s model is somewhat different, with a stronger emphasis on AI Cloud services and GPU-backed customer contracts. But the same theme applies: power access, data center development capability, and customer commitments are being combined into a new business line that is less dependent on cryptocurrency market cycles.
For traders, the question is no longer only how much Bitcoin these companies can mine. Increasingly, the focus is how much power they control, how quickly they can turn that power into operational AI capacity, how much revenue is already contracted, and how much capital is needed before projects generate cash.
The latest deals from Hut 8 and IREN show that the market for AI infrastructure remains active and that former cryptocurrency mining companies are becoming more central to the buildout. Their next challenge is execution: delivering massive campuses, managing customer obligations, and proving that contracted AI revenue can translate into durable earnings over the full life of the agreements.
Explore how AI and blockchain intersect in our guide to AI complementing blockchain and what it means for future data infrastructure.
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.
