Hut 8 Advances a Gigawatt-Scale Campus Near Corpus Christi
Another significant development came November 4, when Hut 8 reported continued progress on its Beacon Point AI data center campus in the Corpus Christi region.
The company said that its previously announced second 352 MW IT-capacity lease had fully commercialized the campus, bringing the total contracted IT capacity associated with the two phases to 704 MW. The campus is designed around approximately 1 GW of utility capacity, a separate measure from the critical IT load under lease. Hut 8 estimates approximately $19.6 billion in base-term contract value across the two phases.
In its third-quarter results, the company said construction of the campus substation had begun, alongside foundation and underground utility work for the first phase. Initial data hall delivery is targeted for the third quarter of 2027, with the second phase targeted for the second quarter of 2028.
The company also disclosed agreements to acquire two energized sites in West Texas, with up to 230 MW of utility capacity at full energization, subject to outstanding approvals and infrastructure development. Those acquisitions point to another important trend: the increasing value of existing electrical infrastructure and sites that can potentially be adapted to serve advanced computing.
As new greenfield projects face lengthy power studies and regulatory requirements, already energized locations may offer an alternative route to capacity. The distinction is especially important in markets such as Dallas-Fort Worth, where DCF’s analysis of CBRE’s first-half 2026 data found that approximately 95% of the region’s 765 MW under construction had already been preleased.
PowerHouse and Big Watt Win Approval for Taylor Technology Campus
Not every large new project is meeting resistance at the municipal level. On October 8, the City of Taylor approved a development agreement with PowerHouse Data Centers and Big Watt Digital for the proposed Taylor Technology Campus northeast of Austin.
The 664-acre project is designed to support as much as 1 GW of data center capacity, incorporating dedicated behind-the-meter generation, including a planned natural gas power facility. The agreement represents an alternative to development on unincorporated land with comparatively limited local zoning oversight.
According to the City of Taylor’s project documentation, the agreement establishes a path for annexation and binding requirements addressing environmental and quality-of-life concerns.
That distinction is significant. The Taylor project illustrates how local government oversight and continued development can coexist, provided developers and municipalities reach agreement on applicable standards. It does not remove the need for other permits, financing, generation infrastructure, or construction milestones.
For PowerHouse, a division of American Real Estate Partners, the development also extends a large national pipeline of potential AI infrastructure sites.
As DCF recently examined in its coverage of Brookfield’s investment in AREP, the transaction reflects the growing strategic value of land development combined with access to electricity and the ability to navigate increasingly complex approvals.
Soluna and Applied Digital Pursue Wind-Connected AI Infrastructure
Another approach to securing electricity is emerging in the Texas Panhandle. On October 21, Soluna Holdings and Applied Digital announced an agreement to form a joint venture for a proposed 50 MW critical IT facility at Soluna’s Project Dorothy 3 site near Silverton in Briscoe County.
The development would be co-located with the 150 MW Briscoe Wind Farm, with its initial phase drawing electricity from both wind generation and the grid. Under the proposed ownership structure, Applied Digital would hold 51% of the venture and Soluna 49%, following closing. ChronoScale, a publicly traded accelerated-computing company majority-owned by Applied Digital, has signed a nonbinding letter of intent to lease the planned facility.
Importantly, the transaction is not yet fully closed. The companies’ October 21 announcement identifies several remaining conditions, including Briscoe County approval of a project tax abatement, definitive lease agreements, power arrangements, and corporate approvals.
That tax-abatement condition takes on added significance as state leaders debate the future of incentives for data center development.
The project illustrates the attraction of pairing AI infrastructure with existing energy assets, while its outstanding conditions also demonstrate how even promising developments depend on a sequence of commercial and regulatory approvals. It is part of the broader shift toward hybrid power strategies that DCF examined in its recent Power First report from the 2026 Data Center Frontier Trends Summit.
Fermi’s Proposed 222 MW Lease Misses Its Closing Deadline
The challenge of converting development plans into binding commercial commitments is especially visible in West Texas.
On November 3, Fermi America disclosed that its proposed lease agreement with TensorWave had not become effective after the October 31 closing deadline passed without extension. The transaction covers a planned 222 MW data center facility at Fermi’s Project Matador development near Amarillo.
The proposed agreement was expected to generate approximately $6.5 billion over an initial 15-year term, assuming the transaction closed and the contractual conditions were fulfilled.
According to Fermi’s November 3 statement, remaining conditions include final documentation, operational agreements, board approvals, and project-level construction financing. Neither party has exercised its contractual termination rights, and Fermi says discussions are continuing. The company indicated that it expects to provide additional information during its November 12 earnings call, or sooner if appropriate.
Project Matador is one of the industry’s most ambitious proposed developments, built around a plan for extensive private power infrastructure to support AI computing capacity.
The TensorWave situation illustrates an increasingly important financing distinction. A proposed long-term lease can provide the foundation for a major infrastructure project, but the agreement must become effective and its associated construction financing must be secured before it can be treated as firm contracted capacity.
That distinction has become more consequential as AI computing itself emerges as a financeable infrastructure asset, a transition examined in DCF’s coverage of NVIDIA’s AI factory financing strategy.
Microsoft’s Nscale Decision Adds Another Qualification
Another development surfaced November 5, when Reuters reported that Microsoft had decided not to exercise an option to expand its existing computing agreement with Nscale at a Texas data center.
The proposed expansion would have substantially increased the contracted capacity at the company’s Ward County development. Importantly, the decision had reportedly been made months before the election, and the existing Microsoft-Nscale agreement remains in place. It therefore should not be characterized as a consequence of Texas’ political shift.
Nevertheless, the decision adds another example of the difference between an announced development pipeline, a customer option, and a firm contractual commitment. In a market measured in hundreds of megawatts and multibillion-dollar agreements, those distinctions are becoming essential to evaluating project readiness.
Dallas Inference Agreement Highlights the Value of Existing Capacity
Not every AI infrastructure deployment requires a new hyperscale campus.
On October 26, Forum Markets announced that its Forum Edge AI joint venture with Edge Node had entered into a $94.4 million, 36-month agreement with Fireworks, an AI inference platform. The agreement covers reservations of NVIDIA B300 GPU computing capacity at an existing Dallas data center, with deployment expected in late December 2026 or early January 2027.
The company’s SEC filing identifies a deployment involving 96 enterprise AI servers.
Unlike a greenfield gigawatt-scale campus, this arrangement relies on an existing powered facility. The distinction matters as the industry moves from an almost exclusive focus on the enormous infrastructure requirements of model training toward a broader mix of AI workloads, including inference services.
Existing facilities with available electrical and thermal capacity may provide a faster route to deployment, particularly while new projects undergo extended regulatory and interconnection reviews.
The opportunity to extract more usable compute capacity from infrastructure already in place is also drawing renewed attention. In a recent DCF Show interview with Hammerhead AI CEO Rahul Kar, the discussion centered on using AI workload orchestration to turn otherwise unused electrical capacity into productive inference infrastructure without compromising existing customer commitments.
ServerDomes Brings an Unconventional Model to Hudspeth County
The proposed ServerDomes Global project near Desert Haven in Hudspeth County illustrates a different response to the industry’s infrastructure challenges.
The company is pursuing a campus at Hueco Ranch, west of the state’s larger AI development concentrations. Plans reported for the site contemplate approximately 350 MW of data center capacity, potentially distributed across as many as ten geodesic dome structures. The proposed first phase would occupy roughly 700 acres within a larger property.
Rather than relying primarily on conventional utility service, ServerDomes proposes dedicated behind-the-meter natural gas generation and closed-loop cooling infrastructure. The company’s design is intended to accommodate high-density computing, including both air- and liquid-cooled configurations.
However, local residents have raised questions about the development’s effects on roads, water resources, wildlife, and the surrounding environment. According to local reporting by KVIA, representatives met with residents on October 26 after concerns that information about the project had not been communicated adequately. County officials have sought further information supporting the developer’s claims about water consumption and electrical generation.
Those claims require careful distinction from independently verified operating performance. A proposed closed-loop system, for example, does not by itself establish that a facility will have no continuing water requirements. The project’s development status and ultimate environmental footprint remain contingent on further documentation and implementation.
Hueco Ranch is representative of a larger challenge for AI infrastructure developers: private generation and unconventional building designs do not eliminate the need to establish credibility with local governments and communities. As DCF has previously examined in its reporting on community opposition, local acceptance is increasingly becoming a material consideration in determining which projects can advance.
San Antonio and Other Cities Tighten Local Controls
Regulatory pressure is also increasing within Texas’ established metropolitan markets.
On October 29, the San Antonio City Council voted 9-2 to adopt new data center development regulations. The rules establish additional land-use review and specific-use authorization requirements for affected zoning districts, along with standards addressing setbacks, landscaping, noise barriers, parking, and site access. The ordinance also prohibits new data center development in the Edwards Aquifer Recharge Zone District.
San Antonio’s action reflects a broader reassessment of the industry’s physical presence in urban and suburban communities. On October 27, League City approved a zoning amendment requiring special-use permits for data centers, giving local officials additional authority to evaluate proposed developments. Coppell has also advanced changes requiring specific review of data centers and cryptocurrency mining facilities in light-industrial districts.
Although the requirements vary, the underlying direction is similar. Communities that previously handled data centers through broad industrial-use classifications are creating more explicit standards for evaluating facilities with substantial electricity, water, generation, and land-use implications.
The result is an increasingly complicated local approval landscape, even before developers encounter the state’s evolving grid and environmental regulations.
A Statewide Permitting Pause Raises the Stakes
The municipal activity is occurring against the backdrop of Abbott’s statewide regulatory actions.
On September 21, the outgoing governor directed the Texas Commission on Environmental Quality to halt permits sought by data center projects until regulators completed audits examining electricity and water requirements. The directive followed earlier instructions to ERCOT, the Public Utility Commission of Texas, and the Texas Water Development Board to gather and verify information about proposed facilities.
Implementation became controversial in October after environmental advocates identified data center-related applications that appeared to have continued advancing. TCEQ subsequently acknowledged problems with its electronic filing process and temporarily shut down automated submissions while making changes intended to ensure applications received appropriate review.
The distinction between the state’s current pause and Hinojosa’s proposed future moratorium is important.
Abbott’s directives were tied to specified audits and regulatory review. Hinojosa has proposed maintaining a broader suspension of new approvals until lawmakers establish additional protections. Whether those measures can be implemented in the form proposed, and how they would apply to projects already permitted or under construction, will require further decisions by state officials.
The situation reflects a larger national shift toward more proof-driven data center development. As DCF explored in How States Are Rewriting the Rules for Data Center Growth, states are increasingly requiring developers to substantiate power availability, infrastructure plans, environmental performance, and community commitments before projects can advance.
The Coming Fight Over Tax Incentives
Among the most consequential policy questions is the future of Texas’ tax treatment of data centers.
The industry’s qualified data center sales-tax exemption has grown dramatically alongside AI infrastructure investment. According to state comptroller estimates reported by The Texas Tribune, the exemption is projected to reduce state revenue by approximately $3.2 billion over a two-year period.
Abbott called for eliminating the exemption and other data center incentives during the next legislative session. Hinojosa has also supported ending subsidies, while advocating additional community protections and greater transparency around proposed developments.
This is a significant area of policy overlap between the outgoing and incoming governors, even though their broader approaches differ.
The issue is particularly relevant for developers whose financial models or pending transactions depend on receiving public incentives. Repeal or modification of existing programs would require legislative action, with important questions about effective dates, existing beneficiaries, and possible transition provisions.
Industry advocates argue that data centers provide construction employment, property-tax revenue, and other economic benefits that should be considered when evaluating incentives. Critics question whether those benefits justify the scale of the exemptions and the public infrastructure required to support rapidly expanding AI loads.
The 2027 debate will help determine how Texas balances continued investment with the costs and obligations associated with data center development.
