Pinned Tweet
$SLNH $MSFT A few days ago, I teased some of the Microsoft connections I’d been digging into around Soluna’s Project Kati 2. The response showed that quite a few of you wanted to see the full research.
I’ve also been fairly quiet on X lately, and this is a big reason why. This took a lot more time and focus than I originally expected.
FOLLOWING THE POWER: The Case for Microsoft at Soluna’s Project Kati 2 --> Full 12-part investigation:
perspez.substack.com/p/follo…
I originally thought I could turn all of this into one detailed X post. That stopped making sense once the research kept expanding.
I ended up going through Microsoft Research, EDF, Masdar, Las Majadas, Project Tumbleweed, Fairwater, Ryan Carver, Siemens, Kati 2’s evolution from an 83 MW second phase into a 350+ MW AI campus, and the anonymous tenant now negotiating around its first 100 MW phase.
This has been weeks of digging through filings, research papers, company materials, historical agreements, infrastructure developments and personnel connections. I’ve built it piece by piece, and in the end I decided I’d rather share the full work than squeeze it into a few conclusions on X and strip away the context that makes the thesis interesting.
There was simply too much that would have been left out, so I put everything together as one long-form Substack instead.
12 parts. Weeks of research. One thesis: why so many independent trails around Project Kati 2 keep converging on Microsoft.
Enjoy the deep dive.
$SLNH $MSFT
Guys, I have some really big evidence pointing to a Microsoft–Soluna connection.
I’ve gone much deeper with the research, and I’m honestly surprised by what I found. There’s far too much to fit into an X post, so I’m going to put together a much more detailed deep dive somewhere I can lay out all the findings, sources, and connections properly.
Stay tuned.
$SLNH | Microsoft's Nscale Withdrawal Puts a Premium on AI Infrastructure That Can Deliver Power Before 2030
Microsoft's reported withdrawal from Nscale's planned data center in Essex offers a revealing glimpse into how the economics of AI infrastructure are changing. The British company's proposed facility in Loughton has encountered repeated delays, and electricity network authorities reportedly do not expect to supply the site with power until the 2030s.
The Telegraph first reported Microsoft's decision, with Bloomberg and Reuters subsequently covering the development. Although Microsoft has not publicly explained its withdrawal, the problems surrounding the project's electricity connection raise an increasingly consequential question for the industry: how much is future data center capacity worth when a customer cannot be certain when it will become operational?
The implications extend well beyond Nscale. Hyperscalers are committing enormous amounts of capital to AI computing infrastructure, frequently planning deployments around successive generations of increasingly powerful and expensive accelerators. Those investments are sensitive to timing. A data center that becomes operational in 2028 can support years of workloads before another facility awaiting grid access in 2031 receives its first megawatt.
For customers making long-term infrastructure commitments, several years of delay can materially change the financial assumptions behind a project. GPUs depreciate, new architectures improve performance per watt, and the opportunity to generate revenue from current-generation computing equipment does not wait for utilities to complete their transmission upgrades.
This increasingly favors developers with credible near-term power delivery schedules over those whose principal assets remain land, development plans and distant interconnection dates.
Soluna Holdings (NASDAQ: SLNH) has been positioning itself around precisely this constraint.
The company's approach differs from conventional data center development, where securing a suitable site can be followed by years of negotiations over electricity supply and grid capacity. Soluna develops computing infrastructure alongside existing renewable generation, using behind-the-meter arrangements to access electricity close to its source while retaining the ability to interact with the wider grid.
That model has already been deployed commercially, although Soluna's operating facilities currently serve primarily Bitcoin-related computing rather than high-density AI workloads.
In September, Soluna completed the final 14 MW of Project Kati 1 in Willacy County, Texas, bringing the campus to its full 83 MW operating capacity and increasing the company's total energized computing portfolio to approximately 206 MW. According to Soluna, every phase of Kati 1 was delivered on schedule, with infrastructure costs remaining within budget.
The achievement matters beyond the additional operating capacity. It provides evidence of the company's ability to build, energize and manage substantial computing loads within the Texas electricity market, a record that becomes relevant when prospective AI customers evaluate execution risk.
September also brought a significant development on the regulatory side. ERCOT conditionally classified the full 166 MW Kati campus and 100 MW at Project Dorothy as Base Load under its Batch Zero framework, preserving their respective interconnection positions subject to verification.
These classifications cover 266 MW of capacity across the two campuses, including existing operating facilities. At Kati, 83 MW is associated with the planned AI expansion at Kati 2. They do not mean that 266 MW of new AI infrastructure is ready for immediate deployment, or that the larger development pipeline has secured all the electricity it requires.
Their importance is more specific: Soluna is advancing its AI developments from established electrical infrastructure and previously studied interconnection positions, rather than beginning the entire process with an unconnected site.
That distinction becomes particularly relevant at Project Kati 2.
Developed in partnership with Metrobloks, Kati 2 is planned as a high-density AI campus with an initial phase exceeding 100 MW of critical IT capacity and a further 250 MW targeted for subsequent development.
The company has progressed through design development, brought a general contractor into the process and secured commitments for certain long-lead electrical equipment. Engineering work is also underway to expand the Las Majadas substation by an additional 100 MW in support of future phases.
One of the more consequential developments concerns the project's electricity architecture. Soluna has executed an access agreement with a natural gas pipeline operator and begun engineering the lateral connection to the site.
That agreement does not establish a completed onsite generation facility or guarantee future gas-fired electricity output. It does, however, move an important part of the project's supplemental power strategy into a more concrete development stage.
Kati 2 is being designed around a combination of renewable generation, grid access, onsite gas-fired generation and energy storage. The intention is to provide the continuous and resilient electricity supply required by modern AI computing, including workloads that cannot simply be interrupted whenever wind generation declines.
It is an important evolution of the behind-the-meter model Soluna established through Bitcoin hosting. AI infrastructure requires a considerably different operating profile, with higher power densities, more demanding cooling systems and stricter availability requirements.
On the commercial side, Soluna has signed a letter of intent with a prospective Kati 2 tenant and has been negotiating the terms of a definitive lease. The company has previously described a construction period of approximately 15 months from contract signing, subject to customer requirements and execution conditions.
That places a potential initial delivery window in late 2027 or early 2028, provided the remaining contractual, financing and infrastructure milestones progress as planned. The timetable is still conditional, but it illustrates why Kati 2 could become more attractive as the market reassesses projects whose electricity connections are several years further away.
Soluna has a second major development opportunity at Project Dorothy 3 in Briscoe County.
In August, the company completed the acquisition of 397 acres adjacent to its existing Dorothy operations and the 150 MW Briscoe Wind Farm, which it acquired earlier this year. The property is intended to support a new AI campus with potential capacity of up to 300 MW.
The wind farm's substation sits directly across from the acquired land. Soluna therefore controls an unusually valuable combination of existing electricity generation, operating computing infrastructure and adjacent development property.
Dorothy 3 still requires additional engineering, utility coordination, commercial commitments and financing before its proposed AI capacity can be delivered. The existing wind farm cannot independently provide the continuous power needed for a 300 MW AI campus. Nevertheless, ownership of the generation asset and nearby land removes several uncertainties that other developers must resolve before construction can begin.
The decision to appoint Ryan Carver as Chief Development Officer adds another dimension to Soluna's development strategy.
Carver previously served as Microsoft's Senior Director of AI Construction and Site Development, where his responsibilities included the company's Fairwater campus in Wisconsin. His experience overseeing hyperscale construction programs gives Soluna direct exposure to the technical, procurement and scheduling standards used by some of the industry's most demanding customers.
For a company seeking to move from Bitcoin hosting into purpose-built AI infrastructure, that experience is relevant to its ability to turn proposed capacity into facilities that customers can actually occupy.
The investment implications reach beyond construction schedules.
Electricity availability increasingly influences where hyperscalers deploy capital, which developers can secure long-term leases and how lenders assess the probability that a project will begin generating cash flow on schedule.
A campus capable of supporting commercial AI operations in 2028 may have a considerable economic advantage over one facing an uncertain 2031 grid connection, even where the latter offers greater theoretical capacity.
Earlier delivery can bring forward rental income, reduce exposure to changing hardware requirements and give tenants greater confidence in their deployment plans. Those advantages can influence both lease negotiations and the financing terms available to infrastructure developers.
For Soluna, this creates the possibility that its existing electrical assets, previously established interconnection positions and relatively advanced development work will attract greater attention as hyperscalers become more selective about delivery risk.
There is no indication that Microsoft's withdrawal from the Essex project is connected to Soluna, nor should a British computing requirement automatically be viewed as demand that can relocate to Texas. The more important development is the potential shift in how customers and investors compare competing infrastructure projects.
Soluna's remaining milestones are substantial. Kati 2 needs a binding customer agreement, project financing, completion of its supplemental power infrastructure and successful construction of an AI facility far more complex than its existing Bitcoin operations. Dorothy 3 has an even longer development path ahead.
Yet the company has already assembled several of the assets that are becoming increasingly difficult to obtain: operating sites, access to renewable generation, established interconnection infrastructure, development land and a team with direct hyperscale construction experience.
Microsoft's reported decision at Nscale underscores the commercial consequences of electricity infrastructure failing to keep pace with AI deployment schedules. For developers targeting 2027–2028, it may also improve the competitive environment at a particularly important moment.
As the industry moves from announcing gigawatts of proposed capacity to financing and occupying actual facilities, investors may increasingly distinguish between developers with ambitious long-term pipelines and those positioned to deliver usable megawatts within the next two years.
Soluna's opportunity is to turn that timing advantage into signed leases, funded construction and operating AI capacity. If it succeeds, the value of its existing power infrastructure could look substantially different from how the market assesses it today.
Their importance is more specific: Soluna is advancing its AI developments from established electrical infrastructure and previously studied interconnection positions, rather than beginning the entire process with an unconnected site.
That distinction becomes particularly relevant at Project Kati 2.
Developed in partnership with Metrobloks, Kati 2 is planned as a high-density AI campus with an initial phase exceeding 100 MW of critical IT capacity and a further 250 MW targeted for subsequent development.
The company has progressed through design development, brought a general contractor into the process and secured commitments for certain long-lead electrical equipment. Engineering work is also underway to expand the Las Majadas substation by an additional 100 MW in support of future phases.
One of the more consequential developments concerns the project's electricity architecture. Soluna has executed an access agreement with a natural gas pipeline operator and begun engineering the lateral connection to the site.
That agreement does not establish a completed onsite generation facility or guarantee future gas-fired electricity output. It does, however, move an important part of the project's supplemental power strategy into a more concrete development stage.
Kati 2 is being designed around a combination of renewable generation, grid access, onsite gas-fired generation and energy storage. The intention is to provide the continuous and resilient electricity supply required by modern AI computing, including workloads that cannot simply be interrupted whenever wind generation declines.
It is an important evolution of the behind-the-meter model Soluna established through Bitcoin hosting. AI infrastructure requires a considerably different operating profile, with higher power densities, more demanding cooling systems and stricter availability requirements.
On the commercial side, Soluna has signed a letter of intent with a prospective Kati 2 tenant and has been negotiating the terms of a definitive lease. The company has previously described a construction period of approximately 15 months from contract signing, subject to customer requirements and execution conditions.
That places a potential initial delivery window in late 2027 or early 2028, provided the remaining contractual, financing and infrastructure milestones progress as planned. The timetable is still conditional, but it illustrates why Kati 2 could become more attractive as the market reassesses projects whose electricity connections are several years further away.
Soluna has a second major development opportunity at Project Dorothy 3 in Briscoe County.
In August, the company completed the acquisition of 397 acres adjacent to its existing Dorothy operations and the 150 MW Briscoe Wind Farm, which it acquired earlier this year. The property is intended to support a new AI campus with potential capacity of up to 300 MW.
The wind farm's substation sits directly across from the acquired land. Soluna therefore controls an unusually valuable combination of existing electricity generation, operating computing infrastructure and adjacent development property.
Dorothy 3 still requires additional engineering, utility coordination, commercial commitments and financing before its proposed AI capacity can be delivered. The existing wind farm cannot independently provide the continuous power needed for a 300 MW AI campus. Nevertheless, ownership of the generation asset and nearby land removes several uncertainties that other developers must resolve before construction can begin.
The decision to appoint Ryan Carver as Chief Development Officer adds another dimension to Soluna's development strategy.
Carver previously served as Microsoft's Senior Director of AI Construction and Site Development, where his responsibilities included the company's Fairwater campus in Wisconsin. His experience overseeing hyperscale construction programs gives Soluna direct exposure to the technical, procurement and scheduling standards used by some of the industry's most demanding customers.
For a company seeking to move from Bitcoin hosting into purpose-built AI infrastructure, that experience is relevant to its ability to turn proposed capacity into facilities that customers can actually occupy.
The investment implications reach beyond construction schedules.
Electricity availability increasingly influences where hyperscalers deploy capital, which developers can secure long-term leases and how lenders assess the probability that a project will begin generating cash flow on schedule.
A campus capable of supporting commercial AI operations in 2028 may have a considerable economic advantage over one facing an uncertain 2031 grid connection, even where the latter offers greater theoretical capacity.
Earlier delivery can bring forward rental income, reduce exposure to changing hardware requirements and give tenants greater confidence in their deployment plans. Those advantages can influence both lease negotiations and the financing terms available to infrastructure developers.
For Soluna, this creates the possibility that its existing electrical assets, previously established interconnection positions and relatively advanced development work will attract greater attention as hyperscalers become more selective about delivery risk.
There is no indication that Microsoft's withdrawal from the Essex project is connected to Soluna, nor should a British computing requirement automatically be viewed as demand that can relocate to Texas. The more important development is the potential shift in how customers and investors compare competing infrastructure projects.
Soluna's remaining milestones are substantial. Kati 2 needs a binding customer agreement, project financing, completion of its supplemental power infrastructure and successful construction of an AI facility far more complex than its existing Bitcoin operations. Dorothy 3 has an even longer development path ahead.
Yet the company has already assembled several of the assets that are becoming increasingly difficult to obtain: operating sites, access to renewable generation, established interconnection infrastructure, development land and a team with direct hyperscale construction experience.
Microsoft's reported decision at Nscale underscores the commercial consequences of electricity infrastructure failing to keep pace with AI deployment schedules. For developers targeting 2027–2028, it may also improve the competitive environment at a particularly important moment.
As the industry moves from announcing gigawatts of proposed capacity to financing and occupying actual facilities, investors may increasingly distinguish between developers with ambitious long-term pipelines and those positioned to deliver usable megawatts within the next two years.
Soluna's opportunity is to turn that timing advantage into signed leases, funded construction and operating AI capacity. If it succeeds, the value of its existing power infrastructure could look substantially different from how the market assesses it today.
$slnh | The AI Data Center Boom Is Entering a New Phase: From Announced Gigawatts to Deliverable Capacity
A new Financial Times investigation raises an important question about the scale of America's AI infrastructure expansion: How much of the announced capacity will actually get built?
According to Bernstein, only approximately 35% of the planned U.S. data center pipeline is considered credible.
The firm's analysis identifies roughly 488 GW of announced nameplate capacity, but considers only about 170 GW credible.
Meanwhile, regulatory intelligence firm RegLayer has identified at least 47 major U.S. projects that have been delayed, withdrawn, or blocked in 2026. The Financial Times also reports that more than 1,000 communities have pursued moratoriums or other restrictions on AI infrastructure development this year.
These figures reveal a growing distinction between the industry's enormous development ambitions and the practical requirements of delivering operational computing capacity.
The constraints are becoming increasingly visible.
Electricity: Securing land and announcing a multi-gigawatt campus does not guarantee sufficient generation, transmission capacity, or an achievable grid connection.
Financing: Higher borrowing costs and greater scrutiny of AI infrastructure debt are forcing developers to demonstrate credible project economics, stronger counterparties, and realistic construction schedules.
Permitting: Community opposition, environmental reviews, and concerns about electricity prices and water consumption are creating additional uncertainty.
Execution: Even projects backed by major AI companies face engineering challenges, equipment constraints, construction overruns, and contractual delivery obligations.
Yet these developments should not automatically be interpreted as evidence that demand for AI computing is disappearing.
Instead, they suggest that the market is beginning to differentiate more aggressively between speculative development pipelines and projects with a realistic path to commercial operation.
That distinction could have important consequences for infrastructure valuations.
If a substantial portion of announced capacity never materializes, access to genuinely deliverable electricity may become more valuable than the headline size of a development pipeline.
Developers that have already secured suitable land, credible power arrangements, regulatory progress, and experienced engineering partners could find themselves in a stronger competitive position as less mature projects encounter delays.
This is particularly relevant to power-first developers such as Soluna Holdings ($SLNH).
Soluna's strategy centers on developing computing infrastructure around energy resources, including behind-the-meter renewable generation and hybrid power configurations.
Its Kati 2 and Dorothy 3 projects represent potential exposure to a market where the ability to deliver reliable electricity may increasingly determine which campuses attract customers and financing.
That does not eliminate the company's execution risks. Tenant agreements, financing commitments, engineering milestones, and construction schedules remain essential before proposed capacity can be treated as commercially secured.
But the broader industry implications are becoming clearer.
For several years, investors have focused heavily on the number of gigawatts developers could announce.
The next phase of the AI infrastructure cycle may place considerably greater emphasis on the number of megawatts they can actually finance, construct, energize, and monetize.
The opportunity in AI infrastructure remains enormous. The amount of capacity capable of capturing that opportunity may be considerably smaller than the announced pipeline suggests.
Source: Financial Times, October 18, 2026; Bernstein research; RegLayer.
$SLNH | John Belizaire: More Hyperscale-Level Talent Is Coming to Soluna
@jbelizaireCEO latest LinkedIn post caught my attention, particularly his comments about why world-class executives are recognizing the opportunity at Soluna and his intention to attract more people of Ryan Carver's caliber.
To appreciate the significance of that statement, it's worth understanding what Carver accomplished before joining Soluna.
At Microsoft, Carver led the construction program at Fairwater in Wisconsin, one of the most ambitious and technologically sophisticated AI infrastructure developments in the world.
He personally described leading construction of its first-of-kind facility, the most technically complex building in the program and the largest individual construction budget in Microsoft's history.
The project was delivered ahead of schedule.
Microsoft CEO Satya Nadella subsequently highlighted Fairwater's achievement, confirming it came online six weeks early and congratulating the teams responsible. He described Fairwater as the world's most powerful AI datacenter.
Think about the experience behind that achievement.
Carver spent more than a decade at Microsoft, ultimately helping deliver infrastructure at the very frontier of hyperscale AI development.
And after all that, he chose Soluna.
Not because Soluna is already operating at Microsoft's scale, but because he sees something compelling in its approach to integrating digital infrastructure directly with energy generation.
That is what makes John's latest comments particularly interesting.
"When fantastic talent makes a move like this, it shows you where the experts see the next big opportunity."
John also expects more leaders from hyperscale companies to move into energy and AI infrastructure development.
And then he added:
"We're going to bring in more people like Ryan."
That's a significant statement coming from the CEO of a company still in the early stages of building out its AI infrastructure platform.
Soluna is looking to assemble leadership with firsthand experience delivering some of the world's most demanding AI projects, while pursuing a fundamentally different approach to how those projects are powered.
Consider what that combination could mean as Kati 2, Dorothy 3, and the broader development pipeline advance.
The experience gained building projects like Fairwater doesn't disappear when an executive changes companies. That knowledge, execution discipline, and understanding of hyperscale requirements become part of the organization they join.
And if John succeeds in attracting more executives with comparable backgrounds, Soluna's organizational capabilities could evolve considerably.
What stands out to me is the direction of that talent movement. People who have worked inside the world's largest AI infrastructure organizations are beginning to recognize opportunities in smaller developers focused on solving the industry's power constraints.
Ryan Carver's decision to join Soluna is already meaningful. John's ambition to attract more people of that caliber makes the bigger picture even more interesting.
SOLUNA HIRES FORMER KINDER MORGAN VP AS TEXAS AI PROJECTS ADVANCE
$SLNH has appointed Ben J. Carranza, P.E., as Director of Community Engagement, filling the Texas-based position the company disclosed it was recruiting for Project Dorothy 3 in July.
Carranza brings a background that extends well beyond community relations. He previously served as Vice President, Regulatory at Kinder Morgan, with years of experience navigating federal regulation of interstate natural gas infrastructure.
Public records offer a closer look at that experience.
In 2016, Carranza was identified as Manager, Rates & Regulatory in a federal filing involving offshore natural gas pipelines in Texas. By 2018, he was Director, Regulatory at Tennessee Gas Pipeline.
That year, a FERC application named him as the regulatory contact for Tennessee Gas Pipeline's Rio Bravo border-crossing facilities in Hidalgo County, South Texas. The company sought authorization to increase capacity from 320 million to 420 million cubic feet per day without constructing additional facilities.
In 2020, he was also named as a regulatory contact in the $171 million Evangeline Pass Expansion application, involving 1.1 billion cubic feet per day of additional gas transportation capacity.
By 2023, Carranza was serving as Vice President, Regulatory. Federal records identify him in proceedings involving Natural Gas Pipeline Company of America, while EPA correspondence addressed to him concerned emissions reporting and environmental regulation of natural gas transmission and storage facilities.
His documented experience spans pipeline capacity, federal authorizations, gas transportation and environmental regulatory processes.
Soluna is bringing that experience into a very different infrastructure business, but one facing some familiar development challenges.
Dorothy 3 is planned as a 300+ MW AI campus alongside Soluna's existing Dorothy infrastructure and its acquired 150 MW Briscoe Wind Farm. Environmental assessments, land surveys, fiber studies and master planning are among the development activities underway.
In July, Soluna specifically identified the need for a Texas-based community engagement specialist to advance discussions with local officials. Carranza now fills that position.
His natural gas background is also worth examining alongside Kati 2, Soluna's planned 350+ MW AI campus being developed with Metrobloks.
Soluna has executed an access agreement with an undisclosed natural gas pipeline operator, with engineering underway on the lateral connection intended to support onsite generation.
Carranza's appointment follows two other additions to Soluna's AI development organization:
• Ryan Carver, Chief Development Officer: Joined from Microsoft in July after more than a decade working on hyperscale data center development. He led the construction program at Microsoft's Fairwater AI campus in Wisconsin and now oversees Soluna's AI development portfolio.
• Neil Robinson, VP of Construction Management: Joined in September directly from Microsoft, where he served as Director of AI Construction. His previous responsibilities included major AI data center developments in the Milwaukee area, overlapping with Microsoft's Wisconsin expansion.
Soluna has now recruited two former Microsoft AI construction executives alongside a former Kinder Morgan regulatory VP.
Their responsibilities address different requirements of large-scale data center development: site selection, engineering, construction delivery, energy infrastructure and relationships with the communities hosting those facilities.
For Dorothy 3, Carranza's appointment fulfills a recruitment objective disclosed two months earlier. His experience with federal energy regulation adds another dimension to the local engagement role as Soluna progresses toward tenant agreements, financing and construction.
$SLNH | Dorothy 3: The 50 MW Timeline, ChronoScale, Microsoft and an $8B+ Development
I've been comparing Soluna's Dorothy 3 announcements with recent Texas government records and ChronoScale's SEC filings.
The chronology is particularly interesting. Soluna began working on a 50 MW expansion months before ChronoScale appeared in the state's Dorothy 3A registration, and several commercial developments have followed.
Here's how the events line up:
April 2026 – Soluna acquires Briscoe Wind Farm
Soluna completes its $53M acquisition of the 150 MW Briscoe Wind Farm in Texas, giving the company ownership of the renewable generation asset supporting its existing Dorothy campus and future expansion.
May 7 – Dorothy 3's initial 50 MW study
In its April business update, Soluna specifically reports that an initial 50 MW expansion interconnection load study is underway for Project Dorothy 3 with Golden Spread Electric Cooperative and Oncor.
At this stage, approximately 400 acres are also under LOI, with fiber studies and community engagement underway.
August 6 – ChronoScale signs with Microsoft
ChronoScale enters into a GPU Services Agreement with Microsoft for approximately 50 MW of AI compute capacity at a Texas data center.
The deployment will use NVIDIA GB300 systems and advanced liquid cooling.
August 21–24 – Two important developments
ChronoScale amends its Microsoft agreement, extending the arrangement to approximately three years.
On August 24, Soluna announces the completed acquisition of 397 acres for Dorothy 3, adjacent to its existing data centers and the Briscoe Wind Farm.
The wind farm's substation is located directly across from the acquired property, providing an important infrastructure advantage for the planned 300 MW AI campus.
August 27 – Microsoft partnership publicly announced
ChronoScale formally announces its 50 MW AI infrastructure partnership with Microsoft.
September 8 – Project Dorothy 3A appears in Texas records
This is one of the most interesting findings.
The Texas Comptroller's official qualifying data center registry lists:
- Facility: Project Dorothy 3A
- Effective date: September 8, 2026
- Owner: STX-01 LLC
- Operator: STX-01 LLC
- Occupant: ChronoScale Corporation
- Registration: DC474523
We now have ChronoScale specifically identified in a government record for Dorothy 3A.
September 14 – Soluna confirms formal 50 MW interconnection request
Soluna announces that its existing 100 MW Dorothy campus has received conditional Base Load classification under ERCOT's Batch Zero process.
The same release confirms that Soluna has formally submitted an additional 50 MW load interconnection request at Dorothy, with the transmission provider already studying it.
Importantly, Soluna's May update had already identified an initial 50 MW interconnection study specifically under Dorothy 3.
September 23 – Briscoe County files its $8B+ development notice
The county files a public notice ahead of its November 9 meeting.
The document identifies:
- Approximately 403.397 acres near the existing Dorothy facilities.
- Proposed Reinvestment Zone #6.
- More than $8 billion in estimated data center improvements, equipment and associated infrastructure.
- Land owned by Soluna Dorothy 3 Energy ServiceCo. LLC and Soluna DV BNSF Energy HoldCo. LLC.
- Tax abatement applications involving STX-01 LLC, identified as a Soluna subsidiary.
The same STX-01 entity named in the Dorothy 3A registration now appears in the county's tax incentive process.
October 2026 – Soluna presents Dorothy 3's phased development
At its Silverton Open House, Soluna displays the campus development plan:
- Phase 1: 75 MW
- Phase 2: 75 MW
- Phase 3: 150 MW
That initial 75 MW phase is particularly interesting alongside ChronoScale's approximately 50 MW Microsoft deployment.
$SLNH | Dorothy 3 Takes Shape: A 300 MW AI Campus Built Around Flexibility and Available Power
Soluna recently published photos on LinkedIn from its Open House in Silverton, Texas, where the company introduced Project Dorothy 3 to the local community.
One of the presentation boards caught my attention. It shows a more detailed development plan for Dorothy 3 than what I've seen in the company's earlier public materials.
The proposed 300 MW AI campus is divided into three phases:
- Phase 1: 75 MW
- Phase 2: 75 MW
- Phase 3: 150 MW
The board also appears to contain information about expected completion, although the relevant details are obscured in the photograph.
The significance here isn't simply the 300 MW figure, which Soluna has discussed for some time. It's how the company appears to be planning its development.
An interesting shift in AI infrastructure demand
Two reports from September make this particularly relevant.
On September 18, CNBC reported that OpenAI and Anthropic were exploring smaller data center deployments of approximately 20–30 MW, alongside their existing commitments to much larger AI campuses.
A day earlier, The Wall Street Journal reported that Crusoe, the developer behind OpenAI's massive Abilene data center, was expanding into smaller, factory-built modular facilities that can be transported to locations with available electricity.
Crusoe's strategy reflects growing demand for faster capacity delivery, particularly as AI inference expands. Large training campuses remain essential, but not every computing workload requires a facility with hundreds of megawatts available from day one.
That distinction matters for Dorothy 3.
An initial 75 MW phase could allow Soluna to match construction and capital requirements more closely to contracted demand, bring capacity online in stages and expand as additional power becomes available.
This isn't confirmation that Dorothy 3 will use Crusoe-style prefabricated modules. However, its proposed phase structure is consistent with the industry's increasing emphasis on flexible capacity deployment.
Dorothy 3 already has an important foundation
Unlike developers starting with undeveloped land and no nearby generating assets, Soluna has spent much of 2026 consolidating its position around Dorothy.
The company now owns the adjacent 150 MW Briscoe Wind Farm, has consolidated ownership of Dorothy 1 and completed the purchase of 397 acres for Dorothy 3.
The Briscoe substation sits directly across from the newly acquired property, providing an important starting point for the planned behind-the-meter development.
Soluna also has approximately 98 MW of existing operating infrastructure at Dorothy 1 and 2.
In September, ERCOT conditionally classified 100 MW at the existing Dorothy campus as Base Load under Batch Zero, while a separate 50 MW interconnection request remained under study. This does not constitute approval for Dorothy 3's new phases, but it highlights the value of developing alongside an established, energized campus.
There is also an important distinction between Soluna's approach and a conventional wind-only data center.
The company's strategy involves combining renewable generation with grid access and additional power resources. Soluna has discussed evaluating onsite gas generation and clustering nearby energy assets to support larger AI campuses.
That combination could be especially valuable for a phased project where power requirements increase as new data halls are developed.
The Open House photographs reveal more than the capacity plan
Another presentation board addresses the project's expected noise and nearby residences.
The material appears to reference approximately 55 dB at the facility fence, along with a map identifying one residence about 1.5 miles northeast and three residences at least 1.8 miles west of the site.
$SLNH $MSFT $APLD
October 11 – ChronoScale and Applied Digital advance a Texas data center arrangement
ChronoScale enters into a term sheet with Applied Digital for a $50M upfront advance covering tenant fit-out costs at a third-party Texas data center.
The company describes a planned lease with a joint venture between Applied Digital and another party that controls the development site.
According to ChronoScale, this facility is expected to provide the capacity required for its Microsoft agreement.
October 14 – ChronoScale's SEC filing provides further detail
Its quarterly filing describes the Microsoft deployment as approximately 50 MW over a three-year service term, with NVIDIA GB300 GPU deliveries planned in two tranches during February and March 2027.
The same filing details the proposed Texas lease and the $50M fit-out arrangement.
November 9 – Briscoe County's next decision
Commissioners are scheduled to consider establishing Reinvestment Zone #6 and approving potential property tax abatement agreements involving Soluna's STX-01 structure.
---
What interests me most is how these developments fit together.
Soluna was already studying 50 MW for Dorothy 3 in May. It subsequently secured ownership of the land, formally submitted the expansion request, and appeared in a state registration naming ChronoScale as the occupant of Dorothy 3A.
Meanwhile, ChronoScale signed a 50 MW agreement with Microsoft and subsequently disclosed a proposed Texas data center lease involving Applied Digital and a third-party site controller.
The first phase of Dorothy 3 is planned for 75 MW, giving additional context to the scale of a potential initial deployment.
The documents establish the individual corporate, infrastructure and commercial developments. The exact location of ChronoScale's Microsoft deployment remains the key connection to establish.
We've spent considerable time investigating Microsoft-related signals around Kati 2, including personnel, power architecture, technical developments and industry relationships.
Dorothy 3 now presents an additional Microsoft-related lead through ChronoScale, a company already named in the project's official Texas registration.
Between Soluna's 150 MW wind farm ownership, existing Dorothy infrastructure, 397-acre land acquisition, phased 300 MW development plan, ongoing 50 MW interconnection process, ChronoScale registration and the county's $8B+ improvement proposal, the project has accumulated a substantial amount of tangible progress.
I'm looking forward to seeing what the November 9 county proceedings reveal about the next stages of Dorothy 3.
Today is a great day to add more $SLNH shares.
The fundamentals are stronger than ever, yet the stock is extremely oversold.
Thank you, Mr. Market, for another opportunity to accumulate at these levels.
The disconnect between price and fundamentals is hard to ignore.
$SLNH | The AI Data Center Selloff: Why I'm Still Bullish on Soluna
The AI data center sector has been taking a beating lately, with IREN, Cipher, TeraWulf, Applied Digital and several other infrastructure developers facing considerable selling pressure.
Soluna has been caught in the same downturn, with its shares once again trading below $1.
Higher financing costs, concerns about AI infrastructure spending and growing scrutiny of capital-intensive development projects have made investors increasingly cautious. Companies with signed long-term contracts are finding it easier to justify their valuations, while developers still working toward major commercial agreements are facing a tougher market.
Some of that caution is understandable. These projects require enormous amounts of capital, and the returns will ultimately depend on lease economics, financing terms and successful execution.
Still, I believe the negative sentiment has become excessive, particularly when looking at the progress Soluna has made throughout 2026.
The financial performance tells a very different story from the share price.
With so much attention focused on Kati 2, the growth of Soluna's existing business has received surprisingly little recognition.
Consider the first-half results:
- H1 revenue: $24.45 million, compared with $12.09 million in H1 2025, representing approximately 102% reported growth.
- H1 hosting revenue: $19.34 million, compared with $5.54 million a year earlier, an increase of roughly 249%.
- Q2 revenue: $15.06 million, up 145% year over year as reported, or approximately 73% on a comparable basis after adjusting for the electricity pass-through accounting change.
- Adjusted EBITDA: The quarterly loss narrowed from $2.09 million in Q1 to $1.56 million in Q2, an improvement of approximately 25%.
- Operating capacity: Kati 1 reached its full 83 MW capacity in September, bringing company-wide operating capacity to approximately 206 MW.
- Liquidity: $113.4 million in unrestricted cash at the end of June.
The accounting change affects reported revenue growth, but the expansion of the hosting business is still substantial.
Adjusted EBITDA has also improved considerably from the $6.4 million quarterly loss reported in Q3 2025. Soluna hasn't reached consolidated adjusted EBITDA breakeven, but the trend gives me reason to believe that positive adjusted EBITDA is becoming an increasingly realistic objective as the operating platform grows.
Kati 1 is particularly relevant here. The final 14 MW was energized in September, with construction completed on schedule and within the infrastructure budget. The recent expansion of the Bitdeer agreement from 28 MW to 35 MW also fully subscribed the K1BC phase.
That additional capacity wasn't contributing for the full second quarter. I'm interested in what the expanded platform can deliver as these deployments ramp and contribute over longer periods.
Briscoe is another example of operational progress that I believe has been overlooked.
When Soluna acquired the 150 MW Briscoe Wind Farm in April, ten of its 81 turbines were out of service.
Management moved quickly to address the maintenance backlog, incurring approximately $1.5 million in repair and maintenance expenses during Q2.
CFO Michael Picchi subsequently confirmed that all ten turbines had been repaired and returned to service by early Q3.
That's encouraging for two reasons. The wind farm has regained previously unavailable generating capacity, and the unusually heavy maintenance burden experienced immediately after the acquisition may ease.9
There is also the seasonal element. Wind conditions are generally more favorable during fall and winter, potentially supporting improved generation, although actual results will depend on production and market conditions.
Beyond its immediate financial contribution, Briscoe gives Soluna direct ownership of an important generation asset serving the Dorothy campus.
Perspez retweeted
Nothing fundamental has changed about @SolunaHoldings.
We keep pushin’
Stay tuned.
$SLNH
$SLNH | Dorothy 3 Takes Shape: A 300 MW AI Campus Built Around Flexibility and Available Power
Soluna recently published photos on LinkedIn from its Open House in Silverton, Texas, where the company introduced Project Dorothy 3 to the local community.
One of the presentation boards caught my attention. It shows a more detailed development plan for Dorothy 3 than what I've seen in the company's earlier public materials.
The proposed 300 MW AI campus is divided into three phases:
- Phase 1: 75 MW
- Phase 2: 75 MW
- Phase 3: 150 MW
The board also appears to contain information about expected completion, although the relevant details are obscured in the photograph.
The significance here isn't simply the 300 MW figure, which Soluna has discussed for some time. It's how the company appears to be planning its development.
An interesting shift in AI infrastructure demand
Two reports from September make this particularly relevant.
On September 18, CNBC reported that OpenAI and Anthropic were exploring smaller data center deployments of approximately 20–30 MW, alongside their existing commitments to much larger AI campuses.
A day earlier, The Wall Street Journal reported that Crusoe, the developer behind OpenAI's massive Abilene data center, was expanding into smaller, factory-built modular facilities that can be transported to locations with available electricity.
Crusoe's strategy reflects growing demand for faster capacity delivery, particularly as AI inference expands. Large training campuses remain essential, but not every computing workload requires a facility with hundreds of megawatts available from day one.
That distinction matters for Dorothy 3.
An initial 75 MW phase could allow Soluna to match construction and capital requirements more closely to contracted demand, bring capacity online in stages and expand as additional power becomes available.
This isn't confirmation that Dorothy 3 will use Crusoe-style prefabricated modules. However, its proposed phase structure is consistent with the industry's increasing emphasis on flexible capacity deployment.
Dorothy 3 already has an important foundation
Unlike developers starting with undeveloped land and no nearby generating assets, Soluna has spent much of 2026 consolidating its position around Dorothy.
The company now owns the adjacent 150 MW Briscoe Wind Farm, has consolidated ownership of Dorothy 1 and completed the purchase of 397 acres for Dorothy 3.
The Briscoe substation sits directly across from the newly acquired property, providing an important starting point for the planned behind-the-meter development.
Soluna also has approximately 98 MW of existing operating infrastructure at Dorothy 1 and 2.
In September, ERCOT conditionally classified 100 MW at the existing Dorothy campus as Base Load under Batch Zero, while a separate 50 MW interconnection request remained under study. This does not constitute approval for Dorothy 3's new phases, but it highlights the value of developing alongside an established, energized campus.
There is also an important distinction between Soluna's approach and a conventional wind-only data center.
The company's strategy involves combining renewable generation with grid access and additional power resources. Soluna has discussed evaluating onsite gas generation and clustering nearby energy assets to support larger AI campuses.
That combination could be especially valuable for a phased project where power requirements increase as new data halls are developed.
The Open House photographs reveal more than the capacity plan
Another presentation board addresses the project's expected noise and nearby residences.
The material appears to reference approximately 55 dB at the facility fence, along with a map identifying one residence about 1.5 miles northeast and three residences at least 1.8 miles west of the site.
These details suggest that Soluna is working through practical development considerations, including the relationship between the proposed campus and the surrounding community.
This follows a process that began earlier in the year. In May, Soluna announced that it had established a community relations task force to engage Briscoe County residents ahead of data center construction.
By August, the company had also initiated preliminary master planning, environmental and water studies, fiber assessments, utility coordination and long-lead equipment planning.
The Open House is a visible continuation of that work.
And the customer interest deserves attention
During Soluna's Q2 earnings call in August, CFO Michael Picchi described the inbound interest in Dorothy 3 as very strong and explained that it had accelerated the company's development efforts.
CEO John Belizaire also indicated that interest extended across hyperscalers and neocloud providers.
That's an important part of the story.
Dorothy 3's development activity isn't taking place in isolation. According to management, customer interest was already influencing the pace of the project months before these Open House photographs appeared.
The latest images don't establish that a tenant agreement has been signed, but they do offer another look at how Soluna is preparing the campus for commercial development.
What stands out to me is the combination of assets and execution.
Soluna has secured the land, acquired the neighboring wind farm, strengthened its existing campus position and advanced the technical and community work necessary for a large AI development.
We can now also see a proposed 75 MW first phase within a much larger 300 MW plan.
The next development I'm particularly interested in is how that initial phase will be matched with firm power arrangements, customer commitments and a construction schedule.
With AI companies increasingly focused on how quickly they can access usable computing capacity, Dorothy 3's combination of phased development, energy ownership and behind-the-meter infrastructure looks increasingly relevant.
Sources: Soluna LinkedIn Open House photos (October 2026); Soluna Q2 2026 Business Update and earnings call; Soluna August 24 land acquisition announcement; Soluna September 14 ERCOT announcement; CNBC (September 18); The Wall Street Journal (September 17).
Combined with the consolidation of Dorothy 1 and the land acquired for Dorothy 3, the company now has considerably greater control over how that site can develop.
And control over land, energy assets and the eventual computing infrastructure is becoming increasingly valuable.
The AI development platform has also strengthened considerably.
Soluna has brought in people with direct experience delivering some of the largest and most technically demanding AI facilities in the industry.
Ryan Carver joined as Chief Development Officer after serving as Senior Director of AI Construction & Site Development at Microsoft, where he led the construction program at Fairwater.
Neil Robinson, another former Microsoft AI construction executive, joined as VP of Construction Management.
Ben Carranza also joined the organization, bringing energy infrastructure and regulatory experience from Kinder Morgan.
I consider these appointments particularly encouraging. Building a hyperscale AI campus involves a completely different level of construction complexity, power engineering and coordination from operating a conventional Bitcoin hosting facility.
Having executives with that experience inside Soluna should strengthen its ability to manage the transition.
Meanwhile, the company has continued advancing Kati 2 through engineering, tenant-informed design work and its definitive Metrobloks joint venture.
The October 7 TDLR registration outlining a planned 112 MW critical IT facility is another indication of how much more concrete the development has become.
Kati 2 has also secured an executed natural gas pipeline access agreement, with engineering work on the lateral underway. That is an important step in developing the planned hybrid power system.
Across Kati and Dorothy, ERCOT has conditionally classified a combined 266 MW as Base Load through its Batch Zero process.
And with a development pipeline now totaling approximately 6.3 GW, Soluna has opportunities extending well beyond its initial AI campuses.
None of these milestones replaces the need for a signed tenant agreement, but the work being done is substantial.
Regarding Kati 2, I suspect some of the recent weakness reflects disappointed short-term expectations.
Throughout the summer, management communicated encouraging progress in the tenant process.
The exclusive LOI signed in July, advanced commercial negotiations and direct tenant involvement in the design process understandably generated excitement.
John Belizaire was optimistic about the opportunity, and given the progress made on the project, there were reasonable grounds for that optimism.
Although management never formally committed to announcing a definitive lease by September, some investors clearly expected Q3 to deliver that milestone.
When September passed without an agreement, I suspect a number of short-term investors lost patience. The broader sector downturn may have amplified the selling, potentially creating opportunities for short sellers as well, although their actual impact would require supporting trading data.
But the absence of an announcement doesn't tell us that the commercial opportunity has deteriorated.
A development of this scale involves complex technical requirements, power arrangements, construction schedules and financing commitments. Getting those elements right can matter considerably more than meeting a timeline anticipated by the market.
I'd rather see Soluna negotiate a strong long-term agreement with attractive economics than rush a deal to satisfy short-term expectations.
Financing will naturally be part of that discussion.
The company is seeking shareholder approval on October 16 to increase authorized common shares from 375 million to 1 billion, alongside a separate proposal concerning potential issuance under the Yorkville equity facility.
I understand why investors are sensitive to dilution, especially after the equity raised this year.
But authorization isn't the same as issuance, and Soluna's larger AI developments are intended to rely substantially on project-level financing and external capital partners.
Management has previously discussed a potential structure involving 70–80% project debt and 20–30% equity.
The eventual tenant's credit quality, lease economics and financing terms will determine how attractive the arrangement is for existing shareholders.
That is what I will be paying attention to when the agreement arrives.
The longer-term outlook for AI infrastructure is another reason I remain optimistic.
The United States is competing with China for leadership in artificial intelligence, and that competition increasingly depends on physical infrastructure.
Advanced AI chips are only useful when sufficient computing capacity and electricity are available to operate them.
According to the International Energy Agency, global data center electricity consumption is projected to more than double to approximately 945 TWh by 2030. The United States is expected to account for the largest share of that increase, with data centers responsible for nearly half of projected US electricity demand growth through the end of the decade.
The IEA has also warned that electricity infrastructure constraints could put approximately 20% of planned data center projects at risk of delays.
China is already building large computing clusters in regions with abundant and relatively inexpensive electricity, taking advantage of its ability to develop generation and data center infrastructure at scale.
Washington is clearly aware of the challenge.
The White House's America's AI Action Plan places the expansion of data centers and domestic energy infrastructure at the center of its strategy for maintaining American AI leadership.
The Department of Energy's Speed to Power initiative is another indication of that priority, focusing on accelerating electricity generation and transmission development.
The challenge isn't simply finding enough electricity on paper. It's getting reliable power to the right locations within the timeframes AI customers require.
And that's where Soluna's model becomes particularly interesting.
Rather than depending entirely on conventional grid-connected development, Soluna has built its business around locating computing infrastructure near renewable generation.
For AI applications, that approach is being expanded into a hybrid architecture combining behind-the-meter renewable power, grid connectivity, planned on-site natural gas generation and energy storage.
This allows Soluna to pursue the cost advantages of renewable electricity while designing for the higher reliability requirements of AI workloads.
Its collaboration with Siemens is also relevant, particularly around power system modeling, grid stability and managing the rapidly changing electrical loads associated with advanced AI computing.
The approach still needs to demonstrate its economics and reliability at hyperscale AI capacity. But the underlying engineering challenges are exactly the ones much of the industry is now trying to solve.
And the economic importance of this infrastructure is growing.
Federal Reserve research has already highlighted how investment associated with computing equipment, software and data center infrastructure has contributed to US economic growth.
The benefits extend into construction, electrical equipment manufacturing, utilities, engineering and other industrial sectors.
A prolonged slowdown in AI infrastructure investment could therefore affect an increasingly important source of American economic activity, while also complicating efforts to maintain technological leadership over China.
I don't expect Washington to guarantee the success of individual developers or prevent every market correction. But there are strong economic and strategic reasons for policymakers to keep working to remove obstacles to domestic AI infrastructure development.
That makes me more confident about the long-term market for companies capable of delivering reliable electricity and computing capacity at competitive costs.
Which brings me back to Soluna and the current valuation.
A year ago, Soluna had a smaller operating footprint, considerably less hosting revenue and a less developed AI infrastructure organization.
Today, H1 revenue has more than doubled, the hosting business has expanded substantially and adjusted EBITDA has improved significantly from Q3 2025.
Kati 1 is complete. Briscoe's previously inactive turbines are back in service. The company has strengthened its liquidity, added experienced hyperscale executives and continued advancing Kati 2 and Dorothy 3.
Yet the shares are trading below $1.
I understand why the market wants a signed Kati 2 lease and greater clarity on financing. Those will be important milestones, and the eventual economics need to justify the capital involved.
But the current selling pressure seems to be giving very little recognition to the growth already taking place.
Soluna has an established hosting business that is generating increasing revenue, alongside a much larger AI development opportunity built around a problem the entire industry is facing: access to affordable, reliable power.
There is still plenty to execute, and the next few quarters will be important. But I don't see the missed Q3 tenant expectations as a reason to disregard everything the company has accomplished.
I'm still bullish on Soluna because its operating business continues to expand, its financial performance is improving and its approach to power infrastructure remains relevant to where the AI industry is heading.
At around $0.90, I believe the market may be placing too much weight on short-term uncertainty while underestimating the value of the business Soluna has already built and the opportunities it is developing.
$SLNH | The AI Data Center Selloff: Why I'm Still Bullish on Soluna
The AI data center sector has been taking a beating lately, with IREN, Cipher, TeraWulf, Applied Digital and several other infrastructure developers facing considerable selling pressure.
Soluna has been caught in the same downturn, with its shares once again trading below $1.
Higher financing costs, concerns about AI infrastructure spending and growing scrutiny of capital-intensive development projects have made investors increasingly cautious. Companies with signed long-term contracts are finding it easier to justify their valuations, while developers still working toward major commercial agreements are facing a tougher market.
Some of that caution is understandable. These projects require enormous amounts of capital, and the returns will ultimately depend on lease economics, financing terms and successful execution.
Still, I believe the negative sentiment has become excessive, particularly when looking at the progress Soluna has made throughout 2026.
The financial performance tells a very different story from the share price.
With so much attention focused on Kati 2, the growth of Soluna's existing business has received surprisingly little recognition.
Consider the first-half results:
- H1 revenue: $24.45 million, compared with $12.09 million in H1 2025, representing approximately 102% reported growth.
- H1 hosting revenue: $19.34 million, compared with $5.54 million a year earlier, an increase of roughly 249%.
- Q2 revenue: $15.06 million, up 145% year over year as reported, or approximately 73% on a comparable basis after adjusting for the electricity pass-through accounting change.
- Adjusted EBITDA: The quarterly loss narrowed from $2.09 million in Q1 to $1.56 million in Q2, an improvement of approximately 25%.
- Operating capacity: Kati 1 reached its full 83 MW capacity in September, bringing company-wide operating capacity to approximately 206 MW.
- Liquidity: $113.4 million in unrestricted cash at the end of June.
The accounting change affects reported revenue growth, but the expansion of the hosting business is still substantial.
Adjusted EBITDA has also improved considerably from the $6.4 million quarterly loss reported in Q3 2025. Soluna hasn't reached consolidated adjusted EBITDA breakeven, but the trend gives me reason to believe that positive adjusted EBITDA is becoming an increasingly realistic objective as the operating platform grows.
Kati 1 is particularly relevant here. The final 14 MW was energized in September, with construction completed on schedule and within the infrastructure budget. The recent expansion of the Bitdeer agreement from 28 MW to 35 MW also fully subscribed the K1BC phase.
That additional capacity wasn't contributing for the full second quarter. I'm interested in what the expanded platform can deliver as these deployments ramp and contribute over longer periods.
Briscoe is another example of operational progress that I believe has been overlooked.
When Soluna acquired the 150 MW Briscoe Wind Farm in April, ten of its 81 turbines were out of service.
Management moved quickly to address the maintenance backlog, incurring approximately $1.5 million in repair and maintenance expenses during Q2.
CFO Michael Picchi subsequently confirmed that all ten turbines had been repaired and returned to service by early Q3.
That's encouraging for two reasons. The wind farm has regained previously unavailable generating capacity, and the unusually heavy maintenance burden experienced immediately after the acquisition may ease.9
There is also the seasonal element. Wind conditions are generally more favorable during fall and winter, potentially supporting improved generation, although actual results will depend on production and market conditions.
Beyond its immediate financial contribution, Briscoe gives Soluna direct ownership of an important generation asset serving the Dorothy campus.
I understand why investors are sensitive to dilution, especially after the equity raised this year.
But authorization isn't the same as issuance, and Soluna's larger AI developments are intended to rely substantially on project-level financing and external capital partners.
Management has previously discussed a potential structure involving 70–80% project debt and 20–30% equity.
The eventual tenant's credit quality, lease economics and financing terms will determine how attractive the arrangement is for existing shareholders.
That is what I will be paying attention to when the agreement arrives.
The longer-term outlook for AI infrastructure is another reason I remain optimistic.
The United States is competing with China for leadership in artificial intelligence, and that competition increasingly depends on physical infrastructure.
Advanced AI chips are only useful when sufficient computing capacity and electricity are available to operate them.
According to the International Energy Agency, global data center electricity consumption is projected to more than double to approximately 945 TWh by 2030. The United States is expected to account for the largest share of that increase, with data centers responsible for nearly half of projected US electricity demand growth through the end of the decade.
The IEA has also warned that electricity infrastructure constraints could put approximately 20% of planned data center projects at risk of delays.
China is already building large computing clusters in regions with abundant and relatively inexpensive electricity, taking advantage of its ability to develop generation and data center infrastructure at scale.
Washington is clearly aware of the challenge.
The White House's America's AI Action Plan places the expansion of data centers and domestic energy infrastructure at the center of its strategy for maintaining American AI leadership.
The Department of Energy's Speed to Power initiative is another indication of that priority, focusing on accelerating electricity generation and transmission development.
The challenge isn't simply finding enough electricity on paper. It's getting reliable power to the right locations within the timeframes AI customers require.
And that's where Soluna's model becomes particularly interesting.
Rather than depending entirely on conventional grid-connected development, Soluna has built its business around locating computing infrastructure near renewable generation.
For AI applications, that approach is being expanded into a hybrid architecture combining behind-the-meter renewable power, grid connectivity, planned on-site natural gas generation and energy storage.
This allows Soluna to pursue the cost advantages of renewable electricity while designing for the higher reliability requirements of AI workloads.
Its collaboration with Siemens is also relevant, particularly around power system modeling, grid stability and managing the rapidly changing electrical loads associated with advanced AI computing.
The approach still needs to demonstrate its economics and reliability at hyperscale AI capacity. But the underlying engineering challenges are exactly the ones much of the industry is now trying to solve.
And the economic importance of this infrastructure is growing.
Federal Reserve research has already highlighted how investment associated with computing equipment, software and data center infrastructure has contributed to US economic growth.
The benefits extend into construction, electrical equipment manufacturing, utilities, engineering and other industrial sectors.
A prolonged slowdown in AI infrastructure investment could therefore affect an increasingly important source of American economic activity, while also complicating efforts to maintain technological leadership over China.
I don't expect Washington to guarantee the success of individual developers or prevent every market correction. But there are strong economic and strategic reasons for policymakers to keep working to remove obstacles to domestic AI infrastructure development.
That makes me more confident about the long-term market for companies capable of delivering reliable electricity and computing capacity at competitive costs.
Which brings me back to Soluna and the current valuation.
A year ago, Soluna had a smaller operating footprint, considerably less hosting revenue and a less developed AI infrastructure organization.
Today, H1 revenue has more than doubled, the hosting business has expanded substantially and adjusted EBITDA has improved significantly from Q3 2025.
Kati 1 is complete. Briscoe's previously inactive turbines are back in service. The company has strengthened its liquidity, added experienced hyperscale executives and continued advancing Kati 2 and Dorothy 3.
Yet the shares are trading below $1.
I understand why the market wants a signed Kati 2 lease and greater clarity on financing. Those will be important milestones, and the eventual economics need to justify the capital involved.
But the current selling pressure seems to be giving very little recognition to the growth already taking place.
Soluna has an established hosting business that is generating increasing revenue, alongside a much larger AI development opportunity built around a problem the entire industry is facing: access to affordable, reliable power.
There is still plenty to execute, and the next few quarters will be important. But I don't see the missed Q3 tenant expectations as a reason to disregard everything the company has accomplished.
I'm still bullish on Soluna because its operating business continues to expand, its financial performance is improving and its approach to power infrastructure remains relevant to where the AI industry is heading.
At around $0.90, I believe the market may be placing too much weight on short-term uncertainty while underestimating the value of the business Soluna has already built and the opportunities it is developing.