Sign a lease. Name-drop an AI tenant. Add a few hundred megawatts. Watch the chart go vertical.

That trade is not dead, but it is getting harder to impress the market.

TheEnergyMag’s analysis of 25 AI and high-performance-computing infrastructure announcements from June 2024 through Aug. 4, 2026 found an interesting split. The deals are arriving faster and carrying slightly richer revenue per megawatt. At the same time, the stocks announcing them are moving less.

The best-fit absolute announcement-day move—the size of the move, regardless of direction—falls from 32.5% at the start of the period to 6.3% at the end. The first eight announcements in the sample produced an average absolute close-to-close move of 24.1%. The median was 14.8%. For the last eight announcements, the average dropped to 10.2% and the median to 7.3%.

Meanwhile, annualized base-term revenue per contracted megawatt for colocation and build-to-suit leases inches up from roughly $1.67 million to $1.90 million.

In other words, the contracts got fatter while the candles got shorter. It is not to suggest that the market is shrugging these deals off. But it appears to be reacting with a polite nod instead of spraying champagne around the room like it used to.

Back when an AI lease could rewrite the whole story

The first wave of deals did something genuinely new. They showed that bitcoin miners and other power-rich operators could turn substations, land and grid connections into credible AI infrastructure businesses.

When Core Scientific announced roughly 200 MW of hosting contracts with CoreWeave in June 2024, the stock jumped 40.2% from the prior close. That agreement did more than add revenue. It gave investors a new way to value the entire company.

Applied Digital’s first 250 MW CoreWeave lease produced a 48.5% move in June 2025. Then TeraWulf’s initial 200-plus MW Fluidstack agreements sent its shares up 59.5% in August.

Those were not ordinary contract announcements. They were corporate identity changes delivered by press releases.

Fast-forward to this summer. TeraWulf signed a $19 billion, 401 MW lease with Anthropic. The stock gained 4.9% on a close-to-close basis. CleanSpark landed a $6.6 billion, 175 MW lease. Its shares rose 8.8%. Bitdeer’s new $4.7 billion, 121 MW Tydal agreement initially pumped its stock by 12% on Tuesday before all of those gains were erased at the market’s close.

These are still enormous contracts. A year earlier, any one of them might have blown the doors off the stock. Now the market wants to see the wiring diagram.

Everyone has an AI pipeline now

Scarcity creates excitement. Familiarity turns excitement into a spreadsheet.

TheEnergyMag’s sample contains just two announcements in 2024, followed by 12 in 2025 and another 11 in the first seven months and four days of 2026. This year’s deals already cover 2,039 MW—slightly more than the 2,001 MW announced during all of last year.

At this point, an AI strategy is no longer a surprise. For many of the miners and power-to-compute companies, it is practically part of the dress code.

That does not mean every announcement is equal. The fitted revenue-per-megawatt line rises over time, but only modestly, and the relationship is noisy. The median 2026 deal is actually below the 2025 median on that measure.

The strongest conclusion is therefore not that lease pricing is marching relentlessly higher. It is that the industry is signing more large deals—and investors have developed a longer checklist for deciding which ones deserve a rerating.

Who is the tenant? Is there an investment-grade backstop? Who pays for the build? When does revenue start? What happens if construction slips? And after debt, equity issuance and project costs, how much of the economics actually reaches each share?

A lease can validate a site while leaving all of those questions open.

The index has a hangover too

The TEM AI Infrastructure Growth Index closed at 6,384 on Aug. 5. It had bounced 23% over the previous seven days after hitting a local low near the end of July.

That sounds healthy until you zoom out. The index remained roughly 28.5% below its June peak and was still down 9% over one month.

That is not a market abandoning the AI infrastructure story. It is a market trying to decide how much of the story it already paid for.

The sector has traveled a long way from the days when access to power and an AI press release were enough. Investors now have a growing stack of signed leases to compare. They can separate scarce sites from ordinary ones, funded projects from financing plans and near-term rent from revenue that may not begin for another two years.

Two years ago, a major AI/HPC agreement could create an entirely new equity story overnight. Today, it is more likely to become another row in the model.

The megawatts are still growing. The easy multiple expansion may be what is running out of fuel.

Regulation News

• Texas pauses AI data center power hookups amid grid fears

Hardware and Infrastructure News

• Crusoe Plans $1B Texas AI Data Center Expansion, Pursues Nuclear Power

• Sphere 3D Plans 50 MW Kentucky Data Center as City Weighs Zoning Rules

• Nexus Seeks $15B for Anthropic Texas AI Data Center With Google Backstop

• DeepSeek Plans 1-Gigawatt AI Data Center in Inner Mongolia

• Bitdeer Signs $4.7 Billion Norway AI Data Center Lease With Volta

• Anthropic Deal Underpins Bitdeer’s $4.7 Billion Norway AI Lease

• CoreWeave Plans 360MW Indonesia Expansion in First Asia AI Data Center Push

Corporate News

• Southern Company Plans $2.15B Raise via Convertible Bond

• American Bitcoin Names Paul Sacks Interim CFO as Matthew Prusak Departs

• Matthew Prusak joins Giga Energy as Chief Business Officer and Interim CFO

• MARA Adds Power, Data Center Veterans to Board Amid AI Push

• NRG Energy Reports $7.48 Billion Q2 2026 Revenue, T.H. Wharton Peaker Plant Begins Operations

• IREN Completes Acquisition of Mirantis

• Nebius appoints Lindsey Irvine as Chief Marketing Officer

• CoreWeave Signs Multi-Year Agreement With Solidigm to Strengthen Its Integrated AI Cloud Platform

Financial News

• Aschenbrenner’s $24B AI Hedge Fund Unwinds Trades After Losses: Report

• Bitzero Secures $25M Through Warrant Issuance for Infrastructure and Debt Reduction

• TeraWulf’s AI Leasing Revenue Reaches 71% as Data Center Shift Accelerates

• Galaxy Digital Reports $85 Million Q2 2026 Net Loss and 526 MW CoreWeave Lease

• Fluence Energy Reports $44.3 Million Net Loss for Q3 Fiscal 2026

• Solaris Energy Infrastructure Reports $219 Million Revenue for Q2 2026

• Canaan Inc. to Repurchase Shares Using Portion of Digital Asset Treasury

Email us

The AI Infrastructure Growth Index tracked by TheEnergyMag fell 11.13% on Wednesday alone, extending its one-month decline to 34%. The index has now retreated about 42% from its record high in June, as selling spreads across chipmakers, neoclouds, power companies and former bitcoin miners recast as data center developers.

The immediate catalyst was a renewed debate over how the AI boom is being financed. Nvidia was reported this week to be considering a guarantee of up to $250 billion to help OpenAI lease a planned 10-gigawatt data-center campus in Ohio, along with separate financing of up to $350 billion for Nvidia chips used at the site.

The discussions revived concerns about "circular financing": chipmakers and hyperscalers invest in AI laboratories or cloud providers, which then spend some of that capital on equipment or computing capacity supplied by their investors.

Such arrangements do not by themselves show that demand is artificial. They can help customers finance infrastructure whose cost and construction timelines would otherwise make development difficult. But they also concentrate risk. If AI laboratories cannot generate enough revenue to honor their commitments, the pressure could flow back through neoclouds, data-center developers, lenders and ultimately chip suppliers.

Other pressures intensified the retreat. Reports of progress in China’s domestic chipmaking industry and the blockbuster debut of memory producer CXMT raised fears of greater competition. Oil prices and Treasury yields also jumped Wednesday, increasing financing costs for a sector whose projected revenue often sits years in the future.

Few examples capture the reversal as clearly as Situational Awareness, the hedge fund founded by former OpenAI employee Leopold Aschenbrenner.

The roughly $20 billion fund recently held discussions with existing investors and lenders about raising fresh capital after suffering losses in the AI-stock rout, the Financial Times reported, citing people briefed on the matter. It also offered some investors the opportunity to buy assets from its portfolio, although one source described the outreach as ad hoc rather than a coordinated fundraising effort.

Situational Awareness had gained 439% on a net basis during the first half of 2026, according to an investor letter reviewed by the FT. The fund used borrowing to magnify its exposure, a strategy that amplified returns during the rally but increased the potential damage when the trade reversed.

Aschenbrenner presented the downturn differently. In the July 24 letter, he argued that the selloff had created some of the most attractive opportunities since early 2025 and invited investors to add capital on Aug. 1. The fund’s fundraising can therefore be read both as a response to pressure and an attempt to buy into the decline.

For mining investors, the broader question is how far the unwind will extend into companies whose valuations increasingly depend on AI rather than bitcoin production.

A 30-day comparison through Wednesday shows bitcoin gaining about 6.84%, while every AI-infrastructure and HPC-related equity included in TheEnergyMag’s comparison declined. CleanSpark, the best performer among the displayed stocks, fell roughly 9%. Cipher Digital and MARA dropped about 16%, while IREN lost 25%, TeraWulf declined 31%, and American Bitcoin slid 46%.

The divergence is striking because mining stocks have historically traded as high-beta bitcoin proxies. That relationship weakened as investors began assigning greater value to miners’ electricity contracts, land, substations and grid connections—all assets that could be redeployed for AI computing.

Multibillion-dollar AI agreements then pushed some miners further away from their original business. Their shares became increasingly sensitive to hyperscaler spending, chip availability, construction financing and the valuations of AI laboratories and neocloud companies.

That shift worked in both directions. Only last week, mining and data-center stocks outperformed during a Nasdaq selloff as investors rewarded new AI contracts. The subsequent rout suggests that the group’s brief insulation from the technology market was less durable than it appeared.

Bitcoin’s relative strength raises the possibility that capital is beginning to rotate back from AI equities to the big orange. Earlier this summer, the movement appeared to run the other way. K33 Research said in June that investors were leaving bitcoin for higher-flying AI stocks, pointing to 62,794 bitcoin of outflows from exchange-traded products over three weeks, the second-largest such streak on record at the time. K33 warned that the opportunity cost of holding bitcoin had become difficult for investors to accept while AI-related equities were soaring.

The latest price action may indicate that dynamic is reversing, but relative performance alone does not establish a capital rotation. Investors selling AI stocks could be moving into cash, bonds or less volatile equities rather than bitcoin. For now, the evidence more clearly supports a decoupling: bitcoin has held its value while the companies that increasingly traded on AI expectations have not.

Mike Alfred, founder of Alpine Fox and a longtime investor in mining infrastructure, offered the countercase at the Energy Investors Forum last week.

“There’s no bubble, so you don’t have to ask,” Alfred said, arguing that demand for power, cooling, land and data-center capacity could expand for another 20 to 30 years. He acknowledged that the buildout would include crashes and downturns, but said those interruptions would not invalidate the longer-term demand for computing.

His exposure is material. Alfred is an IREN director and told the forum that roughly 80% of his fund was concentrated in IREN and Cipher. Both stocks are down around 30% over the past month.

The selloff does not resolve whether AI infrastructure is overbuilt. It does, however, raise the standard for what investors are willing to finance. Signed contracts, creditworthy tenants and funded construction plans are likely to matter more than undeveloped power capacity or prospective AI demand.

Bitcoin may be regaining some independence from the AI trade. For miners that spent the past year persuading investors they were no longer merely bitcoin companies, however, a rotation back to the cryptocurrency would be an awkward kind of victory.

Regulation News

• Russia detains crypto mining firm BitRiver founder on fraud charges

• Hopkinsville weighs limits on data centers as crypto mine proposes expansion

Hardware and Infrastructure News

• Nebius Details Plans for 1.2GW Data Center Campus in Pennsylvania

• NAVER, NVIDIA, and Brookfield to expand South Korea’s AI infrastructure to 200 MW by 2028

• Bitcoin Power Use Jumps 38% with Greener Energy Mix: Report

• Galaxy Digital Acquires 500-Acre Site for Second Texas AI Data Center

• Core Scientific Pays $42M to Exit Block’s Bitcoin Mining Deal as AI Revenue Surges

• NextEra, Brookfield Plan $100B AI Data Center Campus at Former Kentucky Nuclear Site

• EU lays out $11.4 billion for 7 AI gigafactories as it aims to catch up with US and China

• Dynamix-Led Fund Deploys $95M for Texas AI Data Center Grid Deposits

Corporate News

• Ionic Jumps 26% in Nasdaq Debut, Giving Celsius Claimholders Exit Route

• Core Scientific Lands $14 Billion AMD AI Data Center Deal

• Fortitude Invests $45M in Zcash Mining Infrastructure to Drive Vertical Integration

• Nscale Agrees to Buy Anyscale in Push Into AI Software

Financial News

• Bloom Energy Reports $1.06 Billion in Q2 2026 Revenue

• Bitzero Secures $25M Private Placement for AI Infrastructure and Debt Repayment

• Nvidia in Talks With OpenAI to Guarantee $250 Billion Financing for Data Center

• AI Hedge Fund Situational Awareness Seeks Capital Amid July Market Rout

EIF Recap

• The ‘Mullet’ Mining Pitch Has a Catch: Most Bitcoin Mines Aren’t AI-Ready

• AI Infrastructure Isn’t a Bubble, Mike Alfred Says — But Owning GPUs Is a Bigger Bet

• Soluna CEO Sees Stranded Wind Power as a Shortcut to AI Capacity

Email us

Across the EIF panels, generators, utilities, data-center developers, miners, investors and policymakers repeatedly returned to the same physical constraints: generation, interconnection queues, transmission, substations, cooling, equipment lead times, capital, permits and public support. The demand for compute may begin with models and chips, but delivering it ultimately requires electrons at a specific place, on a commercially useful timetable.

No single part of the energy economy can solve that problem alone. Utilities are being asked to accommodate loads at a scale and speed they were not built to handle. Renewable developers need customers for output that would otherwise be curtailed. Power producers need confidence that new demand will support investment in generation. Data-center operators need firm capacity and redundancy. Investors want contracts capable of supporting billions of dollars in infrastructure. Local officials and residents want credible answers about rates, water, noise, taxes and jobs.

Bitcoin mining entered that wider conversation as one of several bridges between energy and compute. Miners have spent years locating underused power, deploying modular loads and curtailing operations when grids become strained. That experience gives some operators a head start as AI developers search for energized land.

But EIF’s broader message was that controlling a grid connection is only the beginning. A parcel with power does not automatically become a financeable AI campus. It may still lack fiber, high-density cooling, transmission capacity, equipment, permits, a creditworthy tenant or community consent.

That distinction is beginning to sort the market. Some energy projects will support hyperscale AI campuses. Others may be better suited to distributed inference, flexible bitcoin mining, grid services or behind-the-meter industrial demand. Some will remain most valuable as power-and-land assets for another developer to build.

EIF therefore offered less a single forecast than a map of the emerging compute-energy stack — and of the bottlenecks that could determine who captures its value.

Mining’s place in the wider power shift

Alexander Neumüller of the Cambridge Centre for Alternative Finance previewed preliminary findings from the next edition of Cambridge’s digital-mining research. He said the underlying survey represented more than half of global bitcoin mining activity.

Among the preliminary findings presented at EIF:

• Estimated annual mining electricity consumption increased to about 190 terawatt-hours, from 138 TWh, between June 2024 and December 2025.

• Estimated emissions rose to roughly 48 million metric tons of CO₂ equivalent, from about 40 million.

• The estimated low-carbon share of mining’s electricity mix increased to *59.4%*, from 52.4%.

• About 10% of respondents said they had already allocated some power to AI or accelerated computing.

• More than 40% of the remaining respondents said they were actively exploring AI or HPC diversification.

• Nearly nine in 10 respondents expected AI/HPC diversification to become a strategic industry theme.

The presentation captured the industry’s ambivalence. Miners want another source of revenue, but many do not want it to destabilize the mining business that gave them their power portfolio in the first place.

The physical layer is the trade

Mike Alfred, founder and managing partner of Alpine Fox, used his EIF fireside chat to explain why he believes power-rich miners are being recast as AI-infrastructure investments.

"I think we have 20 or 30 really good years for building power, data centers and other capabilities and infrastructure for AI," Alfred said. "I think Texas is Mecca. I think Texas is the most important data-center market in the world."

His argument was that miners that secured land and electricity when those assets were undervalued can now use the same positions to pursue contracted data-center revenue. Referring to companies moving from mining into AI infrastructure, Alfred described the transition as going "from a speculative business on the commodity price of bitcoin to now a contractual, repeatable business based on AI."

Alfred also cautioned that the choice of business model will divide the sector’s winners and losers.

"The big divide in the data-center business amongst the folks who started in bitcoin mining and are now in AI is whether or not to own the GPUs yourself," he said. "If you choose to own them, you better have a good reason why. It’s potentially more lucrative, but also potentially more risky."

For miners seeking a less technology-dependent model, Alfred pointed to colocation: the tenant owns the GPUs, servers and racks, while the infrastructure company supplies power, cooling, connectivity and the building.

"That model looks more like a REIT," he said. "It looks more like a real-estate business. It’s much more conservative. It’s easier to finance."

Megawatts have become the strategic reserve

The clearest point of agreement at EIF was that "speed to power" now determines who can compete for AI projects.

John Belizaire, chief executive of Soluna Holdings, described a model built around locating computing facilities beside renewable-generation assets whose output is constrained or curtailed. Rather than entering a new interconnection queue from scratch, Soluna seeks to modify an existing generation interconnection and add data center load.

Belizaire said that approach can take a fraction of the time required to secure a new grid connection. Soluna’s larger thesis is that curtailed wind and solar generation represents "a source of energy that’s hiding in plain sight": electricity that may not reach the grid but can be monetized by colocated compute.

The value is not limited to electrons. A usable AI site needs transmission or distribution access, fiber, water or an alternative cooling design, permitting, equipment and a credible path to operation. That turns an existing interconnection into a form of strategic real estate.

Several panelists also argued that the next phase may be more distributed than the first wave of enormous AI campuses. Clusters of 10-to-20-megawatt facilities could sometimes connect faster, require less upfront capital and offer geographic redundancy. Smaller facilities may also sit closer to inference customers and avoid some of the transmission upgrades associated with gigawatt-scale campuses.

That model would look familiar to miners. The industry spent years deploying modular data centers wherever power was available rather than waiting for ideal locations.

The “mullet” meets operating reality

One panel reiterated the hybrid strategy with a memorable name: the mullet — AI in the front, bitcoin mining in the back.

The premise is straightforward. Mining can monetize a power position while a developer prepares the site for an AI tenant. If a higher-value lease is signed, the operator converts some or all of the location. Until then, the ASIC fleet generates revenue rather than leaving the interconnection idle.

The strategy has helped public mining companies persuade investors to value them less on hashprice and more on controlled megawatts. A panel moderator pointed to TeraWulf’s Kentucky development and a reported long-duration AI lease as evidence that a former mining site could become a major AI asset.

But the panelists disagreed on whether the mullet is a durable operating model or merely a bridge between businesses. Bitcoin mining and AI place different demands on infrastructure:

• Mining can tolerate interruptions and can often earn money by curtailing when the grid is strained.

• AI customers usually require firm power, tight service-level agreements and substantially greater redundancy.

• Mining containers can be deployed quickly and replaced in modules.

• AI facilities require expensive cooling, networking and electrical systems designed for rapidly increasing rack density.

• A mining site can operate in a remote power pocket with limited connectivity.

• AI workloads often need substantial fiber, skilled labor and access to customers or network hubs.

The practical conclusion was that only a portion of miners’ power portfolios will qualify for high-end AI development. The remainder still has value, but it may be worth more as flexible mining load, grid-balancing capacity or a future development option than as a promised hyperscale campus.

The constraint nobody can engineer away

EIF began and ended with a warning about community acceptance.

Data center projects are encountering questions about electricity prices, water, noise, local hiring, tax incentives and the number of permanent jobs they create. Speakers said developers often arrive too late — after residents have encountered a project through social media or a public-hearing notice.

Curtis Harris of Compass Mining argued that developers should show up in communities “early and often,” well before a formal hearing. “It’s not you, it’s us,” Harris said, directing the criticism at the industry. "We’re the ones who need to be better."

He cited a company outreach effort at an Iowa county fair that emphasized three claims residents could evaluate directly: the operation curtails when requested by the local utility, its mining machines use no water, and it hires local contractors and technicians.

Texas state Rep. Jared Patterson similarly argued that lobbying in Austin is not a substitute for local support. Industry participants need to explain the tax base, school funding, grid relationship and actual water design of a project in terms relevant to residents, he said.

The message was uncomfortable but consistent: the industry cannot assume that a technically sound project will receive political approval. Community engagement is not a communications exercise attached at the end. It is part of development.

That lesson applies equally to mining and AI. Residents may distinguish between the two technologies, but they often experience them in the same way — as large industrial loads seeking land, power and public consent.

What we took away from EIF

The forum did not produce a single blueprint for the mining industry’s next phase. It produced a set of tests.

Power is valuable, but only if it can be delivered on a commercially useful timetable. A mining site can become an AI site, but only if its cooling, connectivity, redundancy and location meet a much higher standard. AI contracts can stabilize revenue, but their financeability depends on the customer. Renewable generation can support compute, but intermittency and curtailment must be designed into the operating model. Texas can remain a compute center, but only if generation, transmission, permitting and local consent keep pace.

The miners best positioned for the next cycle may not be those that declare the largest AI pipeline. They may be the ones that classify their megawatts most honestly:

• which sites should remain mines;

• which can support flexible or distributed compute;

• which justify full AI conversion;

• and which are most valuable as power-and-land assets sold or leased to someone else.

That is a less dramatic story than every miner becoming an AI company. It is also a more investable one.

Hardware and Infrastructure News

• Braiins, Enel Launch Energy Tool as Bitcoin Mining Margins Tighten

• AMD, Anthropic Strike AI Chips Deal with up to 2 GW of MI450 GPU Deployments

Corporate News

• Galaxy Digital secures 15-year naming rights for Texas Tech football stadium

• Big Digital, 10NetZero Acquire Texas AI Data Center Site for $10 Million

• Ionic Digital Set to Begin Nasdaq Trading on July 28

• Ionic Digital Forecasts up to $195M in 2026 Revenue Ahead of Nasdaq Debut

• Bitdeer’s Bitcoin Output Climbs to 990 as AI Cloud Run Rate Reaches $76M

• Poolin Files Chapter 11, Lines Up $52M Sale of Texas Bitcoin Mining Assets

Financial News

• Nebius Secures $775M GPU-Backed Loan to Fund AI Cloud Expansion

• Hut 8 Jumps 14% on $9.8 Billion Lease to Fully Commercialize Texas AI Campus

• Greenidge Rebrands as Vulcan After $39.4M Raise to Fund AI Data Center Pivot

• Galaxy Targets $3.5B Notes Sale to Fund Texas AI Data Center Expansion

• Bitcoin Miners Defy Nasdaq Rout as AI Infrastructure Deals Fuel Gains

• Galaxy Locks In 9.875% Rate on $3.5B AI Data Center Bond Sale

Email us

Energy Investors Forum is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Get your tickets today and use code MNRW30 for 30% off exclusively for Miner Weekly subscribers.

For much of the AI infrastructure boom, the industry’s power problem sounded like something that would arrive in the future.

Utilities warned about rising demand. Grid operators pointed to crowded connection queues. Developers talked about waiting years for transmission upgrades.

That future is beginning to arrive.

New York this week became the first US state to impose a statewide moratorium on certain new large data center projects, pausing incomplete state permit applications while regulators develop rules governing how the facilities should be powered, what they should pay for grid access and what benefits they must provide to host communities.

Also this week, PJM Interconnection, which operates the power grid across all or parts of 13 states and the District of Columbia, saw its latest capacity auction reach a federally approved price ceiling of $325 per megawatt-day. Even at that price, PJM still failed to secure enough power resources to meet its reliability target.

Across the Atlantic, Nscale’s planned £2 billion AI data center in Essex reportedly lost its expected grid-connection timeline, forcing the company to explore generating electricity on site.

Taken together, these developments show that the grid constraint is no longer just a risk hanging over the AI infra boom. It is beginning to act as a real brake on it.

New York wants more than jobs

New York Governor Kathy Hochul described the recent rush of data center proposals as an attempt to “flood the zone.”

Her concern is straightforward: the state must keep the lights on while also keeping electricity affordable for households and businesses.

A 50 MW data center, Hochul said in an interview on Bloomberg’s Odd Lots podcast, can consume roughly as much electricity as 50,000 homes. When several large projects arrive at once, they begin competing with residents, manufacturers and other businesses for the same limited power supply.

New York had nearly 12 GW of data center requests in its grid-connection queue as of May, including more than 8 GW added during 2025.

The state’s response is not intended to be a permanent ban, Hochul said. The one-year pause is meant to give regulators time to establish what she repeatedly called the “rules of the road.”

Her summary of those rules was blunt: Data centers should bring their own power or pay a premium to use New York’s grid.

Hochul said the state is considering several requirements. Developers could be asked to fund new generation, contribute toward transmission upgrades and pay into a larger grid-resiliency fund. They may also be expected to provide direct benefits to the towns and counties hosting their facilities.

That last point matters because many proposed data centers are located in smaller communities that do not have experienced planning departments or teams of lawyers negotiating on their behalf.

Hochul, who previously represented a rural part of upstate New York, said those communities may not know what developers are willing to offer.

She suggested that a locality could seek payments based on the size of a project — giving $1 million per megawatt as an example — alongside investments in the local grid and community.

The figure was not presented as a settled statewide fee. But it illustrates the amount of bargaining power Hochul believes host communities may possess.

The state’s goal is to create a standard framework that local governments can use, rather than leaving each town to negotiate alone with some of the world’s largest technology companies.

Who should pay for the grid?

The New York debate is also about who bears the financial risk of the AI buildout.

Utilities may need to spend billions of dollars on power lines, substations and new generation to serve proposed data centers. But developers often examine several locations before deciding which projects to build.

If a utility upgrades the grid for a campus that is later abandoned, ordinary electricity customers could be left paying for infrastructure that is no longer needed.

New York wants more of that risk to sit with the data center developer.

That could mean larger upfront deposits, dedicated power supplies or other financial guarantees before utilities begin major construction.

The policy would amount to a significant change from the early stages of the AI boom, when states and local governments competed to attract data centers with tax breaks, cheap land and promises of abundant electricity.

New York is still competing for technology investment. But it is making clear that access to the grid will not be free or unconditional.

Data centers must compete with factories

Hochul’s interview also revealed a broader political question: how should governments allocate scarce electricity among competing industries?

She contrasted data centers with Micron’s planned semiconductor manufacturing complex in New York, a project expected to create thousands of direct and indirect jobs.

If she had to choose between powering a “largely vacant data center” and a Micron factory employing thousands of people, Hochul said the choice would be obvious.

That comparison gets to the heart of the challenge facing data center developers.

Data centers can produce substantial construction spending, property-tax revenue and technology investment. But after construction is complete, the largest facilities may employ relatively few permanent workers compared with factories consuming a similar amount of electricity.

As power becomes harder to secure, developers may need to show that a project delivers more than electricity demand and a large building.

They may be asked to fund infrastructure, provide community payments, create training programs or supply their own generation.

Hochul even suggested that future data center operators could arrive with their own small modular nuclear reactors, with New York helping to identify suitable land.

The idea remains speculative, but it captures the direction of travel: the biggest new electricity users may increasingly be expected to bring power with them.

PJM shows the shortage is not theoretical

PJM’s latest capacity auction helps explain why states are becoming more cautious.

Capacity markets pay power plants and other resources to be available when electricity demand is highest. The payment is separate from the price customers pay for the electricity they actually consume.

PJM’s auction for the 2028-2029 delivery year reached its approved ceiling of $325 per megawatt-day. Without that cap, PJM estimated the regional price would have risen to about $555 per megawatt-day.

Even at the ceiling, PJM came up 6,831 MW short of the amount of capacity it says is needed to meet its reliability standard.

The price cap limited how high the auction could go. It did not create more power plants.

PJM’s demand forecast continues to rise, in large part because of new data centers. But building dependable generation and transmission takes far longer than announcing an AI campus.

Only 525 MW of new generation and upgrades cleared the auction, far less than the expected growth in demand.

The result shows that the grid bottleneck is no longer just about waiting in a connection queue. It is also beginning to appear in the cost of keeping enough power available for everyone already connected.

When a data center cannot get electricity

Nscale’s proposed AI campus in Loughton, Essex, shows what happens when the grid cannot match a developer’s schedule.

The company was reportedly told that electricity would not be available in time for the planned £2 billion facility to open next year. It is now considering alternatives, including fuel cells from Bloom Energy that could generate power on site using natural gas.

The project highlights a basic mismatch.

A company may be able to secure land, financing, customers and computing equipment within several years. The transmission infrastructure needed to power the campus can take much longer.

On-site generation can help developers avoid connection delays, but it creates new complications involving emissions, gas supply, environmental permits and community opposition.

The power problem does not disappear. It simply moves somewhere else.

What this means for bitcoin miners

The grid bottleneck strengthens the value of mining sites that already have power.

Keel Infrastructure announced this week that it plans to consolidate electricity from three existing bitcoin mines in Québec into a single 96 MW AI and high-performance computing campus in Sherbrooke. The company is not asking for additional power, only approval to transfer its existing allocation.

CleanSpark’s new 20-year lease in Sandersville, Georgia, shows another path. The company is transforming its long-established mining campus into a data center project expected to generate about $6.6 billion in contracted revenue.

Ionic Digital has already leased 234 MW of energized capacity at its former Ward County mining site to Nscale, although its plans to expand the campus further still depend on utility work and regulatory approval.

The examples point to the same advantage: bitcoin miners may already control the land, substations and grid connections that new data center developers are struggling to secure.

But existing power and future power should not be valued equally. A site that can reuse an established allocation, as Keel is proposing, is different from one that still depends on transmission upgrades or new generation.

As grid constraints begin to slow the AI buildout, the most valuable mining assets may be those that allow developers to avoid asking for new power altogether.

Regulation News

• First Statewide Moratorium on New Hyperscale Data Centers Launched by Governor Kathy Hochul

Hardware and Infrastructure News

• Solo Bitaxe Miner Beats 18,000-Year Odds to Mine Full Bitcoin Block

• Meta Expands Louisiana AI Data Center to 5 GW in $50 Billion Buildout

• Nscale’s £2 Billion UK AI Data Center Reportedly Faces Grid Delay

• CleanSpark Signs $6.6B Data Center Lease, Puts Texas Portfolio Under Exclusivity

• Keel Advances 96 MW Quebec AI Data Center With Sherbrooke Approval

Corporate News

• Sphere 3D Adds Tennessee Lobbying, IR Firms to Advance AI Infrastructure Strategy

• BitFuFu Reports June 2026 Bitcoin Production and Hashrate Metrics

• Galaxy Adds Former Xerox CEO to Board as AI Data Center Push Expands

• IREN Appoints Former Nutanix Executive Eric Hammersley as Security Chief

• Soluna Hires Microsoft Data Center Executive Ryan Carver to Lead AI Buildout

Email us

Energy Investors Forum is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Get your tickets today and use code MNRW30 for 30% off exclusively for Miner Weekly subscribers.

For a brief stretch this year, the market seemed willing to believe that every megawatt with a fence around it could become an AI factory.

Bitcoin miners no longer needed to talk only about hashprice, fleet efficiency or the next difficulty adjustment. They could talk about campuses, lease terms, hyperscalers, neoclouds, inference workloads and “critical IT load.” The same substations that once fed racks of ASICs were reintroduced to investors as scarce energy gateways into the artificial intelligence boom. In a power-constrained market, that story worked.

It worked so well that a new question is starting to matter: who got liquidity while the story was working?

That question is moving to the foreground as the TEM AI Infrastructure Growth Index, a basket tracking bitcoin miners, neoclouds, power suppliers and other companies tied to the physical build-out of AI infrastructure, has fallen 16% over the past month. The pullback does not erase the long-term argument for energy-backed compute. Nor does it suggest that recent insider sales or shareholder trims were improper. Many of the transactions were disclosed as prearranged trades under Rule 10b5-1 plans, which are designed to let insiders sell stock according to instructions set in advance.

But market optics change quickly. A planned sale during a rally can look routine. A planned sale followed by a sectorwide drawdown starts to look like a liquidity window.

The recent tape has given investors several examples to digest. Core Scientific’s legal chief sold shares as the company’s AI data center narrative helped lift the stock. Riot Platforms’ chief executive disclosed a prearranged sale after the miner’s stock rebounded. Tether trimmed Bitdeer exposure after buying during an earlier selloff and selling into a recovery. TeraWulf disclosed a new batch of share sales by its chief shortly before one of the most consequential AI lease announcements in the sector.

And at IREN, the controversy is less about insiders selling than about insiders being paid. The company’s board approved more than 18 million restricted stock units for its co-founder co-CEOs, adding a governance and dilution debate to a stock that had become one of the most visible winners of the miner-to-AI pivot.

Together, these episodes mark a shift in the AI infrastructure trade. Investors are no longer only asking which companies have power. They are asking who captures the economics, who absorbs the dilution, who keeps upside exposure and who monetized the rerating before the trade cooled.

TeraWulf Enters the Spotlight

TeraWulf offers the most vivid case study because the company remains one of the sector’s clearest AI-infrastructure rerating stories.

On June 29, Beowulf E&D Holdings, an entity managed by TeraWulf Chairman and Chief Executive Officer Paul Prager, disclosed a sale of 275,000 TeraWulf shares at a weighted average price of $26.596 per share, generating about $7.3 million in gross proceeds. The sale came one week before TeraWulf announced its 20-year AI infrastructure lease with Anthropic.

That June transaction was part of a broader run of disclosed sales by Prager and Beowulf E&D Holdings since late March. In total, Prager and the entity he manages sold about 1.59 million TeraWulf shares for roughly $32.7 million in gross proceeds, implying a weighted average sale price of about $20.55 per share.

Then on July 6, TeraWulf announced a 20-year lease with Anthropic at its Justified Data campus in Hawesville, Kentucky. The lease is expected to generate about $19 billion of contracted revenue over its initial term and support about 401 MW of critical IT load. TeraWulf also agreed to sell its 50.1% interest in the Abernathy joint venture to a Fluidstack-led investor group, monetizing an investment valued at about $450 million and giving the company capital to redeploy into wholly owned AI infrastructure projects.

That is the sort of transaction investors have been waiting for from power-rich miners: a long-term AI customer, a large contracted revenue figure and an argument that legacy mining infrastructure can be upgraded into a higher-multiple asset base.

It is also the kind of moment that makes insider liquidity worth watching.

Cipher, Riot and Core Scientific Show the Same Pattern

Cipher Digital adds the most recent example to the liquidity-window theme.

On July 8, Cipher CEO Tyler Page filed to sell 112,500 CIFR shares with a market value of about $2.38 million, implying an average price of $21.19. The sale was tied to a Rule 10b5-1 trading plan adopted on Dec. 19, 2025. Cipher previously disclosed that Page’s plan covered potential sales of up to 1.5 million shares through Dec. 24, 2026. The 112,500-share notice was 7.5% of the total 1.5 million-share ceiling under the plan.

Riot Platforms had its own version of the story. In May, CEO Jason Les sold 175,000 shares valued at about $4.2 million under a Rule 10b5-1 plan adopted in August 2025. On June 22, he sold another 250,000 shares with a market value of $7.03 million.

Core Scientific has been another focal point for the AI-mining crossover trade. The company emerged from bankruptcy in 2024 and has since repositioned itself around high-density colocation and AI infrastructure, while continuing to report a decline in self-mining revenue.

Core Scientific’s chief legal and administrative officer Todd DuChene filed on July 6 to sell 140,000 shares with a market value of $3.0 million. The planned sale followed 12 prior 10,000-share disposals since April 13, bringing disclosed sales under the plan to about 260,000 shares and $5.9 million in gross proceeds.

These are important caveats. Rule 10b5-1 refers to prearranged trading plans designed to separate insider transactions from later corporate developments, and sales by executives with large equity holdings can reflect diversification, taxes or personal liquidity rather than a negative view of the company. It is not a confession of bearishness. Executives with large stock-heavy compensation packages often sell even when they remain optimistic about a company.

But public markets do not only process legality. They process alignment. When executives sell after a stock has rerated on AI expectations, and the sector then pulls back, investors start asking whether the balance of risk and reward has shifted from insiders to the public float.

The liquidity-window theme is not limited to executives.

Tether’s recent Bitdeer transactions show how strategic holders also used the AI-mining rebound to reduce exposure. As TheEnergyMag reported, Tether trimmed its Bitdeer positions at an average price of around $20 in early June, after buying into Bitdeer for $8.85 apiece during a market selloff earlier this year. While Tether remained one of Bitdeer’s largest shareholders, the trade still fits the pattern: buy into weakness, trim into the AI rerating, and retain a large enough position to keep participating if the story continues.

IREN Adds the Governance Layer

IREN brings a different but related issue into focus.

The company has become one of the most closely watched AI infrastructure names after moving beyond bitcoin mining and pursuing large-scale AI cloud and data center opportunities. But its latest compensation disclosure triggered a backlash among some retail investors and market commentators.

On June 30, IREN’s board approved grants of 9,099,328 restricted stock units each to co-CEOs William Roberts and Daniel Roberts. The awards are subject to a combined six-year vesting and holding period. The company said neither co-CEO will receive another equity incentive grant until fiscal 2031, and that the awards were designed to retain and incentivize the executives through IREN’s next phase of growth.

That explanation did not quiet the debate. Critics focused on the size of the package, its dilution and the fact that the company is still in the middle of proving that its AI infrastructure strategy can generate durable returns. IREN shares fell sharply as governance concerns met a broader selloff in AI-related stocks.

The IREN episode is not an insider-sale story. It is arguably more important: a debate over how much of the AI infrastructure upside founders and executives should receive before the business model has fully matured.

This is where the AI infrastructure trade begins to resemble other capital-intensive booms. The first phase of the rally was about scarcity. The next phase is about governance, capital discipline and execution. In that phase, disclosed insider sales, strategic-holder trims and large founder equity grants become part of the same story. They tell investors where the private incentives sit inside a public-market boom.

Hardware and Infrastructure News

• TeraWulf Surges on $19 Billion Anthropic AI Lease, $450 Million Fluidstack JV Exit

• Bitzero Secures 33-Hectare Land Reservation in Finland for 60MW Data Center

• Galaxy Digital Completes Phase I of Helios Data Center, Delivers 133 MW to CoreWeave

• Bitdeer Breaks Ground on Nevada Factory Despite Bitcoin Hashprice Squeeze

• MARA Buys Texas Site for AI, Bitcoin Mining in Deal Worth Up to $600M

Corporate News

• Keel Hires Digital Realty Veteran Ganesh Aiyer to Drive AI Infrastructure Growth

• Nscale, Nordkraft Form Venture to Run Narvik AI Data Center

Financial News

• Crusoe Reportedly in Talks to Raise $3 Billion at About $30 Billion Valuation

• Core Scientific Legal Chief Sells $5.9 Million Amid AI Rally Boost

• Nscale Secures $900 Million Credit Line to Fund AI Data Center Expansion

• Cipher CEO Files to Sell $2.4M of Stock as AI Rally Triggers Trading Plan

Feature

• EIF Speaker Series: “Bitcoin Is Becoming Increasingly Boring” with Lisa Hough

• This $13 Billion Georgia Family Is Helping Electrify Data Centers - Forbes

Email us

EIF Dallas (July 23) is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Get your tickets today and use code MNRW30 for 30% off exclusively for Miner Weekly subscribers. You can also explore sponsorship opportunities to elevate your brand and connect with leaders across energy, infrastructure, and capital.

Bitcoin miners have long defended their power use with a simple argument: mining may be energy-intensive, but it is unusually interruptible.

A mining facility can curtail quickly when power prices spike or the grid is under stress. It can monetize surplus electricity when power is abundant, then shut down without disrupting households, factories or customer-facing software. That flexibility has become central to the industry’s pitch to utilities, regulators and investors, especially in markets such as Texas where demand response has become part of the mining business model.

But a new line of research raises a provocative question for the AI-infrastructure boom: what if GPU data centers can learn a version of the same trick?

A paper posted to arXiv in late June, titled “Power-Flexible AI Data Centers: A New Paradigm for Grid-Responsive Compute,” argues that AI data centers should not always be modeled as static, inflexible power loads. The authors describe a software architecture that combines grid signals, workload scheduling and power telemetry to reduce, defer or shift AI workloads in response to power-system conditions.

In a real-world deployment on a 130 kW GPU cluster, the paper says the system demonstrated rapid load reduction, sustained curtailment, carbon-aware operation and geographically distributed load shifting while preserving priority service levels.

The headline results were notable for the speed and range of the claimed response. The authors said the system reached 100% compliance across more than 200 National Grid power events in the UK test environment. In a simulated emergency dispatch modeled on a historical grid contingency, the cluster reduced power use by about 30% within 40 seconds. In another fast-response test, it cut load by 40% within roughly one minute. The paper also reported sustained curtailment of 10% to 40% for periods ranging from two to 10 hours, as well as a U.S. geo-shifting test in which 10% of live inference traffic was shifted from Oracle GPUs in Virginia to Illinois.

The paper is not yet peer reviewed. The arXiv version says it has been submitted to IEEE for possible publication, so its findings should be treated as early technical evidence rather than settled academic consensus. Still, the idea is not emerging in isolation. The June paper builds on a related Nature Energy article, “AI data centres as grid-interactive assets,” which reported a field demonstration of software-based demand response at an AI cluster in Phoenix, Arizona. A Nature Reviews Electrical Engineering summary of that work said the researchers showed that data centers can function as flexible, grid-aware loads by adjusting computing workloads in real time rather than relying on new hardware or on-site batteries.

The Phoenix trial is important because it moves the discussion beyond simulation.

According to the preprint version of the Nature Energy work, the test was conducted at an Oracle Phoenix Region Cloud data center on a 256-GPU cluster built with NVIDIA A100 Tensor Core GPUs. The system used Databricks MosaicML for workload orchestration, Weights & Biases for telemetry, and Amperon’s grid-demand forecasting tools. The project involved Emerald AI, NVIDIA, Salt River Project and the Electric Power Research Institute, and was conducted through EPRI’s DCFlex initiative.

The cluster ran representative AI workload mixes, including training, inference and fine-tuning jobs. The researchers classified jobs into flexibility tiers based on how much performance reduction users could tolerate: strict jobs with no performance reduction, and other tiers allowing average throughput reductions of up to 10%, 25% or 50% over a three- to six-hour period. That matters because the system did not simply shut everything off. It selectively slowed, paused or power-capped jobs depending on their tolerance for delay.

The reported result was a 25% reduction in cluster power usage for three hours during peak grid events while maintaining quality-of-service guarantees. The authors said the reduction was achieved without hardware retrofits or battery storage, using workload orchestration and GPU power-limiting techniques such as dynamic voltage and frequency scaling, combined with job pausing.

That does not mean an AI data center is suddenly as flexible as a bitcoin mine.

The Phoenix test involved a 256-GPU cluster, not a 100 MW or 1 GW AI campus. The later arXiv paper’s 130 kW deployment is also small relative to the scale now being discussed by hyperscalers, utilities and public miners. Scaling this approach from a cluster to an entire campus would require coordination across power delivery, cooling, networking, customer contracts and service-level agreements. It would also depend heavily on workload mix.

The researchers acknowledge that limitation. Not all AI workloads are temporally flexible. Batch training, fine-tuning and some inference jobs may be slowed or paused. But strict “Flex 0” jobs — such as real-time inference, streaming and model serving — were not modified in the Phoenix demonstration. Future flexibility may require geographic load shifting, where workloads move between regions to avoid grid stress without creating unacceptable latency or performance penalties.

Even with those caveats, the research cuts into one of bitcoin mining’s most important narratives.

For years, miners have argued that they are uniquely useful to power grids because they can behave less like a conventional industrial load and more like a controllable energy buyer.

AI data centers are more complicated. They are built around expensive GPUs, customer commitments, uptime expectations and high utilization targets. Operators do not want billions of dollars of accelerators sitting idle. But the research suggests the choice may not be binary. An AI facility may not need to run flat out or shut off entirely. It may be able to preserve high-priority workloads while throttling lower-priority jobs, delaying flexible training runs, power-capping GPUs or shifting traffic to another region.

That is the miner-relevant point. Yet there are still reasons for skepticism.

The research has commercial backers like Emerald AI, NVIDIA, Oracle and power-sector partners, with a clear interest in showing that AI infrastructure can be made more grid-friendly. The evidence remains early, and the field tests are far smaller than the campuses now being planned. Utilities will also need market rules, telemetry standards and compensation structures before AI data centers can become meaningful demand-response resources. Customer willingness is another open question: not every AI user will accept slower jobs in exchange for lower power costs or faster interconnection.

But the debate is worth watching.

Bitcoin miners helped popularize the idea that digital infrastructure could act as flexible load. Now AI data centers are beginning to move into that same communication, policy, and engineering conversation.

If the research scales, mining’s grid edge may not disappear, but it may become less unique. The flexibility pitch is no longer just a bitcoin mining story. It is becoming an AI-infrastructure story too.

Regulation News

• DOE Lets PJM Tap AI Data Center Backup Power During Heat Wave

Hardware and Infrastructure News

• Bitdeer Signs Conditional Lease for Norway AI Data Center Site

• Blockfusion Touts Up to $5.4B From AI Lease as SPAC Deal Advances

• Cargo thieves have set their sights on data center supplies

• Bloom, Brookfield Expand AI Power Financing Partnership to $25 Billion

Corporate News

• IREN, Keel Join Russell Indexes as AI Infrastructure Stocks Draw Broader Attention

• Ionic Digital Raises $400M Ahead of Nasdaq Listing as AI Revenue Overtakes Bitcoin Mining

• Sphere 3D Seeks DarkHorse Rebrand as It Pivots Toward AI Infrastructure

• IREN Adds Oracle, Google Veterans to AI Cloud Leadership Team

Financial News

• Applied Digital Expands Credit Facility to $430M as AI Data Center Buildout Continues

• American Bitcoin Sets Reverse Stock Split as Sub-$1 Nasdaq Risk Builds

Email us

EIF Dallas (July 23) is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Get your tickets today and use code MNRW30 for 30% off exclusively for Miner Weekly subscribers. You can also explore sponsorship opportunities to elevate your brand and connect with leaders across energy, infrastructure, and capital.

There are moments in proof-of-work mining when the market briefly reminds everyone that electricity is not tied to a single digital asset, or even a business model.

The same megawatt can secure bitcoin, mine Zcash, chase Kaspa rewards or support an AI/HPC workload. Which one makes the most sense increasingly depends less on ideology than on a moving set of prices: token values, network difficulty, machine efficiency, hosting rates and power costs.

For the past year, public-market investors have largely rewarded the AI side of that equation. Bitcoin miners with large power portfolios have been repriced on the idea that their sites may be more valuable as data center infrastructure than as pure mining operations.

But Zcash’s rally, after years on the sidelines, has complicated that narrative.

The price appreciation of ZEC since late last year, followed by renewed hashrate momentum on the network, has made Equihash mining economics suddenly relevant again. In some cases, the revenue per unit of power now compares favorably not only with other proof-of-work assets, but also with estimated HPC colocation economics.

That backdrop helps explain why Fortitude Mining’s plan to go public is arriving at an interesting time.

Fortitude, a Zcash-focused digital asset mining platform under Digital Curreny Group, announced this week that it plans to list on Nasdaq through an all-stock merger with HeartSciences. The proposed company would operate under the Fortitude brand and is expected to trade under the ticker TUDE, subject to Nasdaq approval.

The transaction is still light on detailed financial disclosure. But Fortitude has released enough operating data to make the broader point clear: this is not just another crypto miner trying to attach itself to the AI trade. It is a test of whether public investors still have an appetite for a crypto-native compute business when the economics of a non-bitcoin proof-of-work network suddenly improve.

A Revenue-per-Megawatt Lens

TheEnergyMag’s latest analysis compares revenue across several proof-of-work mining rigs and a conservative estimated HPC colocation benchmark on a common basis: dollars per megawatt-hour.

The result is a useful snapshot of the compute market’s current hierarchy.

On the high end, a Zcash-focused Z15 Pro generates an estimated $373 per MWh, far above the $223 per MWh conservative estimate for HPC colocation. A next-generation bitcoin miner, the S23 Pro, comes in at about $133 per MWh, while the S21 Pro produces roughly $84 per MWh. By comparison, the DG1+ and Kaspa-focused KS5 and KS3 models generate about $65, $48 and $19 per MWh, respectively, under the assumptions used in the analysis.

That spread matters because it cuts through the usual framing of bitcoin mining versus AI infrastructure. The more precise comparison is not crypto versus AI. It is compute revenue versus power cost, with different machines and workloads competing for the same scarce input: electricity.

Over the past year, the dominant market narrative has been that bitcoin miners should redirect power toward AI and high-performance computing whenever possible. That logic is easy to understand. AI demand has created a scarcity premium for energized land, grid interconnection rights and large-scale power capacity.

But the revenue comparison shows the decision is not always straightforward.

HPC contracts may offer longer duration, lower volatility and stronger visibility than mining, provided that the business development and execution process is done well. That makes them attractive from a financing perspective, especially for public companies trying to reduce dependence on daily crypto-market swings.

Mining, however, retains optionality. When a proof-of-work asset rallies substantially faster than its network difficulty adjusts, mining revenue can temporarily exceed what infrastructure owners might earn from contracted HPC hosting.

Zcash is currently an example of that dynamic. Zcash has historically occupied a smaller niche than bitcoin, with less liquidity, fewer institutional narratives and a narrower base of industrial-scale miners. But the recent ZEC price rally has changed the revenue math for Equihash ASICs.

That does not mean Zcash mining is more valuable than HPC infrastructure on a risk-adjusted basis. Just search Zcash’s price crash and fast rebound earlier this month. Nonetheless, it perfectly shows that crypto-native compute can still produce bursts of exceptional returns, particularly in smaller networks where price and hashrate can move out of sync.

That creates both opportunity and risk for companies like Fortitude. A Zcash-focused platform may enjoy stronger near-term mining economics when ZEC outperforms, but it is also more exposed to the market volatility and the speed at which network difficulty can adjust and erode those gains.

Fortitude’s Timing

Fortitude’s public-market plan lands directly into that debate.

The company has not yet released detailed financials, leaving investors without a full view of its revenue, cost structure, power contracts or margins. But the operating metrics it has disclosed suggest it is not a small exploratory Zcash operation.

Fortitude said it began mining ZEC, the native token of the Zcash network, in 2019 and has since scaled production to an annualized run rate of 157,000 ZEC. Based on ZEC’s current price of about $400, that production rate implies a static annualized revenue of about $62 million.

Either way, the disclosed production rate gives Fortitude a meaningful operating base in a network that remains far smaller than bitcoin but has recently become more economically attractive for specialized miners. The timing is important: ZEC’s price appreciation has improved mining revenue for Equihash ASICs, while rising hashrate suggests more machines are being pointed at the network to capture those returns.

That dynamic gives Fortitude a more differentiated story than public bitcoin miners that are increasingly trying to convince investors their megawatts can be repurposed for AI. Fortitude is entering the market as a crypto-native infrastructure company with direct exposure to Zcash economics.

That identity may help it stand out, but it also narrows the questions investors will ask. Without detailed financials, the key issues will be how much of Fortitude’s revenue depends on current ZEC prices, how quickly network difficulty is rising, what its power costs look like, and whether its fleet can remain competitive as more Equihash capacity comes online.

In that sense, Fortitude’s planned listing could become a useful public-market test case for non-bitcoin proof-of-work exposure. If Zcash economics remain strong, the company may find investor appetite for a miner tied to a differentiated asset. If margins normalize quickly, the listing may instead reinforce how difficult it is for crypto-native miners to turn cyclical revenue spikes into durable infrastructure value.

Regulation News

• US to Fast-Track AI Data Centers as Power Bottleneck Becomes National Priority

• Texas Regulators Approve ERCOT Batch Process for AI Data Center Power Requests

• Thailand seeks to arrest fugitive Chinese businessman over illegal crypto mining

Hardware and Infrastructure News

• Microsoft to Build 2GW AI Data Center Campus in West Texas With Onsite Gas Power

• Constellation, Walmart Sign Nuclear Power Deal for Illinois Supply Chain Expansion

• Bitzero Adds Finland Land Reservation to Nordic AI Data Center Pipeline

• Galaxy Digital Advances $400 Million AI Data Center Plan in McGregor, Texas

• Sphere 3D Signs 30 MW Bitcoin Co-Mining Deal With Bitdeer

Corporate News

• Hut 8 Investors Reach $2.35 Million Settlement Over US Bitcoin Merger Claims

• Riot CEO Sells Another $7M as Bitcoin Miner Briefly Tops $30 on AI Rally

• DOE Announces $17.5 Billion Loan Commitment to Revive American Nuclear Supply Chain

• CoreWeave Expands Swedish Presence with Conapto Data Center Deal

• Bitcoin Miners Move to Defend Exchange Listings as Sub-$1 Shares Persist

Financial News

• DCG’s Fortitude Mining to Go Public Through HeartSciences Merger

• Cerebras Systems Reports Q1 2026 Revenue of $193.4 Million

• HIVE Lines Up Swedish AI Lease as It Seeks $100 Million Note Offering

Feature

• EIF Speaker Series: ‘AI’s Real Buildout Has Barely Begun’ with NG Zhang

• EIF Speaker Series: ‘Inference Could Shift the AI Compute Mix’ with Paul Golding

Email us

EIF Dallas (July 23) is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Get your tickets today and use code MNRW30 for 30% off exclusively for Miner Weekly subscribers. You can also explore sponsorship opportunities to elevate your brand and connect with leaders across energy, infrastructure, and capital.

Bitcoin’s latest difficulty adjustment was a reminder that the network still runs on a simple rule: when mining economics weaken enough, hashrate leaves.

On June 14, Bitcoin’s mining difficulty fell by 10.09% to 124.93 trillion, one of the largest downward adjustments in the network’s history by percentage terms. The drop followed a two-week period of slower block production, signaling that a meaningful amount of computing power had come offline.

But the scale matters as much as the percentage. Bitcoin’s hashrate had been hovering near the 1 ZH/s mark before the decline. At that level, a 10% drawdown implies roughly 100 EH/s of computing power was taken offline during a single difficulty epoch. For context, Bitcoin’s entire network hashrate was only around 200 EH/s four years ago. What disappeared over the past two weeks was roughly half of the total computing power securing the network at that time.

Several factors likely contributed. Bitcoin’s early-June price decline pushed hashprice back toward levels that leave less-efficient fleets near breakeven. Texas miners entered ERCOT’s 4CP summer season, when curtailing during a few peak-demand intervals can materially reduce annual transmission charges.

And across North America, another structural force is becoming harder to ignore: some of the largest public miners are no longer trying to maximize bitcoin hashrate at all costs. They are trying to decide how much of their power should still be used for bitcoin mining.

That distinction matters. If miners are merely unplugging older ASICs because hashprice is weak, the machines can come back when economics improve. But if a site is being retrofitted for high-performance computing, reserved for an AI tenant or financed as a future data center campus, that power may not return to Bitcoin mining in the next bull market.

The latest difficulty drop is therefore more than a network statistic. It is part of a broader reallocation now underway in the mining sector. Bitcoin miners are losing hashrate in one part of the market while gaining a new valuation framework in another.

The Funding Gap Behind the Repricing

This is where the latest difficulty drop connects with the equity market.

The market has already started to value certain public miners less as pure bitcoin producers and more as owners of scarce power infrastructure. That shift — from hashrate to megawatts — is familiar by now. The larger question is no longer whether miners control assets that AI developers want. Many of them do.

The harder question is who can afford to turn those assets into AI data centers.

That is the reality check behind VanEck’s latest framework for valuing bitcoin miners as AI infrastructure. The report argues that miners pursuing AI/HPC opportunities should be evaluated around energized power, execution risk and tenant quality. It also estimates that miners pursuing the transition face a roughly $50 billion near-term funding gap.

The number captures the central tension of the trade. Public miners have been rewarded for owning scarce power at a time when AI developers are desperate for it. But AI data centers require far more capital than bitcoin mining sites. A bitcoin mine can run with relatively simple buildings, modular infrastructure and ASIC fleets that tolerate fast curtailment. AI and HPC facilities require higher standards for uptime, cooling, electrical redundancy, networking and customer support.

That means the next phase of the trade will likely be more selective. Investors are moving from rewarding AI optionality to underwriting AI execution.

A miner with a large undeveloped power portfolio may still have significant strategic value, but it needs capital, customers and time. A miner with a smaller but contracted and financed project may deserve a different multiple. The distinction is between AI option value and AI execution value — and the gap between the two may be where the next repricing happens.

Why the Index Matters

This is also the backdrop for TheEnergyMag’s AI Infrastructure Growth Index.

The index is being built around a market shift that is already visible but still difficult to isolate. Public bitcoin miners are no longer trading only as bitcoin beta. Some are becoming power-backed AI infrastructure bets. Others remain primarily mining companies with optionality. A few sit in between, with operating hashrate, large power portfolios, early AI contracts and unresolved funding needs.

The point of the index is not to treat every miner with an AI slide deck as an AI infrastructure company. It is to track the companies sitting at the fault line between bitcoin mining, power markets and AI compute demand.

That includes miners with energized sites and credible conversion paths, AI infrastructure platforms whose growth depends on power access, and developers assembling land, substations and interconnection rights ahead of future demand.

The index is less a claim that miners have already become AI companies than a way to monitor how much of the market is starting to value them that way.

The Hashrate Question

There is also a consequence for Bitcoin.

The network does not reward miners for holding future AI development rights. It only rewards active hashrate. If North America’s largest public miners increasingly allocate their best sites to AI or HPC, Bitcoin does not get credit for their power portfolios. It only gets the machines that remain online.

That does not mean the AI pivot caused the latest difficulty decline by itself. It did not. Weak hashprice, older machines, seasonal curtailment and power-market dynamics all played a role. But the AI pivot changes how the next hashrate recovery may unfold.

In previous cycles, weaker mining economics pushed inefficient machines offline. When conditions improved, operators could redeploy capital into new ASICs and bring hashrate back. This cycle is different because the alternative use case for power is more compelling.

AI tenants can support longer contracts and potentially more stable revenue than mining. For public miners under pressure to improve multiples and diversify revenue, the opportunity cost of dedicating power to bitcoin mining has gone up.

That raises a strategic question: if the most institutionally visible North American miners keep shifting toward AI infrastructure, who absorbs Bitcoin’s next wave of hashrate growth, if there’s going to be growth at all?

The answer may be private miners that face less pressure to chase AI narratives. It may be lower-cost international power markets. It may be sovereign-backed or stranded-energy projects. Or it may be hybrid operators that keep mining as a bridge load while waiting for AI development to catch up.

What is clear is that the public-miner cycle is changing. The sector is no longer moving only between bitcoin bull markets and mining-margin downturns. It is being pulled into the larger AI infrastructure cycle, where power access, capital formation and data center execution matter as much as ASIC efficiency.

The market has already started repricing bitcoin miners for the AI boom. The next phase will test whether that repricing is supported by funded projects, contracted tenants and delivered capacity — or whether some of it was simply a higher multiple placed on undeveloped power.

Bitcoin’s latest difficulty decline shows that hashrate remains the foundation of the mining business. But the equity market is increasingly looking beyond hashrate.

Regulation News

• North Tonawanda Extends Data Center Pause as AI Conversion Plan Faces Pushback

Hardware and Infrastructure News

• Bitcoin Mining Difficulty Set for Steep Drop as Hashrate Slides After Price Crash

• Avista Pauses Talks Over 500 MW Data Center Power Request After Community Pushback

• Sovright Launches Zcash Mining Pool Testnet With Shielded Payouts

• Bitdeer’s 750MW Ohio AI Data Center Plan Draws Local Scrutiny

• HIVE Wins Approval to Buy 32 MW Sweden Data Center as AI Push Expands

• HIVE Signs $220M AI Cloud Deal With Bell, Cohere for Canadian Sovereign Stack

Corporate News

• Hut 8 Names Former Merrill Lynch CEO Stan O’Neal as Board Chair

• IREN Completes Acquisition of Spanish AI Data Center Developer Nostrum

• Sangha Renewables Weighs M&A for Texas Bitcoin Mine as AI Demand Lifts Power Assets

• SWI Group Acquires $500 Million Stake in Genesis Digital Assets to Reposition Portfolio for AI

• Northern Data CEO to Step Down as AI Infrastructure Firm Advances Rumble Deal

• Cipher Names Former ERCOT Executive as Head of Grid Strategies

• Oman Launches National Bitcoin Mining Pool Omanhash

Financial News

• Dominion Energy Announces $1.5B Junior Subordinated Notes Offering

• OpenAI’s Spending Reportedly Hit $34 Billion in 2025 as AI Race Intensified

• Tether Trims Bitdeer Stake After AI Push Lifts Bitcoin Mining Stock

• NVIDIA Prices $25B Multi-Tranche Senior Notes Offering

Feature

• Two mayors, one $10 billion AI data center, and a growing divide in small-town Texas - Fortune

• EIF Speaker Series: “Bitcoin the Ultimate Flexible Consumer of Energy” with Sean McDonough

Email us

EIF (July 23, Dallas) is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Get your tickets today and use code MNRW30 for 30% off exclusively for Miner Weekly subscribers. You can also explore sponsorship opportunities to elevate your brand and connect with leaders across energy, infrastructure, and capital.

ERCOT’s Four Coincident Peak (4CP) mechanism is not new. For years, large loads in Texas have been able to lower their transmission cost exposure by reducing demand during four 15-minute intervals: the highest systemwide demand interval in each summer month from June through September.

But the old rule is becoming newly important.

Texas is preparing for a much larger transmission buildout as the grid absorbs unprecedented load growth. As detailed in a previous issue, ERCOT witnessed an exponential volume of interconnection requests from large load customers, in particular, data centers. Meanwhile, according to TheEnergyMag, North American public bitcoin miners alone had more than 5.1 GW of operational power capacity in Texas as of Q1.

For these large-load customers, even a small change in transmission cost exposure can translate into a meaningful sector-wide operating cost swing. That is why operators with large but flexible loads are incentivized to navigate the 4CP intervals by curtailing their activities. But the influx of interconnection requests since 2025 has apparently prompted policymakers to review whether large flexible loads are avoiding too much of the cost by curtailing during a narrow set of summer peaks.

The debate also comes at a difficult point for bitcoin miners, the most natural flexible load operators. Bitcoin’s mining revenue remains compressed as the network difficulty remains high. In that environment, every component of the power stack matters: energy price, uptime, curtailment revenue, demand charges, ancillary-service pass-throughs, and transmission charges.

That is why the 2026 4CP season matters even before Texas finalizes any rule change. And if the state eventually moves from the current 4CP framework toward a 12CP design, the problem could become harder. Instead of defending four summer intervals, miners and other large load customers may have to manage coincident-peak exposure across all 12 months of the year.

What 4CP is

ERCOT’s 4CP mechanism allocates transmission costs to certain large consumers based on their demand during four specific systemwide peaks. The relevant moments are the single highest 15-minute demand intervals in June, July, August and September. A large load’s average demand across those intervals becomes the basis for its transmission cost allocation in the following year.

A facility can run profitably through most low-price hours, but if it is consuming power during the wrong 15-minute interval, that load can become embedded in its fixed-cost structure for the next 12 months. Conversely, if the site is curtailed during the coincident peak, it can sharply reduce that exposure.

That is why 4CP has become a summer ritual in ERCOT. Operators watch heat, system load, day-ahead and real-time demand signals, solar output, battery charging, DC tie flows and ERCOT forecasts. They do not know the official 4CP interval in real time. They only know it after the fact. So the decision is probabilistic: curtail early enough to be offline if the interval becomes the monthly peak, but not so often that the cure becomes more expensive than the disease.

LoD’s review of the 2025 season offers a useful illustration. All four coincident peaks last summer landed on weekdays in a narrow late-afternoon window: June 19 at 5:00 p.m., July 30 at 5:00 p.m., August 18 at 5:00 p.m. and September 4 at 5:30 p.m.

The timing looked predictable in hindsight. The difficulty is that each month contained multiple hot days and broad afternoon demand plateaus where several 15-minute intervals were close enough to plausibly become the peak. “Around 5 p.m.” is not precise enough when a quarter-hour interval can set a year of transmission costs.

The 2026 4CP watch has already started

Luxor said in a LinkedIn update that June’s early high-water mark had reached 81.3 GW on the first day of the season last week, prompting the first curtailment dispatch of 2026 for Luxor Energy members. The next day, ERCOT load peaked near 80.0 GW around 3 p.m. CT, close enough to the prior mark to trigger a second dispatch from Luxor’s 4CP engine.

The next test is coming soon. ERCOT’s public dashboard showed demand forecast at 80,343 MW around 6:00 p.m. CT on June 11.

None of those data points means that any specific interval will ultimately become June’s 4CP interval. The official coincident peak is only known after the fact. But the early-June data shows how quickly the monthly peak-defense problem can become active. Even before Texas reaches the hottest stretch of summer, ERCOT can produce demand levels high enough to force curtailment decisions.

For miners, the decision is not simply whether power prices are high enough to justify shutting down. LoD’s 2025 review showed that the largest demand moment of the year occurred around 84 GW while power at that moment settled near $25/MWh. Prices rose later in the evening. That means the 4CP interval can be one of the most financially important moments of the month even when real-time energy prices are not flashing red.

This is the central 4CP tension. Curtailing for 4CP is not about dodging expensive power. It is about giving up potentially profitable runtime to avoid a transmission charge that may not be confirmed until after the season ends.

Why the rule is under pressure

4CP was designed to give large loads an incentive to reduce demand during the system’s summer peaks. In theory, that reduces the need to build transmission infrastructure that only serves a few extreme demand intervals.

But the Texas grid has changed.

Solar generation has shifted parts of ERCOT’s daily risk profile. The system’s highest demand can occur while solar is still contributing enough generation to keep real-time prices relatively low. The more stressful price intervals can arrive later, as solar output fades and evening demand remains elevated.

At the same time, large flexible loads have become better at predicting and avoiding 4CP intervals. That has raised a policy question: if data centers, bitcoin mines and other large loads are part of the reason Texas needs new transmission investment, should they be able to avoid a large share of transmission charges by curtailing during four 15-minute summer intervals?

That is the backdrop for the 4CP-to-12CP discussion. A 12CP design would generally spread the cost-allocation signal across the highest-demand interval in each month of the year, rather than only the four summer months. Depending on the final design, the proposal could also involve a longer measurement interval.

The policy goal would be to make cost allocation broader and harder to avoid, while preserving some incentive for large loads to reduce demand when the grid is strained.

Why transmission cost matters now

Transmission cost is not the same thing as wholesale energy price. Operators can buy power at attractive headline prices and still carry a meaningful all-in power cost once transmission, congestion, retail adders, ancillary services, demand charges and other pass-through items are included.

LoD estimated full 4CP exposure at roughly $50,000 per MW per year, or about $1.25 million annually for a 25 MW site. For a continuously running load, that converts to roughly $5.70/MWh before considering site-specific tariffs, retail structures, hedges, behind-the-meter arrangements, or curtailment behavior.

That may not sound large in isolation. But it becomes meaningful when scaled across a fleet and when margins are thin, especially for bitcoin miners today.

In a high-hash-price market, operators can tolerate more inefficiency. A miner may still care about 4CP, but a few extra dollars per MWh or some unnecessary curtailment can be absorbed by strong mining revenue.

That cushion is much thinner today with Bitcoin’s mining revenue per unit of compute staying near all-time lows.

The 4CP decision also interacts with hashprice in a direct way. Curtailment is not free. Every hour offline means forgone bitcoin production. When hashprice is high, the opportunity cost of curtailment is high. When hashprice is low, the relative value of avoiding a fixed transmission charge rises.

That creates a narrow optimization problem. Large mining sites are flexible loads, but flexibility only has value if it can be controlled precisely. Shutting down too late misses the transmission benefit. Shutting down too early or too often sacrifices uptime. Repeated stop-start cycles also introduce operational complexity, thermal cycling and potential wear.

In a 12CP world, that only gets more complicated.

For Texas miners, 4CP has always been a hidden margin lever. In 2026, it is becoming something larger: a test of whether operators can monetize load flexibility without sacrificing too much uptime, and whether the Texas market will continue to reward that flexibility under the same rule structure.

Hardware and Infrastructure News

• Nebius to Invest £1.7 Billion in UK AI Compute Expansion

• Applied Digital Signs $5.2 Billion AI Data Center Lease for New Southern Campus

• Soluna Formalizes Metrobloks JV for 350 MW Texas AI Data Center Campus

• Crusoe Touts Contracted AI Infrastructure Capacity Nears 5 GW

• OpenAI Weighs 10-Gigawatt Ohio Data Center Lease in Potential Nvidia-Backed Deal

• Texas Adds 456 MW of Gas-Fired Power in Houston as ERCOT Braces for Summer Demand

Corporate News

• Keel Infrastructure to raise $350 million through convertible debt offering

• Dominion Energy Agrees to $825 Million Senior Notes Issuance

• CoreWeave Files Automatic Shelf Registration for Multi-Security Offerings

• Bitzero Wins Nasdaq Listing Approval as AI Infrastructure Push Expands

• Bitdeer Discloses Leadership Shake-Up Amid Deeper Push Into AI Infrastructure

• OpenAI Files Confidentially for IPO, Following Anthropic’s Move Toward Public Markets

• Sphere 3D Taps EA Advisors as Post-Merger AI Push Takes Shape

• KKR, Nvidia and Vistra Launch $10 Billion AI Infrastructure Platform Helix

Financial News

• Hut 8 Plans $4.25B Debt Raise for Texas AI Data Center

• TransAlta Prices C$350 Million Equity Offering for Colorado Gas Asset Acquisition

• Cipher Digital Seeks $810M Debt Raise to Fund Stingray Data Center

• Applied Digital Secures $350M Revolver as CoreWeave Lease Move Takes Shape

• Applied Digital Seeks $1.59 Billion Debt Raise for Fourth Ellendale AI Data Center

• CoreWeave Eyes $3.5B Debt Raise as AI Buildout Stretches Balance Sheet

Email us

Energy Investors Forum (July 23, Dallas) is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Get your tickets today and use code MNRW30 for 30% off exclusively for Miner Weekly subscribers. You can also explore sponsorship opportunities to elevate your brand and connect with leaders across energy, infrastructure, and capital.

Bitcoin has slumped over the past several days, extending its decline this year. But many of the largest public bitcoin mining stocks have moved in the opposite direction, underscoring a growing split between bitcoin and the companies that were once viewed as one of the most leveraged ways to trade it.

The divergence marks a notable change from the 2021 cycle, when listed miners were largely treated as high-beta proxies for bitcoin. This time, the market is increasingly rewarding miners with credible AI data-center strategies, large power portfolios or sites that can be converted from bitcoin mining to high-density computing.

The irony is hard to miss: bitcoin mining stocks are rallying because investors are starting to care less about bitcoin mining.

A basket of major public miners tracked by TheEnergyMag has broadly outperformed bitcoin since January, with the most notable divergence beginning to emerge in early May. Bitcoin is down about 28% year-to-date, while several mining and mining-adjacent equities are up sharply. Hut 8, Keel, TeraWulf, Riot Platforms, Core Scientific, Applied Digital, HIVE, Bitdeer, Cipher Mining, IREN, CleanSpark and MARA are all trading higher. Hut 8 and Keel have more than doubled.

But the rally is not uniform across the bitcoin mining complex. The market has not stopped treating all bitcoin miners as bitcoin proxies. It has stopped treating some of them that way.

American Bitcoin, Cango and Canaan are all down over the same period, even as many larger mining peers have rallied. These names are still trading more like bitcoin-mining stocks in the old 2021 sense: levered bets on bitcoin and mining economics. When bitcoin falls, their shares fall harder. When hashprice weakens, they have fewer alternative narratives to absorb the pressure. Plus, they have significantly smaller market capitalization, which makes them more vulnerable to bitcoin’s price swings.

The result is a two-speed market. One group is still being valued as bitcoin beta. The other is being recast as a backdoor way to own AI’s physical infrastructure buildout.

The broader equity market backdrop is reinforcing that shift. AI-themed stocks far beyond the mining sector have also surged. Micron, Nebius, CoreWeave and Coherent have all been swept into the same AI-infrastructure rally, reflecting investor demand for exposure to memory, cloud capacity, optical components, data centers and compute supply chains. In that context, the strongest mining stocks are not rallying in isolation. They are being pulled into a much larger trade around the physical bottlenecks of artificial intelligence.

That trade is built on a simple premise: AI needs power, space, fiber, cooling and time-to-market. Bitcoin miners happen to own some of those ingredients.

The result is a strange inversion. Bitcoin miners used to be valued for their ability to turn electricity into bitcoin. Now, some are being valued for their ability to stop doing that.

This does not mean the whole sector has successfully crossed over into AI. The actual conversion from bitcoin mining to high-performance computing remains difficult, expensive and uneven. Some miners may lack the locations, fiber access or reliability profile needed to attract AI tenants. Others may have the power but not the capital required for data-center upgrades. Some may use AI language more aggressively than their actual project pipeline justifies.

That makes the current rally both rational and speculative, and shows why the old sector labels are becoming less useful. Traditional AI indexes tend to be dominated by the largest chipmakers, cloud platforms and established technology names. But the AI buildout is also creating a more specialized group of infrastructure-exposed growth companies: power-rich bitcoin miners, data-center developers, compute providers and other listed names tied to the physical constraints of AI.

The question is not just which companies sell chips or train models. It is which companies control the bottlenecks — power, land, interconnection, cooling and time-to-market — that determine how quickly AI capacity can actually be built.

For now, capital appears to be rotating toward that bottleneck and away from bitcoin itself. That does not mean bitcoin is irrelevant. But in public equities, the center of gravity has shifted. Bitcoin is no longer the only source of excitement, and in recent weeks, it has not been the main one.

The phrase “bitcoin mining stock” may now describe a company’s history better than its valuation.

The final test may come from outside the mining sector altogether. Anthropic said this week it confidentially submitted a draft registration statement for a U.S. IPO, setting up what could become one of the most important public-market tests of the AI boom. The company’s latest funding round valued it at $965 billion, putting it within reach of becoming one of the most highly valued companies ever to enter public markets.

That raises the question hanging over miners, AI infrastructure stocks and the wider equity rally: if a near-trillion-dollar AI IPO finally reaches public markets, will it validate the boom by giving investors the pure exposure they have been chasing — or will it mark the moment when the market has to decide how much of the AI future has already been priced in?

Regulation News

• North Dakota officials say data center concerns fueled by misinformation

Hardware and Infrastructure News

• Core Scientific CEO: AI Data Center Build Costs Climb Toward $12M Per MW

• SoftBank Commits Up to €75 Billion for 5 GW AI Data Center Buildout in France

• DMG Plans 50 MW Bitcoin-to-AI Conversion at Christina Lake Site

• Cango’s Bitcoin Mining Pullback Drove 30% Drop in Realized Hashrate

• Bitdeer Breaks Ground on 100 MW Alberta Site With On-Site Gas Power

• HIVE’s Bitcoin-Backed Bitmain Deal Fueled 18 EH/s Growth Spurt

• IREN Secures Connection for 800MW AI Data Center in South Australia

Corporate News

• NVIDIA Launches DSX Platform to Standardize AI Factory Infrastructure and Grid Interaction

• NVIDIA and Microsoft Launch RTX Spark Superchip for Local AI Agent Integration

• NVIDIA and TSMC Integrate AI and Accelerated Computing into Semiconductor Fabs

• Sphere 3D, Cathedra Complete Merger to Build 53 MW Bitcoin Mining Platform

• Google and Voltus Partner on 100 MW Virtual Power Plant to Support Data Center Growth

• Hut 8 Hires Former NextEra IR Chief as It Pursues Investment-Grade Rating

Financial News

• IREN Secures $3.6B Financing for Microsoft GPU AI Buildout at ~5.9% Cost

• Anthropic Files Confidentially for U.S. IPO Amid Race for Public Market Capital

• TransAlta to Acquire Colorado Natural Gas Assets for US$1 Billion

Email us

EIF Dallas (July 23, 2026) is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Get your tickets today and use code MNRW30 for 30% off exclusively for Miner Weekly subscribers. You can also explore sponsorship opportunities to elevate your brand and connect with leaders across energy, infrastructure, and capital.

For years, the story of Bitcoin mining and power grids was relatively simple: miners wanted cheap electricity, grids wanted flexibility, and the debate centered on whether large compute loads helped or hurt the system.

The AI data center boom is making that conversation much more complicated.

A new first-quarter – and 937-page(!!) – report from Monitoring Analytics, the Independent Market Monitor (IMM) for PJM Interconnection, offers one of the clearest warnings yet about what happens when large compute loads arrive faster than the grid can add a reliable power supply.

PJM is the largest US grid operator, serving all or parts of 13 states and Washington, DC. Its territory includes Northern Virginia, the world’s densest data center market, as well as major demand centers across the Mid-Atlantic and Midwest.

In fact, this report reads like the sequel to PJM’s earlier rule-change proposal, as covered in a previous Miner Weekly issue in March, which detailed PJM’s move to overhaul how large electricity users connect to and pay for the grid. Now the market monitor is putting numbers behind the concern: data center load is not just testing interconnection rules, it is already reshaping capacity-market costs.

Miner Weekly

The Grid Is Drawing a Line: PJM Moves to Rein In the AI Power Boom

Read more

5 months ago · 11 likes · BlocksBridge Consulting

At a high level, the report’s message is blunt but nuanced: PJM’s energy market still worked competitively in the first three months of 2026, but its capacity market is under strain.

That distinction matters.

The energy market is where electricity is bought and sold for near-term use — essentially the day-to-day market for power. The capacity market is different. It pays for power plants and other resources to be available in the future, especially during peak demand or emergencies. In plain English: energy pays for the electricity used today; capacity pays for the insurance policy that enough electricity will be there tomorrow.

That insurance policy is getting much more expensive.

PJM’s total wholesale power cost rose 75.5% in the first quarter from a year earlier, from $77.78/MWh to $136.53/MWh. Wholesale power cost is the all-in market cost of serving electricity demand through PJM. It includes energy, capacity, transmission, ancillary services and administrative fees, but the first three categories dominate the total.

Energy is the cost of the electricity itself. Capacity is the cost of making sure enough power resources are available in the future. Transmission is the cost of moving electricity across the high-voltage grid. Together, those three categories made up 98.3% of PJM’s wholesale power cost in Q1.

Each contributed to the increase. Energy costs rose 78.5%, from $54.67/MWh to $97.56/MWh. Capacity costs jumped 398.1%, from $3.57/MWh to $17.78/MWh. Transmission costs rose 5.3%, from $17.86/MWh to $18.80/MWh. Energy drove the largest dollar increase, while capacity had by far the steepest percentage increase.

That does not mean households will see their electric bills rise by 75.5%. Retail power bills include many other items, including local distribution charges, state-regulated utility costs, taxes, riders and long-term power contracts. Wholesale costs also flow through at different speeds depending on the state, the utility and the type of customer.

But the direction seems clear – at least according to PJM’s IMM – that the cost stack pressure behind customer bills is getting heavier.

Not just “supply and demand”

The most interesting part of the report is not simply that power costs are rising. It is how the Market Monitor explains why.

The report pushes back on two easy explanations at once.

The first is the argument that high capacity prices are just supply and demand: demand is rising, supply is tight, so prices should go up. The second is the opposite argument: that high prices prove PJM’s capacity market is broken.

The Market Monitor’s view is more specific. PJM’s capacity market was designed to handle normal load growth — the gradual increase in electricity demand from homes, businesses, factories, weather and broader economic activity. What it is facing now is different: large data center loads arriving in concentrated blocks, backed by customers with deep pockets and urgent timelines.

In that context, “supply and demand” is true but incomplete. The demand is real, but it is not organic. It is being driven by a new class of power buyer whose need for electricity may look almost unlimited compared with the grid’s ability to add dependable capacity.

The report says current capacity-market conditions are “almost entirely” the result of large data center load additions, both actual and forecast. It also says data center load growth is “unique and unprecedented and uncertain,” requiring a different approach than simply treating the results as ordinary supply and demand.

That is the nuance: the Market Monitor is not saying competitive markets should be abandoned. It is saying the market is being asked to absorb a kind of demand it was not designed for.

The report even argues that, without the “paradigm shift” represented by what it calls the almost inexhaustible demand for power from data centers, PJM’s capacity market would have produced reliability at prices more consistent with normal load growth and the cost of new capacity.

In other words, the issue is not whether data centers should get power. The issue is how they should get power — and who should pay for the reliability cushion needed to serve them.

The $23 billion question

The scale of that cost shift appears to be no longer theoretical.

The Market Monitor estimated that existing and forecast data center load increased capacity-market revenues by a combined $23.1 billion across the 2025/2026, 2026/2027 and 2027/2028 capacity auctions. It also said the inclusion of data center load in the last two auctions increased customer bills by $13.8 billion, even after a price cap under the Pennsylvania-PJM agreement limited the impact.

A capacity auction is the process PJM uses to make sure enough power resources will be available years ahead. Utilities and power suppliers buy capacity obligations, and generators get paid for being available. When large new loads are added to the forecast, PJM needs more capacity to serve them. If supply is tight, prices rise.

It’s important to note that the $13.8 billion figure is not charged to every household as a simple flat fee. Capacity costs flow through utilities and load-serving entities, and the actual impact varies by state, utility, customer class and capacity zone. But as a rough systemwide equivalent, spreading $13.8 billion across about two years of PJM electricity demand works out to roughly 0.85 cents per kWh.

The Market Monitor’s concern is that forecast data center demand is being included in the regular auction before enough new generation exists to serve it. In that setup, other customers can end up paying higher capacity costs created by the arrival of large data centers.

This does not mean data centers are the only reason power costs rose in the first quarter. The report also points to higher fuel costs, transmission constraints and winter weather. PJM’s real-time electricity price, known as LMP, rose 67.8% from a year earlier. LMP stands for locational marginal price. In plain English, it is the wholesale price of electricity at a specific place and time. It reflects the cost of generating power, congestion on the grid, and losses from moving electricity.

Natural gas remained the biggest driver of real-time prices. Gas costs accounted for 47.0% of PJM’s real-time load-weighted LMP in the first quarter, and gas plants were the marginal resource in most real-time intervals. A marginal resource is the last and often most expensive power plant needed to meet demand. That plant often sets the market price.

So the cleaner read is this: AI data centers did not single-handedly cause PJM power costs to rise. But they are becoming one of the biggest reasons the region’s future reliability looks more expensive and politically difficult.

PJM wants data centers to bring their own power

The report’s proposed answer is direct: large new data centers should bring their own new generation or be fully curtailable.

Curtailable means the customer can be required to reduce or shut off load when the grid is under stress. “Bring your own new generation” means a data center would need to contract for new power supply that matches its demand, instead of relying on the existing grid and pushing the cost of new capacity onto everyone else.

The Market Monitor recommends that PJM establish a load queue for large new data centers so they are not added until there is adequate generation capacity to serve them. It also recommends an expedited queue option for data centers that bring their own new generation with location and timing characteristics reasonably matched to their load profile.

That recommendation has obvious implications for the computing industry.

For years, Bitcoin miners have had to think like power developers. They searched for cheap or stranded power, built near substations, negotiated curtailment, managed price volatility and learned that energization matters more than equipment announcements.

AI data centers are now running into a similar reality, but at a much larger scale and with less tolerance for interruption.

The lesson from PJM is not that data centers are bad for the grid. It is that large new loads cannot be treated as ordinary demand growth when they arrive in gigawatt-scale clusters and expect firm service. Someone has to pay for the generation, transmission and reliability cushion needed to serve them.

The fight is over who that someone should be.

Why this matters beyond PJM

PJM is not the only grid facing this issue. Across the US, utilities and grid operators are trying to reconcile three forces that do not move at the same speed.

• AI data center demand is moving fast.

• Generation and transmission development move slowly.

• Customer affordability is becoming politically sensitive.

That mismatch is why the PJM report matters beyond its own footprint. It gives a data-rich version of the same question now emerging in other power markets: when a hyperscale data center wants to connect quickly, should the broader customer base help pay for the system upgrades and capacity needed to serve it?

The Market Monitor’s answer is no — or at least, not without clearer rules. Its view is that data centers should bear their own costs and risks, rather than shifting them to households, small businesses and other industrial customers.

That framing is likely to spread.

For Bitcoin miners, the report is both a warning and an opportunity.

The warning is that large-load scrutiny is intensifying. Mining, AI, cloud and high-performance computing will all face more questions about whether they are adding costs to the grid or bringing real flexibility and power supply.

The opportunity is that miners already understand the power-first playbook. As more mining companies pivot toward AI and HPC, their experience with power procurement, curtailment and interconnection could become a competitive advantage.

The AI boom may have started as a race for GPUs. PJM’s report suggests the next phase will be a race for credible power.

And in that race, the most important question may not be who can build the biggest data center. It may be who can prove they are not leaving everyone else with the bill.

Hardware and Infrastructure News

• New West Data Eyes Alberta Oil Deal to Power AI and Bitcoin Mining Expansion

• Texas Towns Pray To Jesus That Land Will Not Become AI Data Centers

• IREN Signs $1.6 Billion Agreement with Dell for Blackwell AI Infrastructure

• Enlight Secures 200 MW Solar PPA with Google for Oklahoma Data Center Operations

Corporate News

• Bitdeer Brings In Ex-Corsair CFO as Bitcoin Miner Pushes Further Into AI

• TeraWulf Stock Rises After 1 GW Kentucky AI Campus Acquisition

• Core Scientific Adds Former Equinix CEO to Board as AI Buildout Scales

Financial News

• Satoshi-era bitcoin whale moves over $200 million in BTC to FalconX, Cumberland: Onchain Lens

• Nebius Jumps as AI Hedge Fund Situational Awareness Reveals 5.6% Stake

• WhiteFiber Secures $100 Million Credit Facility for Data Center Expansion

Feature

• The Future of Bitcoin Mining Is Bigger Than Bitcoin

Email us

EIF Dallas (July 23-24) is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Apply to speak or explore sponsorships for private meeting rooms, branding & more.

Situational Awareness LP has quickly become one of the more closely watched names in the AI investment world.

Founded by former OpenAI researcher Leopold Aschenbrenner, the Artificial General Intelligence (AGI)-focused investment firm has drawn attention for both its explosive growth and the controversy surrounding its founder’s dramatic firing from OpenAI and his subsequent polarizing manifesto about the future of AI.

Its latest SEC filing this week only adds to the intrigue.

The fund’s reported 13F portfolio more than doubled in the first quarter, rising to about $13.68 billion from about $5.52 billion at the end of 2025. Much of that increase came from a larger options book tied to semiconductors and AI infrastructure.

But buried inside the direct equity portfolio was another notable trend: Situational Awareness has been loading up on Bitcoin miners – and former miners. Not just any miners, either.

Across the past few quarters, the fund has built or expanded positions in companies that sit near the center of the Bitcoin-to-AI infrastructure trade: Applied Digital, Bitdeer, Keel (formerly Bitfarms), CleanSpark, Core Scientific, HIVE Digital, IREN, and Riot Platforms. Many of these companies are either actively expanding into AI and high-performance computing or are trying to reposition their power and data center footprints for that market.

At the same time, Situational Awareness has not treated the sector as a single basket. In Q1, it exited Cipher Digital and Hut 8 and trimmed Core Scientific. That makes the filing more interesting than a simple “AI fund buys miners” story.

It suggests the fund may be sorting the sector by a sharper question: which companies offer the most attractive exposure to monetized power infrastructure?

The Miners Situational Awareness Bought and Exited

The biggest increases in Q1 came from names that still have large Bitcoin mining footprints but also meaningful AI or HPC optionality.

Situational Awareness increased its Keel position to 19.88 million shares from 6.90 million at the end of 2025. It boosted CleanSpark to 12.28 million shares from 1.64 million, Riot to 11.50 million from 6.17 million, IREN to 11.70 million from 8.70 million and Bitdeer to 3.44 million from 1.79 million. It also opened a new 3.39 million-share position in HIVE Digital.

The fund also added to Applied Digital, raising its stake to 13.48 million shares from 11.34 million, and increased CoreWeave to 7.18 million shares from 6.10 million.

That group spans several versions of the same broader theme.

Applied Digital is increasingly an AI data center developer. IREN is moving deeper into AI neocloud and AI data centers while still operating a large Bitcoin mining base. Bitdeer combines self-mining, hosting, ASIC manufacturing and early AI cloud exposure. HIVE, while still mining-heavy, is also becoming more active on the HPC cloud front through BUZZ HPC, which has announced AI cloud contracts, a fiber upgrade, and a planned 320 MW sovereign AI infrastructure project in the Greater Toronto Area. Riot has moved beyond pure mining after signing AMD as its first hyperscale data center tenant at Rockdale. Keel has laid out plans to convert existing bitcoin mining facilities to HPC data centers and wind down mining operations altogether. CleanSpark remains more mining-heavy, but each controls energized infrastructure that could become more valuable if power remains the bottleneck for compute growth.

In other words, Situational Awareness appears to be buying across the power-to-compute stack, not merely the Bitcoin mining trade. But the fund’s exits and trims make the pattern more nuanced.

Situational Awareness fully exited Cipher Digital after holding 10.47 million shares at year-end. That is notable because Cipher has one of the sector’s most prominent AI/HPC pivot stories, including major contracted data center capacity.

It also exited Hut 8, another company that has attracted attention for its power-first AI data center strategy. Core Scientific, one of the highest-profile Bitcoin-to-AI colocation conversions, was trimmed slightly to 26.01 million shares from 28.76 million.

So the filing does not read like a blanket bet on every miner with an AI narrative. It looks more like a valuation-sensitive rotation within the group.

One Way to Read the Trade: Enterprise Value per Monetized MW

The market has increasingly valued miners based on power access. But “power access” is a loose term. Some megawatts are already energized and producing revenue. Some are contracted but still under construction. Some are tied to future AI leases. Others remain part of a development pipeline that may take years to monetize.

A stricter way to compare these companies is to look at enterprise value against revenue-generating megawatts: capacity that is already energized and producing revenue through Bitcoin mining, hosting, colocation, AI cloud or HPC services.

That lens helps explain why some of Situational Awareness’ largest additions were in names that look relatively cheaper on currently monetized power. The lower end of the table is where the fund did some of its most aggressive buying.

HIVE was a new position and screened as one of the cheapest names on this basis. Bitdeer, Keel, CleanSpark and Riot also had relatively low enterprise value per revenue-generating megawatt. Those are not all identical businesses, but they share one thing: the market was assigning less value to each currently monetized MW than it was for the more AI-branded infrastructure names.

That may be part of the appeal.

For an AI-focused fund, the bet does not have to be that every miner becomes the next CoreWeave. It may be enough that energized power, operating sites and data center know-how become more valuable as AI demand keeps colliding with grid constraints.

The other side: unpriced HPC optionality

There is another layer to the trade.

Situational Awareness may not simply be buying the cheapest revenue-generating megawatts. It may be buying the cheaper megawatts where future HPC conversion has not yet been fully priced in.

That distinction matters. A miner with a low EV per revenue-generating MW can be cheap for two very different reasons. It may be overlooked, or it may lack a credible path to higher-value AI/HPC monetization. The interesting names are the ones that still screen cheaply on current monetized capacity but also have enough power, sites, management credibility or customer pipeline to plausibly become HPC candidates.

Once a miner signs a major hyperscaler contract, the market quickly stops valuing it like a Bitcoin miner and starts valuing it like an AI infrastructure developer. Enterprise value rises, EV per current revenue-generating MW expands and the “cheap power” trade becomes less obvious.

Cipher may be the cleanest example. The company has one of the strongest AI/HPC pivot stories in the group, with hundreds of megawatts already contracted to hyperscale customers. Cipher disclosed a 907 MW current portfolio, including 700 MW of contracted gross HPC capacity and operating Bitcoin mining capacity. It also pointed to about 3.3 GW of grid pipeline capacity expected to be energized across 2027 to 2030 and beyond.

That success may also help explain why Situational Awareness exited the position. By the end of Q1, Cipher’s future HPC capacity was no longer just an idea. It was tied to announced customer agreements, and a large portion of that upside appeared to have been capitalized into the stock. On current revenue-generating MW, Cipher screened as one of the most expensive names in the group.

In other words, the exit may not have been a rejection of Cipher’s AI strategy. It may have been a valuation call after the market had already rewarded that strategy.

The same logic helps explain why lower EV/MW names such as Bitdeer, CleanSpark, Riot, HIVE and Keel may still be attractive to an AI-focused investor. They have large operating power footprints, but many have not yet announced the same scale of investment-grade hyperscaler contracts. If even a portion of those megawatts can be converted from Bitcoin mining to higher-value HPC or AI data center use, the rerating potential could be meaningful.

That may be the optionality Situational Awareness is buying: not just monetized power today, but monetized power that has not yet been fully repriced as AI infrastructure.

Hut 8 shows the risk of reading 13F filings too literally

Hut 8 is the counterexample.

Situational Awareness exited Hut 8 during Q1, before the company announced a major AI data center lease in May. That later announcement changed the way investors could value Hut 8’s power pipeline.

On May 6, Hut 8 said it had signed a 15-year, 352 MW AI data center lease at its Beacon Point campus in Texas with a confidential high-investment-grade tenant. The company said the base-term contract value was $9.8 billion, inclusive of a 3% annual base rent escalator. The market reaction was immediate with Hut 8 shares jumping nearly 30% to all-time highs over $100.

With the benefit of hindsight, Situational Awareness may have exited too early.

Before the Beacon Point contract, Hut 8 may have looked less compelling on currently monetized capacity compared with the lower EV/MW mining peers. After the announcement, the market had a clearer basis to value its development pipeline as contracted AI data center revenue.

That is exactly how the miner-to-AI trade can reprice.

First, companies get credit for energized power. Then they get credit for secured power. Then they get a much larger premium when that power becomes contracted to investment-grade AI/HPC customers.

Cipher appears to have already moved further into the third stage by the end of Q1. Hut 8 arguably moved there after the quarter ended. Some of the lower EV/MW miners remain earlier in that progression.

The bigger signal

The bigger story is that Situational Awareness’ miner exposure has become another sign of how deeply the AI trade has moved into energy infrastructure.

The fund is not just buying chip stocks. It is buying companies tied to the physical bottlenecks of AI growth: power, data centers, substations, energized land and operating compute sites.

Bitcoin miners happen to own a lot of that infrastructure. Some have cheap power. Some have large campuses. Some have interconnection rights. Some have teams that know how to operate high-density electrical load. That makes them relevant to an AI investor even when their core revenue is still Bitcoin mining.

But the Q1 filing also shows that the trade is becoming more selective.

The first phase of the miner AI pivot was about who had power. The next phase may be about whose power is already monetized, whose power can be upgraded for HPC, and whose future AI value has already been priced in.

The market may be buying the miners where the power is already working, the valuation is still reasonable and the AI optionality has not yet been fully priced in. At the same time, Cipher and Hut 8 show what can happen once that optionality becomes contracted AI data center revenue: the rerating can arrive fast, and sometimes after the 13F snapshot has already moved on.

Hardware and Infrastructure News

• LU-VE Expands Texas Cooling Plant as AI Data Centers Drive Demand

• Canaan to Heat 2,800 Nordic Homes With Bitcoin Miners

• Hut 8 to fund $16 million water infrastructure project for Louisiana AI site

• Soluna Buys Out Dorothy 1B as Bitcoin-to-AI Campus Conversion Advances

• Applied Digital Inks $7.5B Lease for 300 MW at Polaris Forge 3, Expanding AI Factory Footprint

Corporate News

• HIVE Plans CAD $3.5B Ontario AI Factory With 320 MW Power Allocation

• NextEra Energy Confirms Deal to Acquire Dominion Energy in $400 Billion Utility Merger Fueled by AI Power Demand

• Bitfury Entities Secure $81 Million Through Cipher Digital Forward Sale Agreements

• Situational Awareness Deepens AI Power Bets, Adds Bitcoin Miners in Q1 Portfolio Shift

• NextEra Energy and Dominion Energy Outline Post-Merger Leadership and Structure

Financial News

• Bitdeer Reports Q1 2026 Revenue of $188.9 Million and Positive Adjusted EBITDA

• IREN Closes $3 Billion Convertible Deal after AI Push Draws Bigger Backing

• Fluence Energy Announces $420 Million Secondary Offering by Selling Securityholders

• CoreWeave’s New $3.1B Loan Tests Wider Market for AI Infrastructure Debt

• Canaan Posts $62.7M in Q1 Revenue as Mining Footprint Expands to 11 EH/s

Email us

EIF Dallas (July 23-24) is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Apply to speak or explore sponsorships for private meeting rooms, branding & more.

Public Bitcoin miners spent years racing to add more hashrate to the network. In the first quarter of 2026, many of them did the opposite.

Bitcoin’s average network hashrate, based on public blockchain data, declined from roughly 985 EH/s in Q4 2025 to 873 EH/s in Q1 2026. Separately, TheEnergyMag compiled quarterly production disclosures from major publicly traded miners to calculate their respective realized hashrate implied from Bitcoin production results.

At first glance, the aggregate change among large public miners appeared relatively modest. The combined realized hashrate of 10 major ones tracked by TheEnergyMag declined only slightly from approximately 297 EH/s in Q4 2025 to 291 EH/s in Q1 2026. HIVE and Cango were excluded from the comparison because their first-quarter production data was incomplete.

But beneath that seemingly stable aggregate figure was a much more notable redistribution of industrial-scale hashing power.

While companies such as Core Scientific, IREN, Cipher Digital, TeraWulf and Keel Infrastructure sharply reduced realized hashrate as they dismantled or repurposed mining fleets for AI and HPC infrastructure, others, including Bitdeer, MARA and American Bitcoin expanded aggressively to absorb part of the displaced network share.

Among the biggest decliners, IREN’s realized hashrate fell from 42.96 EH/s to 35.83 EH/s, while Cipher dropped from 16.55 EH/s to 11.14 EH/s after fully decommissioning mining operations at its Black Pearl facility in February to begin retrofitting the site for HPC infrastructure. Keel Infrastructure, formerly Bitfarms, declined from 16.52 EH/s to 11.51 EH/s as it continued winding down legacy mining operations and shifting toward North American AI infrastructure development.

CleanSpark witnessed a modest drop but similarly signaled it intends to continue monetizing Bitcoin infrastructure while selectively pursuing AI opportunities. Executives said older ASIC fleets may eventually be sold or relocated once AI deployments become fully operational, though the company acknowledged future site conversions could result in additional impairment charges.

By contrast, Riot Platforms increased realized hashrate from 34.21 EH/s to 42.29 EH/s during the quarter. Bitdeer climbed from 43.20 EH/s to 50.26 EH/s with energization of its SEALMINERs, while MARA rose from 51.92 EH/s to 55.52 EH/s despite simultaneous expansion efforts of their businesses around AI and HPC initiatives.

The divergence highlighted a growing split within the public mining sector and that shift became especially visible in corporate filings and earnings calls, where several miners disclosed large-scale fleet dismantling efforts, asset write-downs and mining infrastructure impairments tied directly to AI conversions.

Core Scientific said mining operations will continue winding down throughout 2026, with management expecting only one or two sites to remain operational for Bitcoin mining by year-end as the company prioritizes high-density colocation infrastructure for CoreWeave. The company recorded a $266.5 million impairment charge during Q1 2026, including $151.6 million related to mining equipment and $114.9 million tied to mining infrastructure.

Cipher Digital separately disclosed $30.8 million worth of mining rigs classified as held for sale after shutting down Black Pearl mining operations. TeraWulf owned approximately 54,100 Bitcoin miners as of March 31, but only around 35,500 were operational at its Lake Mariner campus. The remaining roughly 18,600 miners were categorized as undergoing maintenance, awaiting disposal, or held on standby to replace units under repair.

Instead of merely idling rigs during periods of weak economics, operators are permanently repurposing substations, cooling systems and data center layouts for AI deployments. Once infrastructure is converted for GPU workloads, it is unlikely to quickly return to Bitcoin mining.

American Bitcoin, one of the few companies still expanding its mining fleet, argued the transition could create a long-term opportunity for dedicated Bitcoin miners willing to continue scaling while competitors unplug fleets.

The company increased its owned fleet capacity from 25 EH/s to 28.1 EH/s in April following the reenergization of its Drumheller site, which had remained offline since 2024. Much of that growth, similar to its 2025 ramp-up, was financed through an unconventional structure that used pledged bitcoin rather than cash to acquire new-generation ASIC miners from Bitmain.

As of March 31, 2026, ABTC had pledged a total of 3,090 bitcoin to Bitmain for the purchase of 18 EH/s computing power, which alone represented nearly 64% of ABTC’s 28.1 EH/s proprietary mining fleet. ABTC mined 817 bitcoin during Q1 2026, up 505% from a year earlier. At the current production pace, and assuming Bitcoin network hashrate remains roughly stable, the company could theoretically mine back the equivalent of its originally pledged bitcoin collateral in about six quarters.

If network hashrate continues declining as industrial miners unplug more hashrate to pivot toward AI infrastructure, ABTC’s payback period in bitcoin terms could accelerate even further as remaining miners capture a larger share of block rewards.

All in all, the ongoing migration has altered the financial logic of industrial mining. During previous downcycles, miners typically unplugged rigs because falling Bitcoin prices or rising energy costs rendered operations uneconomic. In 2026, however, miners are increasingly shutting down fleets because AI infrastructure offers more stable long-duration cash flows, stronger financing conditions and higher expected returns on power capacity.

It will be worth watching how the dynamics play out in the quarters to come. But for now, the system remains balanced.

Hardware and Infrastructure News

• HIVE Advances Canadian Sovereign AI Push With New Brunswick Fiber Network Upgrade

• Zcash’s Price Revival Turns ZEC Miners Into Crypto’s Top Power Earners

• Canada Plans 150 MW AI Data Center Expansion in British Columbia Amid Sovereignty Push

• Bitdeer Reports $69M AI Cloud ARR and 783 Bitcoin Mined in April

• Nebius Commences Construction of Gigawatt-Scale AI Factory in Missouri

• Phoenix Group Launches European AI Data Center Push With 18MW France Buildout

Corporate News

• IREN Deepens NVIDIA Partnership With $3.4 Billion AI Cloud Agreement

• KEEL Spent $52M on HPC Expansion, Sold Bitcoin and Latin America Assets to Fund AI Pivot

• CleanSpark Reports $136.4 Million in Q2 Revenue; ERCOT Pipeline Expands to 585 MW

• MARA Signals More Bitcoin Sales as Long Ridge Deal Looms; Confirms 15% Layoffs in AI Pivot

• Riot’s AI-Fueled Bitcoin Mining Rally Triggers CEO’s Trading Plan to Cash Out $4.2M

• Braiins Acquires Commodity Fund CAMMS Capital, Bridging Mining and Traditional Markets

• Bitfury Co-Founder Secures $100 Million in Cipher Digital Share-Backed Deal

Financial News

• MARA Moves to Amend $600M Long Ridge Notes Ahead of Power Asset Acquisition

• IREN Shares Jump 16% as Bitcoin Mining Retreat Fuels AI Expansion

• TeraWulf Reports Q1 2026 Results and New Revolving Credit Facility

• IREN Shares Sink After Unveiling $2 Billion Convertible Offering for AI Buildout

• Nscale Secures $790M Financing for Microsoft-Backed Norway AI Campus

• MARA Holdings Reports $1.26 Billion Net Loss for Q1 2026

• NVIDIA Reports $215.9 Billion in FY2026 Revenue as Data Center Networking Surges 142%

• Nebius Q1 Revenue Jumps 7x Amid ‘Unprecedented Demand’ for AI Infrastructure

Email us

Today’s featured EIF speaker: Lisa Hough, who sits at the intersection of U.S. energy infrastructure, Bitcoin, and AI/HPC data center deployment. She brings institutional energy markets experience from Enron Capital & Trade, Phibro Energy, and PG&E National Energy Group, plus digital asset and regulatory expertise from Unchained and Custodia Bank, and is currently involved in structuring large-scale AI infrastructure initiatives including a roughly 2 GW campus in West Texas.

EIF Dallas (July 23-24) is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Apply to speak or explore sponsorships for private meeting rooms, branding & more.

For much of the past halving cycles, bitcoin miners paid a steep premium to access debt markets—from high single-digit to high double-digit interest rates tied to volatile mining cash flows.

That premium appears to be now compressing.

A wave of project-backed financings tied to AI and high-performance computing (HPC) shows borrowing costs moving decisively lower. Deals that cleared above 9% in 2025 are now landing closer to the 6%–7% range in 2026, according to recent bond offerings across Core Scientific, Cipher Digital, Applied Digital, Hut 8 and more.

At the same time, bitcoin-backed financing is repricing lower too, with Riot Platforms recently cutting its effective borrowing cost to ~6% and Hut 8 refinancing its Coinbase credit with a new deal from FalconX to slash the cost from 9% to 7%.

This isn’t just better market conditions. It’s a repricing of the underlying collateral.

Then: ASICs, Volatility—and Double-Digit Risk Premiums

To see how far the market has moved, it helps to look back at the last cycle.

In 2021–2022, most mining debt was backed by Bitcoin ASIC machines—a collateral base that was highly cyclical, sensitive to bitcoin price and network difficulty, and prone to rapid depreciation.

The result: borrowing costs that routinely sat in the 12%–16% range, with some deals even higher depending on lender and structure. Lenders like Genesis, BlockFi and Celsius dominated the space, extending credit against rigs and bitcoin holdings in what was effectively a crypto-native shadow banking system. It worked—until it didn’t, and dragged big lenders underwater altogether when bitcoin prices fell and liquidity tightened in 2022.

Now: Contracted Compute Is Replacing Machine Collateral

Fast forward to today, and the collateral has fundamentally changed.

Instead of ASICs, lenders are underwriting long-term AI compute contracts, colocation leases with escalators and power-secured data center infrastructure.

That shift moves the credit profile away from commodity exposure and toward infrastructure-style cash flows.

CoreWeave’s trajectory is emblematic—moving from ~9% financing in 2025 to the low-6% range, supported by contracted revenue and investment-grade positioning.

Miners making the AI pivot are now being evaluated less like speculative operators—and more like emerging data center credits.

Miner Weekly

From MinerFi to ComputeFi: CoreWeave’s $8.5B Loan Rewrites the Playbook

Read more

4 months ago · 7 likes · BlocksBridge Consulting

The Spread Tells an Interesting Story

Even as headline rates fall, dispersion remains notably—and increasingly informative.

From TheEnergyMag’s chart above:

• CoreWeave: low-6% range

• Hut 8: low-6%

• Cipher Digital: down from low-7% to low-6%

• Applied Digital: down from low-9% to high-6%

• Core Scientific: still pricing at high-7%

One of the key developments over the past year is the emergence of hyperscaler-linked credit support.

Hut 8, Cipher and TeraWulf have disclosed that Google provides financial backstop mechanisms tied to Fluidstack’s lease obligations—effectively anchoring those projects to hyperscaler demand.

That support matters and has helped pull borrowing costs into the mid- to high-single-digit range, tighter than what miners would achieve on a standalone basis.

This dynamic is clearest in the spread between Core Scientific and Hut 8. Both raised capital against AI infrastructure at roughly the same time with credible counterparties (Core Scientific with CoreWeave and Hut 8 with Fluidstack backed by Google). Yet Core Scientific continues to pay more.

AI Is Doing What Bitcoin Couldn’t

For years, miners argued they were infrastructure businesses. Markets didn’t buy it—because the collateral didn’t support the claim.

AI changed that.

With hyperscalers expected to spend roughly $700 billion on AI infrastructure this year, compute capacity has become a strategic asset. Power, land and interconnection—long the domain of bitcoin miners—are now in short supply.

Miners—who spent a decade building in exactly those constraints—are now being re-rated accordingly.

Hardware and Infrastructure News

• Riot Expands AMD Deal to 50 MW, Cuts Coinbase Loan Rate to 6.15% in AI Push

• IREN Rallies 11% on Sweetwater Energization Milestone Amid AI Power Race

• Peter Thiel Leads $140 Million Series B for Panthalassa’s Ocean-Powered AI Infrastructure

• Nscale Expands Microsoft AI Deal With 66,000 Rubin GPUs in Portugal

• Soluna Closes $53M Wind Farm Acquisition and Reports April 2026 Operational Progress

• American Bitcoin Secured Bitmain Miners at Implied 44% Discount Using BTC Collateral

Corporate News

• $70B Investor Coatue Expands Into AI Data Center Land Grab

• IREN to Acquire Mirantis in $625M Deal to Expand AI Cloud Software Stack

• Applied Digital Completes AI Cloud Spinout Into New Nasdaq Firm ChronoScale

• Nine Texas residents sue MARA, expanding Granbury bitcoin mine noise fight

• Core Scientific Reveals $232M Texas Acquisition, Sold 2,385 Bitcoin to Fund AI Expansion

• Hut 8 Surges to All-time Highs After Landing Second AI Data Center Lease Worth $9.8 Billion

Financial News

• Microsoft Committed $7 Billion Upfront in Nebius AI Deal

• Hut 8 Refinances Bitcoin-Backed Credit Facility with $200 Million FalconX Deal

• Applied Digital Secures $300 Million Bridge Loan for North Dakota AI Data Center

• Cipher Secures $200M Credit Line as Amazon AI Retrofit Cuts Bitcoin Hashrate

• Solaris Energy Prices $1.3B Notes at 6.375% as AI Power Buildout Accelerates

Email us

Today’s featured EIF speaker: Mike Alfred, Founder and Managing Partner of Alpine Fox LP, a value-equities and bitcoin investment partnership, and a board director at IREN and Bakkt. He previously co-founded Digital Assets Data and led it through its acquisition by NYDIG, bringing deep experience across bitcoin markets, public-company governance, and the energy plus compute investment landscape.

EIF is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Apply to speak or explore sponsorships for private meeting rooms, branding & more

The AI boom is rapidly redrawing the map of U.S. data center infrastructure — but the most important signal may lie not just in where facilities exist today, but where power is being requested next.

A dataset compiled from public records by Datacenter.fyi shows there are 3,995 existing data centers across the United States, with roughly 2,900 operational. Virginia and Texas dominate that footprint, with 575 and 410 sites respectively, forming the backbone of the country’s digital infrastructure.

But the next phase of growth is already taking shape.

Texas leads the nation with 336 planned data center projects, outpacing every other state and signaling its emergence as the primary expansion frontier. Virginia follows with 203 projects, while secondary markets such as Ohio (68 planned) and Indiana (37 planned) are gaining traction as developers look beyond traditional hubs. In total, the dataset counts 932 data centers that have been planned and proposed.

At the same time, historically dense markets like California show a more limited pipeline, with just 21 projects planned — a reflection of tightening power availability and permitting constraints in legacy tech corridors.

The geographic spread points to a clear shift: proximity to users is no longer the defining factor. Power is.

That shift becomes even more apparent when looking beyond physical sites to the grid itself.

Data from Interconnect.fyi further shows active interconnection requests surged to 1,687 projects in 2025, representing a record 442 gigawatts of active requested capacity. Just five years earlier, that figure stood at roughly 100 gigawatts.

The growth has been nonlinear. From fewer than 1 gigawatt in 2012, interconnection demand has scaled into the hundreds of gigawatts, with a sharp acceleration beginning around 2021 and intensifying through 2024 and 2025.

Together, the two datasets — one mapping physical infrastructure, the other mapping power demand — reveal the full scope of the AI buildout. What is visible on the ground is only a fraction of what is queued at the grid level.

Meanwhile, the capital behind this expansion is increasingly concentrated among a handful of technology giants.

Companies including Amazon, Google, Microsoft and Meta are not only among the largest spenders on AI infrastructure, but also rank among the most prolific data center operators in the U.S., according to Datacenter.fyi’s breakdown of facility ownership.

Amazon alone accounts for 479 data centers, making it the largest operator by count, followed by CyrusOne (221), QTS (186) and Digital Realty (170). Microsoft (153), Equinix (129) and Google (79) also rank prominently, underscoring how hyperscalers and large colocation providers dominate the physical footprint of compute.

That ownership concentration is being reinforced by an unprecedented surge in capital spending.

In the first quarter alone, Amazon, Google, Microsoft and Meta collectively spent more than $130 billion on capital expenditures, largely directed toward AI data centers and supporting infrastructure. The figure marked another record and represented a sharp increase from a year earlier.

The spending trajectory shows little sign of slowing. Combined outlays from the four companies are expected to approach $700 billion this year, as each races to secure capacity for increasingly compute-intensive AI workloads.

Taken together, the data center map, the interconnection queue, and the capex surge point to the same conclusion: the current buildout is not just broad-based, but highly capital-intensive and led by a small group of firms with the balance sheets to sustain multi-year infrastructure investments.

That dynamic also helps explain the pressure building in interconnection queues. As hyperscalers commit to gigawatt-scale deployments, their demand is translating directly into large-load requests that now dominate grid pipelines across multiple regions.

The pattern feels familiar

Long before AI drove hyperscale demand, Bitcoin miners were already optimizing for the same constraint. Following the China’s Bitcoin mining ban, mining infrastructure rapidly migrated to the U.S., with Texas emerging as a global hub that effectively replaced regions like Sichuan and Xinjiang.

That migration triggered an earlier wave of power and infrastructure development — from substations to large-load interconnections — built around energy-intensive compute. In hindsight, it offered a preview of the dynamics now playing out at a much larger scale.

Industry executives have increasingly drawn that connection. At a recent panel discussion at Bitcoin 2026, featuring MARA and CleanSpark, participants described Bitcoin miners as having “plowed the road” for AI, establishing the operational and infrastructure blueprint that data center developers are now expanding upon.

The implications are becoming clearer as interconnection queues swell.

Grid access — not land or hardware — is emerging as the primary bottleneck. Projects are measured not just by capital invested, but by their position in increasingly crowded queues, where timelines can stretch years.

For the mining sector, the convergence is both validating and competitive. The same regions that once attracted hashpower are now being targeted by AI developers with deeper balance sheets and longer-term contracts, intensifying the race for power.

In that sense, mapping the U.S. data center landscape today is less about counting facilities and more about understanding the underlying energy dynamics.

The physical footprint — nearly 4,000 sites — tells the story of where computing has been. The 442 gigawatts sitting in interconnection queues tells the story of where it’s going.

Regulation News

• Maine Governor Vetoes Data Center Moratorium, Citing $550M AI Project

• China Blocks Meta’s $2 Billion Acquisition of AI Startup Manus

• Keel Wins Zoning Approval for AI Data Center Expansion at Former Bitcoin Mine

Hardware and Infrastructure News

• Central North Carolina communities pause to consider data center development

• Duke Energy Extends Robinson Nuclear Plant to 2050 Amid AI-Driven Power Surge

• Luxor, MicroBT Tie Investment to $100M Miner Deal Amid Weak Bitcoin Hashprice

• Core Scientific to Expand Pecos Campus Capacity to 1.5 GW for AI Data Center Development

• Solaris Energy Signs 600 MW Power Contract, Raises 2026 Guidance

• Galaxy Targets Full 133 MW CoreWeave Delivery at Helios by End of Q2

• Canaan Lands Tether Order for Customized Bitcoin Mining Systems

• MARA to Buy Ohio Gas Power Plant in $1.5 Billion Bet on AI Data Centers

Corporate News

• Brookfield-Backed Csquare Files Confidentially for AI-Fueled US IPO

• Tether Pushes Deeper Into Bitcoin Mining Stack With Open-Source Software Play

• Riot Amends $200M Coinbase Credit Line to Cushion Bitcoin Swings

• Galaxy Hints at Trillion-Dollar Hyperscaler Behind Helios AI Buildout

• Nscale Hires Former Oracle Data Center Lead to Drive AI Buildout

Financial News

• NextEra Energy Reports $6.7 Billion in Q1 2026 Operating Revenues

• Hut 8 Taps $3.25 Billion Debt for Fluidstack-Backed AI Data Center

• Vistra Corp. Completes $4 Billion Private Offering of Senior Notes

• NRG Energy Issues $2.6 Billion in Senior Notes and Secures $900 Million Term Loan

Feature

• For AI and Bitcoin Investors: The Grid’s ‘Secret Sauce’ Just Got Even Sweeter

• ‘We Plow the Road’: Miners Pitch Role in AI Buildout at Bitcoin 2026

• Who Owns the Stack: From Bitcoin to AI, the Race for Power Is Going Off-Grid

Email us

EIF is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.

Apply to speak or explore sponsorships for private meeting rooms, branding & more.

Core Scientific grabbed headlines this week with its proposed $3.3 billion senior secured notes priced at 7.75%, the latest in a wave of high-yield debt offerings tied to the AI infrastructure boom.

While the financing itself has been widely covered, a lesser-noticed supplemental filing tied to the deal offers something more revealing: a detailed look at the actual infrastructure assets being pledged to back the debt — and, more importantly, the cash flow profile expected to service it.

At the center of the deal is Core Scientific’s long-term hosting relationship with CoreWeave. The document shows that the two parties have already funded roughly 80% of the estimated $5.5 billion in total capex for the projects, underscoring how far along the buildout is.

It also lays out how Core Scientific is packaging the CoreWeave-related data center assets into a financing vehicle with the underlying sites — and their future revenues — forming the backbone of the credit story. In doing so, it provides one of the clearest disclosures yet into how a Bitcoin miner’s AI pivot is being translated into a project-financed, debt-backed business.

A Site-by-Site Look at the Buildout

Based on the document, Core Scientific is developing six data center facilities across five campuses, having secured roughly 900 MW of grid capacity to support about 590 MW of contracted critical IT load exclusively for CoreWeave.

As of mid-March, around 350 MW has been energized, with more than 185 MW already billing — a notable milestone, as it marks the transition from construction to cash-generating operations.

The buildout, however, is not yet complete, and the timeline stretches into the second half of 2027.

The filing breaks down progress by campus, offering rare visibility into how much capacity is coming online and when revenue is expected to fully ramp up.

• Austin, TX: Fully completed in 2024, with 16.5 MW of IT capacity. This site is already generating base license fees and serves as the earliest proof point of the model.

• Denton, TX: The largest project in the portfolio at 262 MW. As of March 2026, about 132 MW — roughly half — is already billable. Full completion is expected in the second half of 2026.

• Marble, NC: A 65 MW campus that is about 57% complete, with 37 MW currently billing. It is expected to reach full completion in the first half of 2026.

• Muskogee, OK: A 70 MW campus where construction began in early 2025, with both delivery and completion expected in the first half of 2026.

• Dalton 1, GA: A smaller 30 MW site scheduled to begin delivery and reach completion in the first half of 2026.

• Dalton 4, GA: A larger 145 MW expansion that will come later in the cycle, with delivery starting in the second half of 2026 and full completion targeted for the first half of 2027.

Taken together, the portfolio shows a staggered ramp, where a meaningful portion of capacity is generating revenue even as the bulk of megawatts are still under construction.

The $10B Headline — and What Actually Drives Revenue

Core Scientific frames the exclusive long-term contract as generating roughly $10 billion in total revenue over the life of its agreements with CoreWeave, net of rent repayments tied to the upfront construction funding contributed by CoreWeave.

Notably, that headline figure includes power passthrough, which materially inflates the top line without contributing to economic margin. Under this structure, CoreWeave reimburses electricity costs paid by Core Scientific, with those payments recorded as both revenue and operating expense on Core Scientific’s balance sheet.

A more grounded way to think about the underlying revenue engine is the base license fee.

Across the portfolio, CoreWeave pays a fixed rate of $100 per kW per month for most sites and $115 per kW per month for the Austin campus, with built-in annual rent escalators of 3.5% (and 3.0% for Austin). Applying those rates to the full 588.5 MW of contracted capacity translates into approximately $59 million per month in base license fees once all sites are fully built and operational, or roughly $700 million annually. This figure is a closer proxy for the net revenue generated by the infrastructure itself, before operating costs and financing effects.

Revenue vs. Levered Free Cash Flow

The illustrative financials included in the filing further underscore that distinction.

For 2026, Core Scientific projects:

• $737 million in total revenue, including power passthrough

• $418 million in operating expenses, also including power passthrough

• That works out $319 million in net operating income

• But just $6 million in levered free cash flow

Stripping the power passthrough from both revenue and operating expenses, the net operating income reflects the actual value of the hosting agreements rather than grossed-up utility flows.

The more meaningful drop occurs below that line.

The sharp compression from net operating income to levered free cash flow is largely driven by what the filing categorizes as “Other Cash Flow Items.” While this includes maintenance capital expenditures, a substantial portion relates to CoreWeave repayment mechanics — specifically, the revenue credits that reduce the cash Core Scientific receives as reimbursement for upfront construction costs. Under the agreements, CoreWeave is entitled to credits equal to 50% of the base license fee, capped at $1.5 million per MW of deployed capacity, effectively allowing it to recoup a certain portion of the buildout costs over time through reduced cash payments.

In effect, what appears as strong contracted revenue is partially being recycled back to the customer in the early years, delaying cash realization. When combined with interest expense on the $3.3 billion debt stack, the result is a business that generates significant accounting income but very limited near-term free cash flow.

Why This Matters for the Bitcoin-to-AI Pivot

This supplemental filing may not have made headlines, but it offers a critical window into how the industry’s AI pivot is being financed.

The takeaway is nuanced. Core Scientific’s deal with CoreWeave delivers long-term, contracted revenue with built-in escalation — a stark contrast to the volatility of Bitcoin mining. But it also introduces a capital-heavy structure where cash flow is constrained in the early years and heavily shaped by financing terms.

In many ways, this mirrors what has already been seen on the other side of the relationship. Just weeks ago, CoreWeave was tapping debt markets with GPU-backed financing structures, effectively using contracted compute demand as collateral to raise capital. Here, Core Scientific is doing something similar — but with powered land and data center infrastructure instead of chips.

Both cases point to the same underlying shift: AI infrastructure is increasingly being built not just with equity and operating cash flow, but through structured, asset-backed financing models where future contracted revenue is pulled forward to fund present-day capex.

That makes this more than just a Bitcoin miner diversifying into AI. It is another example of how the entire stack — from GPUs to data centers — is converging toward a project finance model, where assets are ring-fenced, cash flows are pre-allocated, and returns are shaped as much by capital structure as by demand.

Hardware and Infrastructure News

• Soluna Eyes 98 MW Bitcoin-to-AI Conversion With $16.5M First-Step Dorothy Buyout

• Data centre delays threaten to choke AI expansion

• Alcoa Nears Sale of Idled Smelter to NYDIG as Bitcoin and AI Chase Industrial Power

• Microsoft’s $3.3B Fairwater AI Data Center Goes Live Ahead of Schedule in Wisconsin

• Soluna Expands Blockware Deal With 3.3 MW at Texas Wind-Powered Bitcoin, AI Data Center

• Keel Infrastructure Finalizes $13 Million Sale of Paraguay Site, Exits Latin America

• American Bitcoin completes installation of over 11,000 miners at Drumheller facility

• Applied Digital Secures $7.5 Billion Lease for 300 MW AI Campus

Corporate News

• Fermi Shares Sink 20% as CEO Exit Clouds AI Data Center Ambitions

• TransAlta Appoints Mike Politeski as CFO and Grant Arnold as CCO

• Amazon to invest up to another $25 billion in Anthropic as part of AI infrastructure deal

• Cango Appoints New CFO and Director Amid Strategic Shift

• Nscale, BT Plan UK Sovereign AI Data Centers With NVIDIA Stack

Financial News

• Tether Takes 8.2% Stake in Bitmain-Tied Bitcoin Mining Lender Antalpha

• Core Scientific Seeks $3.3B Bond as Bitcoin Miner Rides AI Data Center Debt Boom

• Soluna Saw 10X Volume Spike After Nasdaq Warning as Bitcoin-to-AI Pivot Gains Focus

• HIVE Announces Closing of Private Offering of US$115 Million of 0% Exchangeable Senior Notes

• Ex-Credit Suisse Team Plans $1 Billion Fund for Data Center Risk

Features

• New Gas-Powered Data Centers Could Emit More Greenhouse Gases Than Entire Nations - Wired

Email us

Today’s featured EIF speaker: Matthew Roszak, chairman and co-founder of Bloq and Hemi, and founding partner of Tally Capital, a globally recognized investor who has backed category-defining companies across the bitcoin and digital asset ecosystem, including Binance, Coinbase, Blockstream, and BitGo.
EIF is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.
Register to attend today or explore sponsorships for private meeting rooms, branding & more.

Public bitcoin miners have liquidated their BTC reserves at a pace not seen since the depths of the last crypto bear market, as a prolonged slump in mining economics pushes operators into survival mode.

Several major public miners, including MARA, CleanSpark, Riot, Cango, Core Scientific, and Bitdeer, have already sold more than 32,000 BTC in the first quarter of 2026, according to data analyzed by TheEnergyMag. The dataset remains incomplete, as first-quarter earnings reports are still pending.

Even so, the figure already exceeds total net sales across all four quarters of 2025 and sets a new industry record, surpassing the roughly 20,000 BTC that public miners liquidated in the second quarter of 2022 during the market turmoil triggered by the Terra-Luna collapse.

The reversal is striking. Just over a year ago, miners were accumulating aggressively, ending 2024 with a net addition of 17,593 BTC and pushing combined reserves above 100,000 BTC.

The shift comes as hashprice — a key industry metric measuring expected mining revenue per unit of computing power — hovers in the low $30/PH/s range, near all-time lows. At those levels, margins are either compressed or outright negative, especially for operators running older, less efficient fleets or paying higher power costs.

The root of today’s pressure can be traced to the industry’s aggressive hashrate expansion following China’s mining ban in 2021 — a period that, in hindsight, fueled exponential growth at each company’s own expense.

For context, bitcoin’s current price — despite retreating from its all-time high above $120,000 — remains higher than the previous cycle peak. Yet network difficulty is now roughly 10 times higher than in 2021, and block rewards were cut in half in 2024. In effect, mining profitability has compressed by an order of magnitude, helping explain the wave of recent selling.

But the record liquidation does not tell a uniform story. Instead, it reveals an industry beginning to diverge — with some operators forced to sell into weakness, while others lean on structural advantages or capital discipline to ride out the downturn.

For many, the immediate priority is liquidity. Selling bitcoin remains the fastest way to shore up balance sheets, fund operations, and meet debt obligations in a financing environment that is both selective and expensive.

Others are taking a more measured approach. American Bitcoin (ABTC), the proprietary mining carve-out of Hut 8, has doubled down on accumulating bitcoin through both mining and market purchases. As of early April, it had built reserves of more than 7,000 BTC, up from zero a year earlier, while ramping its proprietary hashrate to 28 EH/s.

But the company is not repeating the hashrate-at-all-costs playbook of the previous cycle. Matt Prusak, president and interim CFO of ABTC, told TheEnergyMag that its focus is on quality growth under current market conditions.

“We won’t do deals that we don’t think will win. … When you see people chasing hashrate and chasing big exahash numbers, that’s never been our style,” Prusak said. “Having the biggest fleet doesn’t make a difference to me.”

Unlike many publicly traded peers, ABTC expanded its hardware fleet when demand for ASICs had already cooled. In the summer of 2025, it acquired roughly 15 EH/s of Antminer S21 series from Bitmain by pledging about 3,000 BTC — redeemable within 24 months — rather than paying cash. Such a structure would have been unlikely in prior market conditions. The value of the pledged bitcoin has since declined by about 40%, and Bitmain cannot liquidate the collateral unless ABTC chooses not to redeem it.

Based on Q4 2025 data analyzed by TheEnergyMag, ABTC’s all-in cash cost of production was around $55,000 per bitcoin, or roughly $25/PH/s — among the lowest in the public mining cohort. That allows it to accumulate newly mined bitcoin at a discount to prevailing market prices.

Even if bitcoin falls below that level, Prusak said the company retains the flexibility to allocate capital dynamically. ABTC raised $240 million through at-the-market offerings in 2025 and another $110 million in the first quarter of this year.

“We don’t have to flip into AI. We are a bitcoin allocator. If bitcoin’s expensive relative to the cost to mine, we mine. If bitcoin’s cheap relative to mining, then we buy.” he said, adding: “ At this time, we have no intent ot sell. … We are accumulating.”

But for private operators without comparable access to capital, the divergence in strategy is increasingly shaped by one of the industry’s oldest variables: power costs.

Sean McDonough, president and CEO of New West Data, a Canadian oil producer that mines bitcoin using off-grid power generated by flared natural gas from its own oil sites, said the company’s effective power cost is below $0.02 per kilowatt-hour. That is, in some cases, roughly one-third of what large-scale public miners pay.

At that level, even less efficient machines remain profitable. With hashprice around $30/PH/s, a miner paying $0.02/kWh can sustain fleet efficiencies of roughly 60 J/TH. McDonough said this enables the company to acquire older-generation equipment at lower upfront costs while maintaining margins, especially as ASIC prices have fallen alongside hashprice.

That cost advantage has allowed New West Data to expand despite the downturn. The company tripled both its oil production and bitcoin compute capacity in 2025 and expects to triple again this year. It currently operates about 15 MW of computing capacity, all powered by flared gas from its own sites.

Still, flared gas represents a niche model, requiring expertise in upstream oil production rather than traditional power procurement through utilities or long-term power purchase agreements.

Absent ultra-low-cost power, miner operators are also turning to operational optimization to preserve margins.

Luxor, a bitcoin mining pool operator, ASIC broker, and software provider, launched a fleet management tool called Commander earlier this month. The platform uses automated algorithms to evaluate hashrate and power market conditions every five minutes, dynamically adjusting power settings across a fleet based on real-time economics.

The goal is to optimize output from existing infrastructure. Luxor says internal benchmarks show an 8% to 14% improvement in profitability compared with traditional on/off curtailment strategies.

The shift toward software reflects a broader recalibration across the industry. With hashprice under pressure, upgrading to the latest generation of machines often requires capital outlays that are difficult to justify on a standalone return basis.

Instead, operators are focusing on extracting better margins from existing fleets — gaining incremental efficiency wherever possible.

Ethan Vera, Luxor’s chief operating officer, said the Commander platform has already scaled to about 5 EH/s of customer hashrate since launch. It complements LuxorOS, the company’s firmware solution introduced in 2022, which now supports roughly 45 EH/s, or about 5% of the global network.

In one recent case study, Luxor claimed that Soluna, a publicly traded bitcoin miner with colocation and proprietary mining in Texas, was able to speed up the recovery time via LuxorOS for its 1.1 EH/s fleet by 50% after curtailment events, improving uptime without additional operational expenditure.

All told, the industry is no longer moving in lockstep. What was once a relatively uniform business model defined by scale and hashrate growth is fragmenting into a range of survival and quality growth strategies shaped by power economics, balance sheet flexibility, and operational sophistication.

Regulation News

• Missouri town fires half its city council over data center deal

Hardware and Infrastructure News

• Cango Launches Commercial AI Inference Operations at 50-MW Georgia Facility

• Microsoft Secures Former OpenAI “Stargate” Site in Norway for AI Infrastructure

• Microsoft Secures 3,200 Acres for Massive Cheyenne Data Center Expansion

• Bitdeer Ramps to 70 EH/s as ‘Seasonal Factors’ Hit Bitcoin Mining Uptime

Corporate News

• CoreWeave Secures $21 Billion Meta Contract and Targets $4.7 Billion in New Debt Financing

• CoreWeave Lands Anthropic Deal, Upsizes Convertible Notes to $3.5B Amid AI Boom

• Riot Data Center Chief Departs Less Than a Year After Bitcoin Miner’s AI Push

• CoreWeave Lands $6B AI Deal With Jane Street, Secures $1B Equity Investment

Financial News

• NRG Energy Announces Multi-Tranche Debt Offerings and $900 Million Term Loan

• Jane Street in Talks to Back Fluidstack at $18 Billion Valuation

• TeraWulf raises $900 million through stock sale for Kentucky infrastructure and debt settlement

• HIVE secures $75 million through private bond issuance and seeks to list on TSX

Email us

Today’s featured EIF speaker: Congressman Earl “Buddy” Carter (R-GA), a member of the House Energy and Commerce Committee and the House Budget Committee, where energy reliability, infrastructure investment, and technology policy collide.
EIF is a high-signal, curated conference convening leaders across energy, compute infrastructure, and capital to map the next decade of power and data center buildout.
Register to attend today or explore sponsorships.

If 2021 was the year of ASIC-backed loans, 2026 may go down as the moment compute financing grew up.

CoreWeave’s $8.5 billion delayed draw term loan—secured primarily by GPU infrastructure—has already made headlines as the first and largest deal of its kind. The headline numbers alone are notable: the facility allows CoreWeave to draw capital through June 2027, with final maturity in March 2032. It is split between floating-rate debt priced at roughly SOFR + 2.25% and fixed-rate tranches around the high-5% range, putting the overall cost of capital in the mid-single digits—remarkably low for an asset class that didn’t meaningfully exist a few years ago.

But buried inside the credit agreement is something far more intricate: a blueprint for how Wall Street now wants to finance compute. This isn’t just “GPU-backed debt.” It’s compute treated as infrastructure.

And in many ways, it’s what MinerFi tried—and failed—to become.

The lesson from 2021: hardware alone isn’t enough

Back in 2021, bitcoin miners tapped a wave of ASIC-backed loans to expand aggressively. The premise was simple: machines generate bitcoin, bitcoin repays debt.

What that model underestimated was how volatile and fragile both sides of the equation were.

When bitcoin’s price fell in 2022, the hashrate went up relentlessly as a result of the MinerFi boom, deteriorating the hashprice even further. At the same time, the resale value of ASIC machines declined. So lenders were left with collateral that was worth less and generating less cash.

That double exposure—asset value risk and cash flow risk—was fatal.

The real collateral isn’t GPUs

CoreWeave’s credit agreement emphasizes the Data Center (DC) Funding Conditions, which are really the spine of the whole deal.

Simply put, CoreWeave cannot simply point to a purchase order for GPUs and borrow against it. The financing is tied to whether those machines are actually becoming usable compute inside a functioning data center, for a customer contract with the customer’s acceptance and satisfaction.

That matters even more because the advance rate is high.

The agreement defines the borrowing base for a funding date as the “Funding Date GPU Amount,” which equals 90% of Funding Date Capital Expenditures, plus certain fees and expenses tied to the transaction. Those capital expenditures include the cost of the infrastructure itself and installation costs.

In practical terms, lenders are saying they are willing to fund almost the entire build cost of the GPU deployment. But they only do that because the deal does not treat a GPU as collateral the moment it is bought.

Instead, the agreement requires the financed infrastructure to satisfy the DC Funding Conditions. Those conditions tie funding to real-world deployment: the equipment has to be physically associated with the relevant data center site, prior loans must already have been used properly, and for the applicable infrastructure, the DC Funding Conditions must be satisfied at the time of the borrowing. The contract also includes language elsewhere that accepted GPU clusters count toward Funding Date Capital Expenditures only when the customer has accepted them in writing and the borrower certifies testing has been completed for the relevant data halls.

So the key idea is this:

CoreWeave is getting very high leverage — roughly 90 cents of debt for every dollar of qualifying deployment cost — but only after proving that the machines are not just purchased, but installed, tested, accepted, and on the path to generating contracted revenue.

That is a notable difference from the MinerFi era.

In 2021, ASIC-backed lenders often financed miners largely on the assumption that once the machines were delivered, the economics would take care of themselves. Here, the lenders are willing to go bigger precisely because they are financing not a box of chips, but a controlled, trackable, revenue-linked compute system. That is what makes this feel less like equipment lending and more like infrastructure finance.

Following the machines down to the rack level

Once you start from that premise—only funding “ready-to-earn” assets—the rest of the structure builds around visibility and control.

Every GPU financed under the facility is tracked:

• It must be located at an approved site

• It must be tied to a specific deployment

• Lenders must have contractual access to the facility

This is very different from early ASIC lending, where collateral could be loosely defined and, in some cases, difficult to locate or control in distress scenarios.

Here, lenders know not just what they are financing, but where it is and how it is being used.

Cash doesn’t flow freely—it follows a strict order

Once those GPU clusters start generating revenue, the cash doesn’t simply go to CoreWeave.

It flows through a controlled system—known as a cash waterfall.

Here’s the idea:

Every dollar that comes in is a pre-assigned job.

1. First, pay the costs to keep the system running (power, operations)

2. Then, pay lenders (interest and principal)

3. Only after that can any remaining cash go back to the company

Why it matters: it ensures that debt repayment is prioritized automatically, rather than relying on management discretion.

In MinerFi, that level of control was often opaque.

And the agreement goes one step further with what’s called a cash trap.

If the project starts underperforming—for example, if it’s not generating enough cushion above its debt obligations—then excess cash stops flowing back to the company.

Instead, it stays within the system to prevent situations where value leaks out of the structure just as risks are rising.

Even the electric bill is part of the credit story

One of the more telling details is how seriously the agreement treats power costs.

Running GPUs at scale is energy-intensive, just like running bitcoin miners, and electricity is one of the largest—and most volatile—expenses.

The structure requires mechanisms to manage that risk, including hedging or other arrangements to stabilize power pricing.

In simple terms: the loan isn’t just underwriting compute—it’s underwriting the cost of the electricity that makes that compute possible. And that variable is actively controlled rather than left exposed.

Borrowing only happens when the math works

All of this feeds into a final layer of discipline: the project must prove it can pay for itself.

That’s where the debt service coverage ratio (DSCR) comes in. DSCR compares how much cash a project brings in versus how much it needs to pay lenders (principal+interest).

• If a project generates $120 and owes $100 in debt payments, its DSCR is 1.2x

• That extra $20 is the cushion—it allows for things to go slightly wrong without immediately causing stress

In CoreWeave’s case:

• To take on new debt, the projected DSCR needs to be around 1.20x

• To stay in compliance, it needs to maintain roughly 1.15x

That buffer is what gives lenders confidence that even if things don’t go perfectly, the system can still hold.

The bigger shift: from machines to systems

Put all of this together, and the contrast with 2021 becomes clear.

MinerFi was built around machines: Finance the hardware and expect the revenue to follow.

This deal is built around systems:

• Verify the infrastructure is live

• Confirm the cash flow is real with customer contracts

• Control how that cash is used

It’s a fundamentally different way of risk management.

At $8.5 billion, backed by major banks and structured to investment-grade standards, this isn’t just a large deal. It’s a signal.

Compute—specifically AI compute—is being redefined as infrastructure that can be financed with the same discipline as power plants or toll roads.

Bitcoin mining showed that energy can be turned into digital output at scale.

CoreWeave’s financing shows how that output can be turned into something lenders are willing to underwrite at massive scale.

Of course, structure can’t eliminate risk. Both MinerFi and this new wave of compute financing ultimately depend on demand holding up.

The difference is where the risk sits.

• In mining, it was concentrated in network and market-driven revenue (hashprice)

• Here, it’s embedded in contract durability and utilization

Whether that shift makes the system more resilient—or simply changes how stress emerges—remains to be seen.

But one thing is already clear: Compute financing has entered a new phase—and this time, it looks a lot more like Wall Street than crypto.

Hardware and Infrastructure News

• Soluna Buys Texas 150MW Wind Farm for $53M to Power AI, Bitcoin Data Centers

• Bitdeer Unveils 9.45 J/TH Miner as Bitcoin Hashprice Remain Under Pressure

Corporate News

• DeepSeek’s V4 model will run on Huawei chips, The Information reports

• CleanSpark Shutters 25MW Tennessee Bitcoin Mine Tied to GRIID Noise Dispute

Financial News

• Riot Sells 3,778 Bitcoin in Q1, Posts More Collateral as BTC Slides

• Cango Sells 2,000 Bitcoin in March After 4,616 BTC February Exit to Cut Debt

• CoreWeave Signs $21B Meta AI Cloud Deal as It Seeks Another $4B Debt Financing

Email us

Introducing EIF: A new conference in Dallas (July 23-24, 2026) for executives, investors, and operators across mining, AI/HPC, power, and infrastructure. Full details in today's press release.

The AI infrastructure boom has no shortage of capital.

It has a shortage of conversion.

Across the industry, billions of dollars are being raised, allocated, and announced at a pace that would have been unthinkable just a few years ago. But turning that capital into functioning data centers, energized megawatts, and ultimately revenue-generating compute is proving far more complex — and far slower — than the headlines suggest.

That gap is beginning to define the next phase of the AI supercycle.

Capital is flooding in

By TheEnergyMag’s analysis, nine publicly listed bitcoin mining data center operators with HPC/AI initiatives collectively raised $17 billion in net financing inflows in 2025, up from $10.9 billion in 2024. Over the same period, their net spending on property, plant, and equipment reached $6.3 billion, compared with $4.5 billion a year earlier.

The direction is clear: more capital is entering the system, and more is being deployed.

But the gap between capital raised and capital converted into productive infrastructure remains wide.

Zooming out, the same pattern is now playing out across the broader AI ecosystem.

Major technology companies, including Microsoft, Amazon, Alphabet and Meta Platforms, are expected to spend as much as $635 billion on AI infrastructure in 2026, up sharply from $383 billion in 2025.

The capital is there.

The question is how quickly — and how efficiently — it can be turned into revenue.

The conversion problem

At NVIDIA’s GTC 2026, the tone from infrastructure investors and operators was less about fundraising and more about execution. The Luxor team summarized their key takeaways from attending the event in a recent newsletter post.

One point stood out: money is no longer the scarcest input.

Instead, the constraints are increasingly physical and operational.

A single gigawatt-scale AI campus can require more than 9,000 workers across dozens of states. Equipment supply chains — from transformers to cooling systems — remain tight. Interconnection queues are clogged with projects that may never materialize. And rapidly evolving compute architectures are forcing redesigns even after construction has begun.

In other words, the bottlenecks are no longer just about securing capital or even power in isolation.

They sit at the intersection of:

• labor availability

• grid access and transmission

• equipment manufacturing

• site design and cooling architecture

• tenant readiness and utilization

Each layer introduces friction between capex committed and cash flow realized.

The grid is no longer a passive input

Nowhere is this friction more visible than in power markets.

In the Electric Reliability Council of Texas (ERCOT), as of February 2026, the Large Load Interconnection queue contained approximately 400 individual requests totaling 239,000 MW. This volume is roughly 2.8 times the current ERCOT record for system peak demand.

The headline number suggests overwhelming demand.

But it also highlights a deeper issue: not all megawatts are equally real.

In PJM Interconnection, regulators and grid operators are already moving to separate speculative load from bankable demand. PJM’s latest long-term forecast places greater emphasis on “firm” commitments — projects backed by actual construction progress or binding service obligations — rather than early-stage requests.

Meanwhile, the Federal Energy Regulatory Commission has stepped in to push for clearer rules governing co-located large loads such as data centers, reflecting mounting concern over how these facilities interact with existing grid infrastructure.

The shift is subtle but important.

The market is moving from who can secure power on paper to who can actually deliver power in practice.

From gigawatts to monetization

Even once a site is financed and energized, the conversion process is not complete.

Revenue depends on:

• securing long-term customers

• achieving high utilization rates

• aligning infrastructure with evolving compute requirements

• maintaining uptime and service guarantees

This is particularly relevant for bitcoin miners pivoting into HPC and AI colocation.

Their advantage lies in power sourcing, site development, and operational expertise. But participating in the AI infrastructure stack requires adapting to a different economic model — one that is contract-driven, utilization-sensitive, and increasingly dependent on creditworthy counterparties.

The result is a growing divergence between announced capacity and monetized capacity

The deeper story is that the AI boom is no longer confined to software or semiconductors.

It is now fully entangled with the physical economy.

Grid operators are rewriting interconnection frameworks. Utilities are reassessing cost allocation. Local communities are pushing back against large-scale developments. Labor markets are tightening around specialized construction and engineering roles.

Taken together, these dynamics point to a broader conclusion: The AI infrastructure cycle has shifted from capital scarcity to execution complexity.

Raising billions is no longer the defining challenge. Aligning capital, power, infrastructure, labor, and demand into a functioning, revenue-generating system is.

That alignment does not happen automatically. It requires coordination across stakeholders that have historically operated in silos — investors, developers, utilities, grid operators, equipment suppliers, and end users.

Where mining, AI/HPC, energy, and capital markets converge

As the gap between capital raised and revenue realized becomes more visible, the industry is entering a phase where execution, not ambition, will separate winners from the rest.

That is why we are launching the Energy Investor Forum (EIF).

EIF will bring together the companies and investors shaping the next phase of AI and power-intensive infrastructure, not just for on-stage discussion but for curated meetings and real business development across power, sites, interconnection, financing, customer relationships, and the partnerships needed to convert announced capacity into operating, monetized infrastructure. To get involved, readers can explore sponsorship opportunities, apply to speak, or request an investor attendee pass.

Because the next phase of the AI boom will not be defined by who can raise the most money. It will be defined by who can put it to work.

Regulation News

• Chip Security Bill Advances in House After Super Micro Case

• Proposed GOP Bill Targets Foreign Crypto Mining Hardware

Hardware and Infrastructure News

• Meta boosts investment in West Texas AI data center by over sixfold to $10 billion

• TSMC’s 3nm Chip Capacity Has Become So Constrained That Only “Long-Term, Loyal” Customers Are Getting Priority

• Bitdeer Contracts DCI for 180 MW AI Data Center Development in Norway

• Nebius Plans 310 MW AI Data Center in Finland

• Applied Digital Modifies Ellendale Data Center Leases Following CoreWeave Credit Upgrade

Corporate News

• Bitfarms Nears Keel Rebrand, Plans to Sell $161M in Bitcoin to Fund AI Push

• Duke Energy concludes $2.48 billion divestment of Tennessee gas assets to Spire

• Cango Faces NYSE Delisting Risk Over Low Share Price

• Luxor Launches Commander, a Fleet Management and Profitability Optimization Software for Bitcoin Mining Operations

Financial News

• Google Backs $5B Texas Data Center for Anthropic, Deepening AI Infrastructure Push

• Mistral AI Secures $830 Million Debt Deal for France AI Data Center

• CoreWeave Lands $8.5 Billion GPU-Backed Financing in Latest AI Deal

• NVIDIA Invests $2 Billion in Marvell to Expand AI Infrastructure Ecosystem

• OpenAI Closes $122B Raise at $852B Value as AI Capex Surge Accelerates

Feature

• Microsoft is picking up a Texas data center project OpenAI didn’t want, in a telling sign of how far they’ve drifted apart

• The Same Engine Behind Bitcoin and AI - TheEnergyMag

Email us

Across rural America, the AI infrastructure boom is increasingly taking shape behind closed doors.

For much of the past year, some of the biggest names in AI infrastructure have been hunting for land in rural America the same way they hunt for power: quietly, quickly and often before the public knows what is happening.

The playbook has become increasingly familiar. Developers approach local officials under nondisclosure agreements (NDAs), use intermediaries or shell entities to mask the ultimate buyer, and begin assembling hundreds of acres before residents can organize around what a project might mean for farmland, water use, noise and electricity demand. That secrecy may help companies lock up strategic sites in a fiercely competitive market for AI infrastructure, but it is also becoming a political liability.

Now the backlash is getting harder to ignore.

In Wisconsin, at least five communities have now been linked to NDAs tied to proposed data center developments, according to Wisconsin Watch, which found secrecy arrangements in Beaver Dam, Kenosha, Janesville, Menomonie and most recently Beloit. In Beaver Dam, Meta used shell companies in the early stages of a 520-acre, roughly $1 billion campus. Across the state, seven major data center projects are pending with a combined value of more than $57 billion, making Wisconsin one of the clearest examples of how the AI buildout is colliding with demands for public disclosure.

The pushback has become forceful enough that Microsoft — one of the most aggressive builders of AI data center capacity globally — said this month it would stop using NDAs with local governments for data center development. The company said the old practice had been meant to protect commercially sensitive information, but that transparency with communities had become "paramount." Microsoft added that it is working to identify and terminate active NDAs with local governments, even as it reserved the right to keep using confidentiality protections for certain trade secrets and land acquisitions.

That policy shift says as much about politics as it does about process. Microsoft President Brad Smith said this week that winning over local communities has become essential as towns across the U.S. increasingly protest data center projects. Reuters reported that opposition in parts of the Midwest and Northeast has already contributed to canceled developments over concerns ranging from power prices and water usage to pollution from related energy infrastructure.

What is changing is not just the scale of the projects, but the scale of what communities feel they are being asked to absorb before anyone has had an honest public debate.

In Mason County, Kentucky, for instance, officials have backed a proposed hyperscale campus that would span about 2,080 acres of agricultural land outside Maysville. The plan envisions six data center buildings, multiple substations, stormwater ponds and phased construction beginning in summer 2027 if approvals move forward. County officials have pitched the project as an economic engine that could bring 400 full-time jobs and thousands of construction jobs, while also keeping the identity of the company behind it undisclosed unless the development is ultimately approved.

That secrecy has helped turn the project into a local flashpoint. Realtor.com and local broadcaster WKRC reported that several farming families have rejected lucrative offers to sell. One mother and daughter turned down more than $26 million for part of their land, while another family previously refused nearly $8 million. Officials say at least 20 residents were approached and 18 signed conditional sale agreements, but the holdouts have become symbols of a broader resistance: not necessarily to data centers as such, but to the idea that farmland can be quietly assembled for an unnamed AI customer and only debated in earnest after the map is largely drawn.

A similar drama is playing out in Pennsylvania, where 86-year-old farmer Mervin Raudabaugh rejected more than $15 million from a data center developer for his 261 acres in Cumberland County. Instead, he sold the development rights for about $2 million to preserve the land for agriculture. Realtor.com reported that the developer’s outreach had been persistent, while local preservation advocates said deep-pocketed data center buyers are increasingly targeting large tracts of open farmland that conservation groups cannot match on price.

Taken together, the cases point to a deeper tension in the AI infrastructure boom. For hyperscalers and their development partners, secrecy is often treated as a practical necessity. Land is scarce, power is scarcer, and once word gets out, prices can jump, competitors can swoop in and local opposition can organize before a project is mature enough to present in full. Microsoft said as much in explaining why NDAs became common in the first place.

But the communities being targeted are increasingly viewing that same secrecy as a warning sign.

In Wisconsin, backlash over confidential negotiations has fueled legislative efforts to restrict local governments from signing NDAs with data center developers. Critics say the issue is bigger than whether companies are entitled to protect trade secrets. It is about whether elected officials should be negotiating projects that can reshape land use, tax policy, infrastructure planning and local utility systems without telling the people who will live next to them.

That tension is likely to intensify as the AI buildout moves beyond established data center corridors into rural counties with abundant acreage and relatively easier paths to large-scale development. The industry still has money, urgency and political support on its side. But the farmers turning down eight-figure offers in Kentucky and Pennsylvania are a reminder that not every acre is just another site-control exercise.

In the AI race, land is becoming as strategic as chips and megawatts. The more quietly developers try to secure it, the more likely rural communities are to ask what else they are not being told.

Regulation News

• Pennsylvania House Passes AI Data Center Bill, Setting Up Grid Cost Showdown

Hardware and Infrastructure News

• Illinois City Approves $20B AI Data Center After Marathon Public Hearing

• SoftBank, DOE Back 10GW Ohio AI Data Center and Power Buildout

• Arm Enters Merchant Silicon Market with AI-Focused “AGI CPU”

• Cipher Digital Lands Third AI Data Center Lease, Secures $200M Credit Facility

• Iowa Bitcoin Miner Pushes Expansion as Cedar Falls Weighs Gas Plant

• Bell and HIVE Launch Merritt AI Facility, Formalizing 16.6 MW Canadian Expansio

Corporate News

• Bitfarms Shareholders Approve US Relocation and Rebranding to Keel Infrastructure

• BitFuFu Sues Over Alleged Bitcoin Mining Hosting Fraud in Mississippi

• NVIDIA and Emerald AI collaborate with U.S. energy firms to develop adaptable AI infrastructure for grid management

• Hut 8 leans into modular ‘LEGO block’ model to switch between AI and bitcoin mining

• Auradine Rebrands as Velaura AI, Shifts Teraflux Inventory to In-House Bitcoin Mining

• Crusoe and Form Energy Ink 12 GWh Iron-Air Battery Deal for AI Infrastructure

Financial News

• NextEra Energy Issues $600M in Junior Subordinated Debentures

• Core Scientific increases credit line to $1 billion with J.P. Morgan commitment

• US to pay total $1bn to switch from wind to oil and gas development

• WhiteFiber Reports 61% Revenue Growth in Q4 2025 and Secures $865 Million Nscale Contract

• MARA Sells $1.1 Billion in Bitcoin to Retire $1 Billion in Debt

Email us