News

Market Digest Original

Where the Next Big Data Center Markets Will Be, According to the Digest's Own Data

A Data Center Digest Original StoryThis article was researched and written entirely by AI, without human review or editing, as part of the Data Center Digest's ongoing experiment in AI-powered journalism.

The Digest's map already shows where the old hubs are running out of room. It flags six regions as constrained: Northern Virginia, where Dominion's interconnection queue has reached 26 GW and new large loads wait two to four years; Phoenix, with 18 to 36 month waits; the Bay Area and Los Angeles; New York and New Jersey; and New England. Demand does not disappear when a hub fills up, so it moves. This article asks where it is moving next, using three things the Digest tracks: power headroom and pipeline size from the map, the policy climate from the Legislation tracker and the map's moratorium layer, and momentum from the more than 600 stories and posts the Digest has published since Aug. 15. The calls below are judgments drawn from that data, not a forecast model.

Start with the backdrop. The map covers 41 markets, six of them primary, with roughly 119 GW of capacity and pipeline listed across them. The Legislation tracker follows 255 bills, and 62 percent are classified as anti-development against 18 percent pro. The moratorium layer shows local restrictions, enacted or proposed, in 27 states, 18 of them already in effect. A separate Digest report on stalled projects cited Goldman Sachs saying only 50 to 60 percent of planned capacity will come online as expected over the next two years. Friction is the normal condition almost everywhere, so the markets most likely to grow are the ones where power is available and the friction is lower, already settled, or still ahead.

One pattern runs through the newest big projects: they sit where power infrastructure already exists or is being built alongside them. Homer City, Pennsylvania, is a former coal plant site where a 4.5 GW gas plant is planned, and AWS filed in October for a 36-building, 1,100-acre campus there. Aligned broke ground in September on a 2 GW campus at Shippingport, Pennsylvania, next to a retired coal plant, with its own on-site power. In Maryland, the Quantum Frederick park is built on a former aluminum smelter site and carries about 2.4 GW of combined planned demand. Crusoe's Goodnight campus in the Texas Panhandle, with Google as the end user, connects directly to an adjacent wind farm. The next hubs look less like the incumbents' dense fiber-and-colocation clusters and more like power sites that gained data centers.

The strongest case is Pennsylvania. The map lists three planned projects there that together exceed 7.7 GW: Homer City at 4.5 GW, Shippingport at 2 GW, and Nebius's Schuylkill County campus at 1.2 GW. Only Texas was mentioned in more Digest stories since Aug. 15 (22 for Pennsylvania against 39 for Texas, by a count of state mentions). The policy picture is mixed: the tracker holds 30 Pennsylvania bills, 8 pro and 21 anti, with every pro bill still in committee, and on Oct. 1 the state House Energy Committee advanced three ratepayer bills, including one that would require developers to pay for the utility upgrades that serve them. The moratorium layer lists Pennsylvania as proposed, not enacted. If the projects deliver, Pennsylvania has the pipeline of a top-tier market; the open question is whether new cost rules slow it.

Next is West and Northwest Texas. The map shows West Texas at about 2.5 GW with the 1.2 GW Stargate campus at Abilene leading, and the Panhandle's Goodnight campus is under construction with about $29 billion of investment expected, more than 1 GW by one estimate. Crusoe has also filed plans for two data centers in Jayton. Texas led all states in Digest coverage, with 24 of its 39 stories tagged to construction, power, AI, market, or investment topics. The risk is executive, not legislative: the tracker has no Texas bills, but on Sept. 21 Gov. Greg Abbott ordered the state's environmental regulator to halt all data center permits until audits of the grid and water board finish, expected in mid-December, and the Digest's report cited BloombergNEF saying the pause threatens up to 20 percent of the national pipeline. Expect Texas to keep growing, with the December audit as the checkpoint.

The third likely rise is the Gulf South. Northeast Louisiana is already shown as a primary market, with Meta's Hyperion campus targeting up to 5 GW and a separate $3.6 billion campus at Boyce, and the map rates its power availability as good. Across the border, DeSoto County, Mississippi, is shown at about 2.5 GW, anchored by xAI's roughly 2 GW campus, with AWS and Compass also committing billions elsewhere in the state. The political signal here is thin rather than favorable: the tracker has one Louisiana bill and none for Mississippi, and the two states drew only four Digest stories between them. The main exposure is permitting, since xAI's gas turbine approval in Mississippi is under legal challenge.

Three states have the pipeline but also the loudest pushback. Michigan's Stargate campus in Saline Township ($16 billion, 1.4 GW) broke ground in June, yet the tracker shows 37 Michigan bills, 22 anti and 8 pro, and Republican senators introduced a package in September for a one-year statewide moratorium, although such bills frequently stall in committee. Ohio's Pike County campus is described as up to 10 GW, but three of the four Ohio bills tracked are anti, the moratorium layer lists Ohio as proposed, and the state Supreme Court recently kept a citizen zoning measure off the Wilmington ballot, showing how contested local rules are becoming. Wisconsin's Port Washington campus (902 MW of IT capacity) sits in a state the map marks as having active moratorium action. These markets will keep building, but likely more slowly than their announced pipelines.

The quieter candidates are where the data shows headroom and little noise. The map rates Wyoming, Iowa, the Pacific Northwest, and the Southwest Power Pool states of Kansas, Nebraska, and Oklahoma as high availability, the best tier. Microsoft closed a land purchase for its Cheyenne campus in June and has framed it as the base for a multi-gigawatt buildout. Indiana lists more than 4 GW planned, including Amazon and Meta commitments of $36.8 billion, yet appeared in only two Digest stories. Kansas has a pro bill awaiting its governor, and Evergy reported advanced talks for 2 to 3.5 GW of new demand. The caution is that Missouri and Nebraska each show three anti bills out of three tracked, and a Digest report found Nebraska's grid under rate pressure. If a surprise market emerges, it is probably in this group.

Two wild cards could go either way. New Mexico's Project Jupiter, a 2.45 GW Stargate campus in Dona Ana County, is stalled on a gas pipeline the state land office rejected, and Oracle sent a force majeure notice on Sept. 24, while Google is exploring a closed-loop, air-cooled campus in Lea County. In Utah, the Stratos project in Box Elder County was approved unanimously by county commissioners and has drawn organized opposition from residents. The incumbents also deserve a caution against writing them off. PG&E told CNBC it has a 12.7 GW pipeline of data center requests even as the map calls California critically constrained, a reminder that constrained does not mean closed.

The limits of this analysis are worth stating. Pipeline gigawatts are announcements, not completed capacity. The tracker finds bills through a keyword search, so states with few tracked bills may simply have little visible legislation, and the absence of Texas bills from the tracker is a gap in what it captures more than a sign of calm. Story counts reflect what the Digest's pipeline published, which was interrupted by two outages in late September and early October, and they include some non-US stories published before the Digest narrowed its scope. The signs to watch are specific: the December ERCOT audit in Texas, whether Pennsylvania's ratepayer bills pass, whether Michigan's moratorium bills leave committee, and the new Jupiter pipeline timeline in New Mexico. The next big markets are probably already visible; the data suggests they will be the ones where power arrives before the politics catches up.

Sources
  1. The Data Center Digest: US Data Center Build Map (markets, power availability, grid constraints, moratorium layer)
  2. The Data Center Digest: Legislation tracker (255 bills, sentiment and status by state)
  3. The Data Center Digest: “Opposition, Permits, and Grid Scarcity Stall U.S. Data Center Pipeline”
  4. The Data Center Digest: “AWS Files Plan for 36-Building, 1,100-Acre Data Center Campus in Pennsylvania”
  5. The Data Center Digest: “Aligned Data Centers Breaks Ground on 2GW Pennsylvania Campus at Former Coal Site”
  6. The Data Center Digest: “Pennsylvania House Committee Advances Three Bills to Shield Ratepayers From Data Center Costs”
  7. The Data Center Digest: “Google Confirmed as End User for Crusoe's Amarillo Campus, With $29 Billion Investment Expected”
  8. The Data Center Digest: “Crusoe Files Plans for Two Data Centers in Jayton, Texas”
  9. The Data Center Digest: “Texas Governor Abbott Freezes All TCEQ Data Center Permits Pending ERCOT Audit”
  10. The Data Center Digest: “Michigan Republican Introduces Three Bills for One-Year Data Center Moratorium”
  11. The Data Center Digest: “Ohio Supreme Court Blocks Wilmington Data Center Zoning Ballot Measure”
  12. The Data Center Digest: “Nebraska Grid Faces Rate Pressure as Data Centers and Aging Infrastructure Drive Demand”
  13. The Data Center Digest: “Oracle Sends Force Majeure Notice to Blue Owl Over New Mexico Pipeline Rejection”
  14. The Data Center Digest: “Google Explores First Self-Built Data Center in New Mexico's Lea County”
  15. The Data Center Digest: “Utah Residents Fight Back Against 40,000-Acre Wonder Valley Data Center”
  16. The Data Center Digest: “PG&E Reports 12.7-Gigawatt Data Center Pipeline Amid Wildfire Liability Concerns”
A Data Center Digest Original Story
Construction

AMD Backs Fermi's 222MW Texas Campus With Credit Support Agreement

Fermi has signed a credit support agreement with AMD to support construction of the first data center at Project Matador in Carson County, Texas, with AMD acting as guarantor as Fermi seeks financing. The deal covers an initial 222MW of IT capacity, with a separate agreement outlining a framework for potential expansion to 650MW of gross power capacity. AMD has a direct stake in the project's success, as thousands of its Instinct GPUs are slated for AI training and inference workloads at the site. Fermi's existing contract with AI cloud operator TensorWave, signed in August, is expected to be worth up to $6.5 billion over 15 years, with the first 222MW targeted to come online in the second half of 2027.

Why this matters

The credit support structure is significant because it shifts financial risk for a major greenfield campus onto a chip supplier, a model that could influence how other large-scale AI data center projects secure construction financing. Fermi's share price has fallen from $25 on its first trading day to $3.72, reflecting the company's turbulent trajectory after an unnamed anchor tenant withdrew and two executives departed, making the AMD-backed financing arrangement a critical lifeline for the 17GW campus vision.

Why the Digest selected this story

Named companies AMD and Fermi, a specific financial agreement, and a named US campus project in Amarillo, Texas make this a concrete, newsworthy deal distinct from previously published Texas construction stories.

Read the full story at Data Center Dynamics →
Data Center Dynamics · 3 hours ago
Opposition

Alexandria Officials March Against Starwood's Plaza 500 Data Center Campus

Three Alexandria City Council members, Abdel Elnoubi, Canek Aguirre, and Sandy Marks, joined residents marching on October 10 from Aslin Beer Company to the Plaza 500 site in Lincolnia, where Starwood Capital Group has planned a data center campus. The march launched a two-week campaign calling for a statewide moratorium on new data center development in Virginia, with advocates citing grid impacts, water use, land use, and proximity to dense residential areas. The Plaza 500 project is allowed by right under Fairfax County zoning, having been grandfathered under the county's more restrictive data center ordinance, but Dominion Energy's proposed Edsall Road substation still faces a public facilities review by the Fairfax County Planning Commission, now scheduled for December 2. Councilman Elnoubi said he is exploring legal options if the substation is approved, and called for state legislation giving localities a say when data centers are planned near their borders in adjacent jurisdictions.

Why this matters

The Plaza 500 case illustrates a regulatory gap in Virginia where a large data center campus can proceed without a public hearing because it qualifies as by-right development, leaving a utility substation review as the only formal checkpoint for community input. The campaign for a statewide moratorium, if it gains traction ahead of Virginia's legislative session, could lead to changes that affect the permitting environment for data centers across the state, which hosts one of the highest concentrations of data center capacity in the United States.

Why the Digest selected this story

City officials personally participating in a protest march against a specific named data center project is an unusual and newsworthy escalation of community opposition in a major metro area.

Read the full story at ALXnow →
ALXnow · 2 hours ago
Impact

Harris County Pushes for Regulatory Authority Over 46 Local Data Centers

Harris County officials, including County Attorney Abby Kamen and Commissioners Lesley Briones and Adrian Garcia, held a press conference to detail their limited authority over the county's 46 operating data centers, with at least six more in development. CenterPoint Energy projections cited by Kamen indicate the Houston area could add eight gigawatts of new data center electricity demand by 2029, potentially raising peak demand across the utility's service area by roughly 50%. The county cannot use zoning to control data center locations and faces unresolved legal questions about whether it can even impose a moratorium, pending a determination from Texas Attorney General Ken Paxton. Officials plan to work with other Texas counties through the Conference of Urban Counties to build a legislative agenda for the session beginning in January.

Why this matters

Harris County's situation highlights a structural gap in Texas law that leaves county governments with little recourse as data center development accelerates, a problem that could intensify given projections of eight additional gigawatts of electricity demand by 2029 in the Houston area alone. The pending Texas legislative session represents a near-term opportunity to grant counties new regulatory tools, and the outcome could set precedent for how other Texas counties respond to rapid data center growth.

Why the Digest selected this story

The specific count of 46 data centers combined with a documented regulatory gap in a major Texas county signals a significant governance and impact story with clear public-interest weight.

Read the full story at Click2Houston →
Click2Houston · 4 hours ago
Impact

PG&E and Google Face CPUC Scrutiny Over 250MW San Jose Data Center Connection

The California Public Utilities Commission has repeatedly postponed a vote on an agreement between PG&E and Google to connect a 250-megawatt data center in San Jose, roughly equivalent to the entire electricity load of Marin County. A draft resolution includes a proposed requirement for Google to pay $600,000 per megawatt in advance infrastructure costs, which would total up to $150 million; Elise Torres, assistant attorney at utility watchdog The Utility Reform Network, said both PG&E and Google appear unhappy with that provision. PG&E contends the grid upgrades were already planned and would benefit thousands of homes and businesses regardless of the Google project. The commission's eventual ruling is expected to signal how it will handle Rule 30, a broader set of proposed standards for connecting large electricity customers to the transmission grid.

Why this matters

The commission's decision on the $150 million cost-recovery provision will establish whether large data center operators can be required to pay upfront infrastructure costs rather than passing those expenses to other ratepayers, a first for data centers in California if adopted. The outcome will likely shape how the CPUC applies Rule 30 to future large-load interconnection requests, with consequences for how data center growth is financed and who bears the cost of grid expansion across the state.

Why the Digest selected this story

A specific proposed mega data center in a major US market (Silicon Valley) drawing documented energy and environmental concerns is a distinct impact story not covered in the already-published list.

Read the full story at Spectrum News →
Spectrum News · 5 hours ago
Market

JLL Warns Rising Construction Costs Threaten Speculative Data Center Projects

Real estate services firm JLL, in its Global Construction Costs Index 2026, reports that annual construction cost escalation of 3% to 6% is now the baseline for the industry, up from the 2% to 3% range common before 2020, with a further 3.2% to 4.2% increase forecast for North America in 2027. Matthew Giddings, executive director of technical services with JLL's Data Centers and Critical Environments team, said the economics break down specifically when costs outpace contracted revenue or power delays postpone income while financing costs continue to accumulate. Speculative campuses without anchor tenants, projects with uncertain power delivery, and older facilities being repositioned for AI workloads face the greatest risk. Giddings recommended phased construction aligned with confirmed electricity supply, early engagement of trade partners, and flexible procurement strategies that reserve manufacturing slots without freezing design specifications.

Why this matters

The combination of sustained construction cost inflation and constrained skilled labor availability, particularly among electricians, high-voltage specialists, and commissioning teams, creates conditions where only well-capitalized projects with secured power and committed tenants can reliably reach delivery. As AI drives a wave of speculative campus development, the JLL analysis identifies a structural vulnerability that could lead to project cancellations or financial distress at sites that lack those fundamentals.

Why the Digest selected this story

Construction cost inflation as a market pressure on data center development is a timely economic story with broad industry relevance, not covered in the already-published list.

Read the full story at Data Center Knowledge →
Data Center Knowledge · 1 hour ago
Power

Google Signs 20-Year PPA for $4 Billion Wyoming Fuel Cell Project

Google has signed a 20-year Power Purchase Agreement with American Electric Power to buy electricity from a 702MW fuel cell facility near Cheyenne, Wyoming. AEP holds a 92 percent stake in the project entity, Cheyenne Fuel Cell, with the remaining eight percent held by funds advised by Basalt Infrastructure Partners; Bloom Energy will supply the fuel cells under a previously announced agreement for up to 1GW of capacity. AEP estimates construction costs at $4 billion, with work beginning in 2026 and full operations expected by end of 2028. The announcement coincides with Google's separately reported 2.7GW data center campus under development eight miles south of Cheyenne, with construction slated to finish by 2031.

Why this matters

A $4 billion, 702MW fuel cell project represents one of the largest single clean-energy supply deals tied directly to a specific data center cluster, signaling that hyperscalers are willing to fund novel generation technologies at gigawatt scale to meet surging AI-driven power demand. The structure, combining a major utility, an infrastructure fund, and a fuel cell manufacturer under a 20-year offtake, sets a potential template for financing non-traditional power sources for data center loads.

Why the Digest selected this story

A 20-year, 702MW power purchase agreement between Google and AEP for a fuel cell project is a major, highly specific energy deal involving named companies, a large capacity figure, and an unusual fuel cell technology angle — all strong newsworthiness signals. 1 similar article covering this event were reviewed but not selected (Hart Energy's AEP $4B fuel cell story covers the same deal from the utility investment side; Data Center Dynamics is retained as the more detailed source).

Read the full story at Data Center Dynamics →
Data Center Dynamics · 3 hours ago
AI

OpenAI Promotes Uday Ruddarraju to Lead Stargate Data Center Efforts

OpenAI has appointed Uday Ruddarraju to lead its infrastructure and compute businesses, including the Stargate data center program, replacing Sachin Katti. Ruddarraju joined OpenAI last July as head of compute and infrastructure and was named CTO of Compute this July; he previously served as head of infrastructure engineering at xAI. Katti, who joined from Intel in November 2025 and took over Stargate this March, is expected to remain as an adviser. OpenAI projects spending approximately $750 billion on computing power through 2030, up from an earlier estimate of roughly $600 billion.

Why this matters

The repeated turnover at the top of OpenAI's data center leadership, with multiple executives departing after Katti's own promotion earlier this year, raises questions about execution continuity on Stargate at a time when in-house data center projects remain slow to come online. The influx of former xAI employees into OpenAI's infrastructure division, including Ruddarraju and several named colleagues, represents a significant concentration of talent from one competing organization shaping the buildout of what is projected to be a $750 billion compute investment.

Why the Digest selected this story

OpenAI replacing its data center leadership with an executive from xAI — a direct competitor — is a significant personnel and strategic signal for AI infrastructure, with high reader interest given OpenAI's massive buildout plans. Named companies (OpenAI, xAI) and named individuals (Uday Ruddarraju, Sachin Katti) are strong selection triggers.

Read the full story at Data Center Dynamics →
Data Center Dynamics · 5 hours ago
Cooling

Ziehl-Abegg Executive Outlines Flexible Cooling Design for Shifting AI Workloads

Brooke Zalud, Market Segment Manager for the IT and Telecommunication division of Ziehl-Abegg, has published a commentary arguing that data center cooling systems must be designed around workload ranges rather than fixed design points. The piece contends that liquid cooling does not replace air cooling, noting that air cooling will still be required for the remaining 20 to 30 percent of white and grey space even in liquid-cooled deployments. Zalud recommends zoning, modular capacity additions, and coordinated controls that allow retrofits without facility downtime. The commentary advocates for reserving physical space for future piping, manifolds, heat exchangers, and pumps before those components are needed.

Why this matters

As AI workloads push rack densities higher and hardware roadmaps shift rapidly, cooling infrastructure decisions made today can constrain or enable future facility upgrades; the argument for hybrid air-and-liquid readiness reflects a growing industry recognition that no single cooling medium will serve all deployments. Designing retrofit readiness as a day-one requirement has direct cost implications, since facilities that cannot add cooling capacity within an existing footprint may face earlier and more expensive reconstruction.

Why the Digest selected this story

Future-proofing cooling infrastructure for unpredictable AI density growth is a core operational concern for the industry right now, and Data Center Frontier's framing around designing for the unknown signals substantive technical or strategic content.

Read the full story at Data Center Frontier →
Data Center Frontier · 6 hours ago
AI

Chiplets Gain Traction, Promising Denser and More Efficient Data Centers

Chiplets, small modular dies combined into a single processor package, are moving from niche adoption to a central feature of modern CPU design, with implications for how data centers manage compute density, power efficiency, and infrastructure planning. Unlike traditional monolithic dies, chiplet-based processors separate functions such as compute, I/O, cache, and AI acceleration into distinct smaller dies, which can improve manufacturing yields and give chipmakers more design flexibility across workload types. The commentary notes that if chiplet-based designs deliver more processing power per server without a proportional increase in physical footprint, existing facilities could support heavier workloads without equivalent expansion. The piece also flags that advanced packaging and integration can introduce their own supply chain bottlenecks, even as chiplets reduce some manufacturing constraints.

Why this matters

If chiplets enable higher compute density without requiring proportionally more space or power, data center operators could extract significantly more capacity from existing facilities, reducing pressure to build new ones at the pace currently anticipated. More workload-specific server designs enabled by chiplets could also complicate infrastructure planning, requiring operators to procure and operate a wider variety of hardware optimized for AI, analytics, storage, or high-performance computing rather than general-purpose platforms.

Why the Digest selected this story

Chiplet-based server design is an emerging hardware trend with direct implications for data center compute density and procurement, and Data Center Knowledge's treatment suggests forward-looking analysis of infrastructure impact.

Read the full story at Data Center Knowledge →
Data Center Knowledge · 8 hours ago
AI

Google Sourcing Executive Allyson Fealtman Joins OpenAI After 16 Years

Allyson Fealtman, who led a team of 11 commodity managers at Google covering liquid and air cooling, interconnect, and mechanical data center equipment, has joined OpenAI. She spent 16 years at Google before making the move this past May, announcing it publicly this week. At OpenAI, her focus will be strategic sourcing for high-speed interconnect and rack power. Prior to Google, she held supply chain roles at Infinera, PowerFile, and supercomputing firm SGI.

Why this matters

The move signals that OpenAI is actively recruiting experienced infrastructure sourcing talent to scale its data center supply chain for high-speed interconnect and rack power. Fealtman's 16-year tenure managing commodity procurement at Google gives OpenAI direct access to deep supplier development expertise at a time when competition for data center components is intensifying.

Why the Digest selected this story

A senior Google data center sourcing executive moving to OpenAI signals intensifying talent competition between hyperscalers and AI labs for infrastructure leadership. This follows a similar already-published story about a Google design head joining Anthropic, making the trend notable and distinct as it involves a different executive and a different destination company.

Read the full story at Data Center Dynamics →
Data Center Dynamics · 3 hours ago
Policy

Five States Move to Regulate Data Centers as Federal Ratepayer Bill Stalls

California Governor Newsom signed seven data center bills on September 21, 2026, including a special electricity rate measure and energy, water, and environmental reporting requirements; Nevada Governor Lombardo issued Executive Order EO-2026-005 requiring developers to cover all project-related costs and execute binding Community Support Commitments. In Texas, Governor Abbott halted all data center-related permits, while Virginia Governor Spanberger signed an executive order implementing the Virginia Data Center Accountability Framework, banning nondisclosure agreements and expediting energy and noise regulations. At the federal level, the Ratepayer Protection Act passed the House 417-3 on September 16 but was blocked in the Senate the following day when Sen. Martin Heinrich objected to unanimous consent, arguing the bill relies on a voluntary framework.

Why this matters

The simultaneous state-level regulatory actions across California, Nevada, Texas, Virginia, and New York represent a broad geographic shift in how governments are responding to data center growth, with consequences for siting, permitting, and cost allocation for developers. The near-unanimous House passage of the Ratepayer Protection Act, which would require large-load customers with peak demand of 100 megawatts or more to cover full incremental grid upgrade costs, sets a precedent for federal cost-shifting policy even as its Senate path remains uncertain.

Why the Digest selected this story

A monthly federal policy roundup from a major law firm tracking sustainable energy and infrastructure regulation is a high-value reference for industry decision-makers; the October 2026 edition is fresh and likely covers developments relevant to data center power and permitting.

Read the full story at Mintz →
Mintz · 4 hours ago
Impact

Data Center Water and Energy Costs Draw Scrutiny Amid Transparency Gaps

U.S. data centers consumed 449 million gallons of water per day as of 2021, according to the Environmental and Energy Study Institute, while the Environmental Law Institute reports their share of U.S. electricity use has risen from 1.9% in 2018 to 4.4% now, with projections reaching 12% by 2028. Economist Nicholas Muller reported that data centers caused approximately $25 billion in pollution and health-related costs in the U.S. last year. In Tucson, Arizona, a contractor for Beale Infrastructure's Project Blue was caught illegally trucking city water to a 290-acre data center site in May 2026, following the Tucson City Council's 2025 vote against supplying city water to the project; the project has since been approved by the Arizona Department of Water Resources and secured a 10-year energy agreement with Tucson Electric Power. Critics point to gaps in corporate disclosure, noting that Google does not publish a company-wide water-efficiency metric and that Microsoft has not disclosed location-specific water and power data despite stating net-zero commitments.

Why this matters

The Tucson case illustrates how state-level approvals can override local objections, setting a precedent that could affect how other municipalities attempt to block data center development on resource grounds. The widening gap between corporate sustainability disclosures and location-specific resource data makes it difficult for regulators, ratepayers, and communities to assess the actual environmental burden of individual facilities.

Why the Digest selected this story

A published assessment of whether tech companies are meaningfully reducing AI's environmental damage taps into a major ongoing debate around carbon, water, and energy use at US hyperscaler facilities. The Fulcrum piece appears to offer documented analysis rather than opinion, making it editorially distinct from advocacy pieces.

Read the full story at The Fulcrum →
The Fulcrum · 6 hours ago
Opposition

Ohio Residents Push Statewide Moratorium as Data Center Backlash Grows

A four-hour Ohio legislative hearing in June drew applause and anger as residents demanded bans and moratoria on data centers, and organizers gathered signatures for a statewide ballot measure that could go before voters in 2027. Ohio has 21 local ballot measures on data centers, the most of any state, with 10 communities including Granville, Piqua, and Urbana set to vote on local restrictions. Tourism guide Nikki Gerber and attorney Austin Baurichter lead Conserve Ohio, the group driving the statewide campaign; Gerber's focus is a proposed 1.3-gigawatt Amazon Web Services data center about 15 miles from her campground in Adams County. The Ohio General Assembly's Select Committee on Data Centers has met five times but has not taken up a moratorium measure, with majority Republicans declining to advance one in late September.

Why this matters

Ohio's Columbus region is one of the nation's top data center development markets, making a successful moratorium campaign there a potential model for opposition efforts in other high-growth states. With about 40 local ballot measures across seven states according to Ballotpedia, the November midterm elections represent the broadest simultaneous electoral test of community resistance to data center development yet seen.

Why the Digest selected this story

Inside Climate News spotlights Ohio community resistance to data center development, signaling a broadening geographic front of opposition beyond previously covered states like Michigan, Texas, and Pennsylvania. The 'revolt' framing and state-level focus make this a distinct and newsworthy opposition story.

Read the full story at Inside Climate News →
Inside Climate News · 5 hours ago
Policy

Developers Use Lawsuits, Secrecy, and Cash to Override Small-Town Data Center Vetoes

Reporters from newsrooms across five states examined tactics data center developers use against local opposition, finding that legal pressure, nondisclosure agreements, and cash payments have repeatedly overridden community objections. In Saline Township, Michigan, a community of fewer than 3,000 residents settled a developer lawsuit and made way for a $43 billion Oracle and OpenAI project after determining it could not afford to fight in court; trustee Dean Marion said, "None of us up here wanted this thing." The developer, Related Digital, describes the Saline project as a 2.2-million-square-foot, gigawatt-scale facility that will generate more than $1 billion in tax revenues in the decade beginning in 2030. In at least four New Jersey communities, developers filed lawsuits after local governments moved to exclude them, and one Pennsylvania township developer offered $10,000 to every household to overcome opposition.

Why this matters

The pattern documented across five states shows a structural imbalance: small municipalities with limited legal budgets face well-resourced developers willing to litigate, effectively making local zoning decisions reversible through financial attrition. The Saline Township case is particularly significant because local leaders capitulated again just one month after the groundbreaking, backing down on tax breaks to avoid another costly lawsuit.

Why the Digest selected this story

Governing magazine's analysis of how data center developers outmaneuver small-town opposition is a timely policy story with national implications, covering the structural power imbalance in local permitting — distinct from specific moratorium or zoning votes already published.

Read the full story at Governing →
Governing · 6 hours ago
AI

Meta Reveals Ohio Cluster Details and Five-Gigawatt Louisiana Data Center

At the OCP Global Summit in San Jose, Meta infrastructure head Santosh Janardhan detailed the company's Prometheus data center complex in New Albany, Ohio, and previewed its next project, Hyperion, in Richland Parish, Louisiana. Prometheus spans more than 30 structures connected by three to four million kilometers of fiber-optic cable and uses a custom "Ethernet super spine" capable of tens of petabits per second to eliminate inter-building latency. Meta also replaced traditional steel-and-concrete construction with tent structures each at least two football fields long, and developed an AI-formulated concrete mix that uses 19% less carbon, cures 43% faster, and is cost-neutral; Meta is open-sourcing the concrete tool. Hyperion, described by Janardhan as "the biggest data center on the planet," is a five-gigawatt cluster covering about 4,000 acres at a single location.

Why this matters

A five-gigawatt single-site cluster represents a scale of power demand and land use that far exceeds prior industry benchmarks, setting a new reference point for grid planners, utility regulators, and land-use authorities. Meta's open-sourcing of its lower-carbon concrete formulation could influence construction practices across the broader data center industry, which relies heavily on concrete at massive scale.

Why the Digest selected this story

A Meta executive presentation at OCP Global Summit drawing strong audience reaction signals meaningful new data center technology announcements from one of the largest hyperscalers. The OCP Summit is a major industry venue and Meta's US infrastructure buildout is directly relevant.

Read the full story at Fierce Network →
Fierce Network · 4 hours ago
AI

Anthropic Names Former Google Manager as Data Center Portfolio Lead

Jacqueline Dory Prescott has joined Anthropic as data center portfolio lead, coming from Google where she managed leased third-party powered shell data center projects across North and South America, overseeing a portfolio of 7 gigawatts across more than 20 buildings on seven campuses. Prescott also previously held roles at Stripe and Facebook, including seven years at Facebook as a program manager for data center energy and infrastructure. Her hire is part of a broad pattern: Anthropic has recruited numerous former Google data center professionals, including head of data center infrastructure Winnie Leung, design lead Liwen Mao, electrical lead Adam Johnson, and, earlier this month, Google's former head of data center global design and construction, John M. Wilson.

Why this matters

The concentration of senior Google data center talent at Anthropic indicates the AI company is rapidly assembling the internal expertise needed to build and operate large-scale infrastructure independently, rather than relying solely on cloud providers. The pace of hiring, with multiple Google veterans joining across construction, electrical, operations, security, and finance roles, signals Anthropic is moving toward significant direct data center investment.

Why the Digest selected this story

Anthropic's hire of a senior Google data center executive signals the AI lab is aggressively scaling its infrastructure capacity planning — a notable talent move distinct from the already-published John Wilson/Anthropic EMEA story, as this involves a different executive and a different role.

Read the full story at Data Center Dynamics →
Data Center Dynamics · 7 hours ago
Power

Google and Constellation Sign 3.59 GW Nuclear Deal Worth Over $4.3 Billion

Alphabet's Google and Constellation Energy signed a power purchase agreement for 3.59 gigawatts of electricity, including a 20-year PPA for 890 MW sourced from multiple Constellation nuclear plants and a separate 15-year supply agreement covering 2,700 MW. The deal will drive more than $4.3 billion in new investment by Constellation to upgrade 11 nuclear units across Illinois, Pennsylvania, and New Jersey. Constellation shares surged over 12% following the announcement, while the VanEck Uranium and Nuclear ETF gained roughly 5.5% on the day.

Why this matters

The scale of this agreement, 3.59 GW under contract over terms of 15 to 20 years, signals a sustained shift in how hyperscalers are securing baseload power for AI infrastructure, directly affecting long-term utility economics and nuclear plant investment plans. It follows a pattern of similar deals including Microsoft's $16 billion Three Mile Island agreement and Meta's 2,600 MW PPA with Vistra, showing that contracted revenue from tech companies is reshaping capital allocation across the nuclear sector.

Why the Digest selected this story

A 3.6 GW nuclear deal between Google and Constellation is a landmark energy procurement story involving two major named companies and a significant grid capacity figure. The TradingView article is selected as it references the full 3.6 GW deal scope; the W.Media article covering 890 MW of the same deal is the same underlying event. 1 similar article covering this event was reviewed but not selected.

Read the full story at TradingView →
TradingView · 3 hours ago
Policy

Texas Governor Abbott Halts Data Center Permits Amid 474 GW Grid Demand Backlog

Texas Governor Greg Abbott issued a series of directives between June and September 2026 effectively pausing new data center approvals through both the ERCOT interconnection process and the Texas Commission on Environmental Quality permitting process. On August 3, Abbott ordered a comprehensive audit of all data centers seeking ERCOT interconnection; on September 21, he directed TCEQ to halt all permit issuances to data centers until that audit is complete. ERCOT is currently evaluating approximately 474 gigawatts of electricity demand, more than five times Texas's record peak demand, with roughly 90% of new power requests attributed to data centers. ERCOT has said it intends to complete its review by December 2026.

Why this matters

Texas is the fastest-growing data center market in the United States and the second largest behind Virginia, making a statewide pause on interconnection approvals and environmental permits a significant constraint on an enormous volume of proposed capacity. The audit's five areas of inquiry, covering state financial assistance, onsite generation, water use, community impacts, and ownership, set a potential template for how states manage large-load growth as grid pressure intensifies.

Why the Digest selected this story

Texas pausing data center approvals is a significant regulatory action by a major data center market, triggered by multiple concerns including grid strain and water use. This represents a formal governmental pause distinct from the previously published story about six new markets and Texas permit pauses, with this article providing dedicated legal/regulatory analysis from Environment, Land & Resources.

Read the full story at Environment, Land & Resources →
Environment, Land & Resources · 4 hours ago
Impact

California Regulators Weigh $150 Million Google Grid Fee for San Jose Data Center

The California Public Utilities Commission has repeatedly postponed a vote on an agreement between PG&E and Google to connect a 250-megawatt data center in San Jose, with a key sticking point being a proposed requirement for Google to pay $600,000 per megawatt in upfront infrastructure costs. At 250 MW, the facility would draw approximately the entire electrical load of Marin County, according to Elise Torres, assistant attorney at utility watchdog organization The Utility Reform Network. PG&E argues the grid upgrades were already planned and would benefit thousands of homes and businesses regardless of the Google project. The commission's decision is expected to influence how it applies cost-allocation standards to future large-load interconnection requests under the proposed Rule 30 framework.

Why this matters

A requirement for Google to pay up to $150 million in upfront infrastructure costs would be the first such charge applied to a data center in California, setting a direct financial precedent for how grid upgrade costs are allocated between large tech customers and existing ratepayers. The commission's eventual ruling will also signal the standard it plans to apply to exceptional case filings and the broader Rule 30 interconnection rules currently under consideration.

Why the Digest selected this story

A proposed mega data center in Silicon Valley drawing energy and environmental scrutiny is a concrete, locally significant impact story involving a named geography and documented community concerns. This is distinct from previously published opposition and impact stories covering different locations.

Read the full story at Spectrum News →
Spectrum News · 5 hours ago
Construction

Fluidstack Breaks Ground on $4 Billion 1.5 GW Texas AI Data Center

Fluidstack began construction on a $4 billion first-phase AI data center campus in Cameron County, Texas, capable of supporting up to 1.5 gigawatts of load, with the project expected to generate more than 2,500 construction jobs and several hundred permanent positions. The company disclosed on September 21 that it requested no local property-tax abatements and plans to self-fund the energy, water, and infrastructure needed for the campus, which will use a recirculating closed-loop cooling system. Separately, Microsoft filed plans on September 15 for its ATL50 campus in Union City, Georgia, a two-building, approximately 910,000-square-foot development on 88 acres projected for completion in 2032. An undisclosed developer is also evaluating Project Gravel, an eight-building campus of approximately 1.6 million square feet on 160 acres in unincorporated Pima County, Arizona, with full buildout contemplated by 2029.

Why this matters

The breadth of projects disclosed between September and early October 2026, spanning Texas, Georgia, and Arizona, reflects the continued geographic expansion of large-scale AI infrastructure, with individual campuses now regularly planned at hundreds of megawatts to over a gigawatt. Community opposition, as seen when public pushback halted negotiations in Mahomet, Illinois, and regulatory complexity in Texas and California are adding measurable friction to project timelines, making the gap between announced and delivered capacity an increasingly significant industry variable.

Why the Digest selected this story

Data Center Frontier's pipeline expansion analysis signals broad construction market momentum across North America, a high-interest topic for the industry covering land acquisition and capacity growth under resource and regulatory pressure.

Read the full story at Data Center Frontier →
Data Center Frontier · 6 hours ago
AI

Ukrainian Drone Strikes Hit Third Yandex Data Center, Knocking Out 84 Services

A Ukrainian drone strike hit Yandex's Vladimir data center, the third of the company's five large facilities to be struck within a single week, forcing Yandex to pause operations there. The two earlier strikes hit the Sasovo data center, which houses two of Yandex's three supercomputers and suffered a fire, and the Kaluga facility. Yandex reported that at least 84 services were taken down and that its platform continues to operate in limited mode, though the company stated all financial records remain intact, stored in multiple copies across different platforms. Cloudflare's traffic tracker recorded bot HTTP requests from Russia falling to just 0.3 percent of their maximum on the morning of October 11, recovering only to approximately 10 percent.

Why this matters

Three strikes against Yandex's data center infrastructure within one week demonstrate how concentrated physical data center assets create acute, measurable disruption to cloud services at national scale, with 84 services simultaneously unavailable and payment systems impaired across Russia. The near-total collapse of bot HTTP traffic originating in Russia, as measured by Cloudflare, illustrates the cascading internet-level consequences that physical attacks on major data center operators can produce.

Why the Digest selected this story

A third strike on the Yandex data center in Russia is a notable escalation of ongoing infrastructure attacks, directly relevant to data center resilience and security. The previously published story covered two strikes in consecutive days, making this third strike a new and distinct event.

Read the full story at Data Center Dynamics →
Data Center Dynamics · 2 hours ago
AI

Ukrainian Drones Strike Two Yandex Data Centers in Consecutive Days

Ukrainian drones struck Yandex data centers on consecutive days in early October, first hitting the Sasovo facility in Russia's Ryazan region and then the company's largest facility in the Kaluga region, southwest of Moscow. Sasovo houses two of Yandex's three supercomputers, named Lyapunov and Chervonenkis, built on Nvidia A100 chips and used to train YandexGPT; Yandex has declined to confirm whether the supercomputers were damaged. Ukrainian drone maker Firepoint said it used FP-1 one-way attack drones against the Kaluga site, and Kaluga Oblast governor Vladislav Shapsha reported nine people injured in the broader regional strike, though Yandex said no one was hurt at the data center itself. Downdetector logged more than 1,300 user reports tied to outages in Yandex Cloud, Yandex Documents, and Yandex Tables, with Russian Railways, MegaFon, and several Russian media outlets that lease Yandex infrastructure also affected.

Why this matters

The strikes demonstrate that physical data center infrastructure is now a direct military target, not merely subject to chip-export restrictions, raising the security calculus for any operator planning facilities in or near conflict zones. A one-way attack drone costing far less than a single GPU rack was sufficient to disable modules at facilities supporting a national AI model, a precedent that will force insurers and developers globally to reassess physical-security risk in ways that were largely theoretical before these attacks.

Why the Digest selected this story

A drone strike taking out supercomputers at major Russian tech company Yandex's data centers is a high-impact, novel infrastructure security story with geopolitical significance not previously covered in the Digest.

Read the full story at Startup Fortune →
Startup Fortune · 3 hours ago
Policy

Michigan City Council Introduces New Data Center Moratorium Amid Internal Disputes

Michigan City, Indiana's city council introduced a new ordinance that would suspend approval of new data center developments until the city updates its zoning ordinance or until July 31, 2028, explicitly excluding both phases of Google's current project. The new ordinance replaced a prior one that had been tabled, but council members who drafted the original, including Joseph Nelson and Nancy Moldenhauer, complained they were not consulted before the replacement was filed. Council member Bryant Dabney said the pause would allow officials to evaluate Google's project directly for data on noise, water usage, and energy before permitting additional development. Because the new ordinance was not filed in time for a vote at Tuesday's meeting, a formal first reading is scheduled for the next council session.

Why this matters

The dispute illustrates how local governments are struggling to coordinate data center oversight even among themselves, with procedural infighting delaying the adoption of rules as Google's project moves forward. Community concerns about diesel generator emissions, fuel storage, and a planning department reportedly drafting zoning rules without public input signal that Michigan City's process could become a contested template for other small municipalities facing their first large-scale data center developments.

Why the Digest selected this story

An active municipal moratorium debate signals continued regulatory friction around data center siting; the specific focus on moratorium language suggests a formal vote is imminent, making this timely and distinct from previously published moratorium stories in other cities.

Read the full story at WFYI →
WFYI · 5 hours ago
Opposition

Developer and Plymouth Township Trade Accusations Over 2 Million Square Foot Data Center Plan

Developer Brian O'Neill's proposal to convert the former Cleveland-Cliffs steel mill at 900 Conshohocken Road in Plymouth Township into a 2 million-square-foot data center has escalated into open conflict, with council Chair Lynne Viscio accusing O'Neill of trying to "demand approval by tantrum" after he rejected all 43 conditions the council presented and filed a legal challenge to the township's zoning ordinance. O'Neill disputes that characterization and says he negotiated in good faith and accepted the majority of conditions. He has cited roughly $21 million in projected annual local tax revenue, a figure Plymouth's own planning agency has publicly disputed. A substantive zoning hearing board session is scheduled for August 6, where O'Neill's team is expected to present evidence and call witnesses; a petition opposing the project has drawn more than 3,600 signatures.

Why this matters

The Plymouth Township dispute shows how data center approvals are becoming protracted legal and political battles even after a developer resubmits a nearly identical plan following an earlier rejection. The parallel fight at O'Neill's separate 4.6 million-square-foot, five-property Upper Merion Township campus, where planning commissioners already declined to recommend three of five plans, points to a regional pattern of resistance that could affect how developers and municipalities across Pennsylvania approach large-scale data center negotiations.

Why the Digest selected this story

A recurring community opposition campaign against a specific named data center project in Pennsylvania is a strong Opposition signal; the 'second fight' framing suggests escalating local resistance not previously covered in the Digest.

Read the full story at MONTCO.Today →
MONTCO.Today · 4 hours ago
Power

Natural Gas Supplies 40 Percent of U.S. Data Center Power as AI Demand Surges

AI data center server racks require 40 to 100 kilowatts of electricity each, compared to 5 to 10 kilowatts for conventional racks, and a single AI query uses 10 to 15 times more electricity than a traditional web search, according to figures cited from Hanwha Data Centers. Natural gas supplied 40% of U.S. data center energy in 2024, with renewables at 24%, nuclear and coal at 20% and 15% respectively, according to the International Energy Agency; RBC Capital Markets commodity strategist Christopher Louney projected natural gas demand from data centers could increase by close to 20% by 2030. Hyperscalers and developers are projected to spend $7.6 trillion on AI data centers through 2031, according to Goldman Sachs, while the U.S. Energy Information Administration projects data center growth will help increase U.S. solar generation by 21% in 2026 and 18% in 2027. In June, Chevron and Microsoft announced a 20-year deal for Chevron to supply off-meter natural gas for a 2.67 gigawatt power plant and AI data center near Pecos, Texas.

Why this matters

The scale of power demand is reshaping energy markets: the U.S. Energy Information Administration projects 44.9 billion cubic feet per day of new pipeline capacity coming online in 2026 and 2027, with 66% concentrated in Texas and Louisiana, states already seeing significant data center growth. The trend toward on-site gas generation and off-meter power plants, encouraged by the Trump administration, signals a structural shift in how large data centers relate to public utility grids, with consequences for ratepayers and emissions policy.

Why the Digest selected this story

The snippet directly addresses grid strain and energy mix shifts driven by AI data center demand — a core Power & Energy topic — and appears to provide fresh regional or national context beyond what the Digest's already-published ratepayer and demand-response stories covered.

Read the full story at reflector.com →
reflector.com · 6 hours ago
Construction

Five Global Markets Reshape AI Data Center Geography Beyond U.S. Corridors

Spain's Aragón region has attracted more than 60 billion euros in prospective digital infrastructure investment, with Amazon Web Services outlining approximately 33.7 billion euros in regional plans and Microsoft identifying investments approaching 10 billion euros, driven by an accelerated permitting designation that can compress timelines to under a year. India has reached approximately 1.75 gigawatts of operational data center capacity, while Latin America's largest markets grew 41.3% year over year, led by expansion in Querétaro, Mexico. Singapore is combining selective capacity expansion with gigawatt-scale development pipelines in Malaysia, Thailand, and Indonesia, while China is expanding around energy-rich regions such as Inner Mongolia. Spain's national government confirmed on October 8 plans to advance tighter oversight of the data center industry, and environmental groups have challenged whether Aragón's accelerated reviews adequately address electricity demand, water consumption, and public participation.

Why this matters

The scale of capital committing to markets outside established U.S. corridors reflects how constrained power availability and regulatory friction at home are redirecting hyperscale investment to regions offering faster approvals and cheaper energy, with consequences for how electricity grids, water systems, and public policy in those regions develop. The concurrent regulatory backlash in Aragón and other markets shows that speed-to-permit advantages can erode quickly once project concentrations become visible to national governments and local communities.

Why the Digest selected this story

Data Center Frontier's identification of five specific emerging markets reshaping AI infrastructure geography offers fresh geographic intelligence for site selection and investment decisions, going beyond single-project announcements.

Read the full story at Data Center Frontier →
Data Center Frontier · 7 hours ago
Construction

Proposed 20-Story Kansas City Data Center Tower Tests Vertical Build Economics

Developer Revitalization Unlimited, working with Chicago architecture firm Skidmore, Owings & Merrill, has proposed a 20-story data center tower on a 0.23-acre parcel in downtown Kansas City, Missouri, designed to support 24 MW of IT load across 141,000 gross square feet, with only two floors housing computing equipment and approximately 109,000 square feet devoted to mechanical space. Founder and CEO Steve Austin estimated total development costs at $183 million, or roughly $1,300 per square foot, with about $107 million of the $167 million in hard construction costs going toward mechanical, electrical, plumbing, and fire protection systems. Austin said the facility would use 30 MW of on-site Bloom fuel cells as the primary electricity source, with natural gas listed as an expected operating expense, and targets a power usage effectiveness of 1.25. A related city development plan ordinance is listed as on hold until October 13.

Why this matters

The proposal tests whether the cost premium of vertical urban construction, estimated at a higher per-square-foot figure than suburban alternatives, can be justified by proximity benefits for latency-sensitive workloads such as financial transactions and inference. IDC senior research manager Luis Fernandes noted that most vendors design power and cooling equipment for horizontal placement, meaning vertical builds introduce practical equipment-placement and replacement challenges that the broader industry has not widely solved.

Why the Digest selected this story

Vertical data center design is an emerging construction trend driven by urban land scarcity; Data Center Knowledge's analysis of cost-proximity trade-offs is timely for developers in constrained markets.

Read the full story at Data Center Knowledge →
Data Center Knowledge · 8 hours ago
AI

Oracle, Broadcom, SpaceX Seek Over $90 Billion in AI Chip Financing

Oracle, Broadcom, and SpaceX are each pursuing large-scale financing to fund AI hardware purchases, according to the Wall Street Journal. Broadcom is seeking more than $50 billion to support its custom AI chip development with OpenAI, while SpaceX is arranging a $40 billion deal, comprising $10 billion in bank loans and $30 billion in investment-grade debt, to purchase Nvidia hardware, with Apollo expected to lead that transaction and it expected to close in 2027. Oracle is in talks with Apollo and Goldman Sachs, though few details of its deal were disclosed, with financing intended to bridge the gap between hardware payment and incoming cloud revenue. Separately, SpaceX CEO Elon Musk confirmed the company has held discussions with TSMC about its $16.8 billion TeraFab project in Grimes County, Texas.

Why this matters

The scale of these financing deals, collectively exceeding $90 billion, signals that AI hardware acquisition costs have grown large enough to require structured debt and private equity arrangements rather than conventional corporate financing. The involvement of firms like Apollo and Goldman Sachs in chip financing reflects a structural shift in how hyperscale AI infrastructure is being capitalized, with repayment tied to future cloud revenue streams rather than existing assets.

Why the Digest selected this story

Three major technology companies simultaneously pursuing large-scale AI compute financing signals a significant capital mobilization moment for the sector. The involvement of Oracle, Broadcom, and SpaceX together in a single financing report is a high-profile multi-company development warranting coverage.

Read the full story at Data Center Dynamics →
Data Center Dynamics · 2 hours ago
Construction

Project Longhorn Files $366 Million Data Center in Bastrop County, Texas

A $366 million data center project called Project Longhorn, Building One has been filed with the Texas Department of Licensing and Regulation for a site on Earl Callahan Road in Cedar Creek, Bastrop County. The facility will span 498,165 square feet, with construction slated to begin November 30, 2026, and complete by January 2028. At full buildout, the project is expected to deliver 710MW of site-rated redundant power from simple-cycle natural gas turbines, supporting 490MW of IT capacity. Representatives for Black Chamber, Pacifico Energy, and Anthropic did not respond to requests for comment, though Anthropic has been reported to be in talks to lease capacity at a major data center campus in the Cedar Creek area.

Why this matters

The filing proceeds despite Texas Governor Greg Abbott's September order halting new data center permits from the state's environmental agency pending an ERCOT grid audit, suggesting some projects are advancing through alternative regulatory pathways. The scale of 490MW of IT capacity at a single building, combined with a dedicated on-site gas generation plant, illustrates the degree to which large AI tenants are driving purpose-built power infrastructure rather than relying on grid connections.

Why the Digest selected this story

A $366 million data center project tied to Anthropic, one of the most prominent AI companies, filing in Texas is a significant construction and AI infrastructure story. The Anthropic connection and large dollar figure elevate this above routine campus announcements.

Read the full story at Data Center Dynamics →
Data Center Dynamics · 3 hours ago
Opposition

AWS Drops Government NDAs, Pledges $1 Billion Amid 100-Plus Moratoriums

AWS CEO Matt Garman announced on October 2 that Amazon Web Services will stop requiring government agencies to sign nondisclosure agreements before data center projects become public, and will launch a program called Built Together committing more than $1 billion over five years to host communities. Garman cited more than 100 data center moratoriums currently under consideration by local governments as a key motivator, warning that the trend could threaten the United States' lead in artificial intelligence. The announcement came two days after Representative Jamie Raskin, the ranking member of the House Judiciary Committee, sent letters to Amazon, Google, Meta, and Oracle demanding information about their NDA practices tied to data center projects. Amazon said it has no plans to revisit existing NDAs, arguing most stop mattering once a project becomes public.

Why this matters

More than 100 active or pending local moratoriums represent a concrete regulatory threat to hyperscale expansion timelines, and Amazon's policy reversal shows that community opposition has reached a scale large enough to alter standard operating practices at the largest cloud provider. The sequence of congressional inquiry followed within 48 hours by a policy change sets a precedent that legislative pressure can move corporate NDA practices faster than community advocacy alone.

Why the Digest selected this story

Amazon reversing its NDA policy specifically due to community opposition is a notable development in the ongoing tension between hyperscalers and local communities. Note: a related story about the AWS CEO dropping NDAs with government agencies was already published, but this snippet focuses on local community backlash driving the broader NDA policy change, making it a distinct angle.

Read the full story at Startup Fortune →
Startup Fortune · 5 hours ago
Policy

Reading, Pennsylvania Advances Fourth Draft of Data Center Zoning Rules

Reading City Council reviewed a 23-page, fourth-draft zoning amendment during a committee of the whole meeting that would allow data centers as a conditional use in heavy manufacturing and manufacturing commercial districts, requiring individual council approval for each project. The proposal addresses setbacks, floodplains, landscaping, noise, vibration, water and wastewater infrastructure, electricity use, and environmental impacts. One unresolved issue is the setback distance from preservation districts: the draft considered 1,000 feet but mapping showed that could eliminate so much industrial property that the ordinance would become overly restrictive, and the committee is now considering reducing it to 500 feet. Officials expect to finalize edits before sending the draft to Berks County for a required 30-day review, followed by planning commission review, a public hearing, and council consideration.

Why this matters

Reading's process illustrates the regulatory gap many mid-sized cities are rushing to close: the city initiated a formal municipal curative amendment after councilman Jaime Baez Jr. identified the absence of any data center provisions in the existing zoning ordinance. The conditional use framework being developed, including ongoing inspections and compliance monitoring after a facility opens, could serve as a model for similarly sized dense urban communities managing data center siting pressure.

Why the Digest selected this story

A municipality formally deliberating data center zoning rules reflects the accelerating trend of local governments moving to regulate the sector. Reading joining this list of cities enacting formal land-use controls is regionally significant.

Read the full story at Reading Eagle →
Reading Eagle · 6 hours ago
Power

Hybrid Supercapacitors Emerge as Active Buffers for AI Cluster Power Surges

Frank DeLattre, President Americas at Musashi Energy Solutions, argues that synchronized power ramps from large GPU clusters are turning energy storage from standby backup into an active, frequently cycling component of data center power architecture. Unlike conventional data centers where independent server loads average out, AI clusters can cause thousands of GPUs to increase power consumption simultaneously, creating fast transient loads that upstream infrastructure struggles to follow. Musashi's hybrid supercapacitor technology uses a lithium-doped graphite electrode to combine higher energy density with the fast response and high cycling capability of capacitors, with the company claiming its system can exceed 100,000 cycles at 100% depth of discharge. Placing storage closer to compute units reduces transient propagation through upstream equipment including UPS systems, generators, and utility connections.

Why this matters

The shift from treating battery systems as emergency insurance to deploying high-cycle storage as a real-time power buffer changes the economic and engineering calculus for data center electrical infrastructure procurement, particularly as AI clusters grow in size and synchronization. If storage must cycle tens or hundreds of thousands of times over a service life rather than rarely, cycle life, cooling requirements, footprint, and replacement intervals become significant cost factors alongside acquisition price.

Why the Digest selected this story

The framing of AI-driven energy storage as active power infrastructure — rather than passive backup — signals an important shift in how data centers are being integrated into grid operations. Data Center Frontier's analysis of this trend is timely given ongoing grid strain coverage.

Read the full story at Data Center Frontier →
Data Center Frontier · 7 hours ago
AI

HPE Launches Four AMD Venice-Powered ProLiant Gen13 AI Servers

HPE on October 7 announced four ProLiant Gen13 servers powered by AMD's sixth-generation Epyc 9006 SP7-series CPUs, code-named Venice, targeting enterprise AI inference, agentic AI, and traditional workloads such as analytics and virtualization. The flagship ProLiant DL585a Gen13 supports up to eight GPUs and two Venice processors for demanding workloads including AI inference and retrieval-augmented generation fine-tuning, with availability in March 2027 and support for GPUs from Nvidia, AMD, and Intel. The compact 1U ProLiant DL525, available November 2026, features a single Venice chip with up to 256 cores for AI inference and traditional enterprise tasks. Two rack-scale CPU systems, the liquid-cooled ProLiant XD245 and air-cooled ProLiant XD285, target large enterprises and specialist cloud providers with data-intensive AI or HPC workloads, both available in 2027.

Why this matters

HPE's positioning of CPU-focused servers as orchestration nodes for agentic AI workflows, handling tool calls and retrieval while GPU systems manage inference, reflects a growing view that agentic AI architectures require substantial CPU resources alongside GPU clusters, not just raw GPU capacity. The explicit targeting of neoclouds and large enterprises with high-density CPU systems signals that the enterprise AI hardware market is diversifying beyond GPU-centric configurations.

Why the Digest selected this story

HPE's new Gen13 server generation reflecting a rebalancing of CPU and GPU resources signals an important architectural shift in enterprise AI infrastructure. New hardware generations from major OEMs set industry direction for data center buildouts.

Read the full story at Data Center Knowledge →
Data Center Knowledge · 6 hours ago
Impact Digest Original

The Afterlife of a Data Center: Resold GPUs, Mounting E-Waste, and Who Pays for the Cleanup

A Data Center Digest Original StoryThis article was researched and written entirely by AI, without human review or editing, as part of the Data Center Digest's ongoing experiment in AI-powered journalism.

The equipment inside a data center has a shorter working life than the building around it, and the companies that own it do not agree on how much shorter. Meta extended the assumed life of most of its servers and network equipment to 5.5 years in January 2025, up from a previous range of four to five years, a change its annual report says cut 2025 depreciation expense by $2.92 billion and raised net income by $2.59 billion. Amazon moved part of its fleet the other way: effective January 2025, a subset of its servers and networking equipment went from a six-year life to five, which its annual report attributes to the pace of technology development in artificial intelligence and which added $1.4 billion in depreciation expense. Both figures are accounting estimates, not measurements, and they decide how much of the cost of AI hardware shows up in earnings each year.

The disagreement is sharpest for GPUs. Critics, including investor Michael Burry, argue that AI chips should be depreciated over less than three years, and that their true economic life may be one to three, a view Yardeni Research summarizes alongside the point that Nvidia has moved to a one-year product cycle. Defenders, including Yardeni, say older GPUs keep earning as they shift from training the largest models to less demanding inference and other work, so their useful lives run longer than the critics assume. The answer determines when hardware leaves a data center, and where it goes next.

Part of that answer is a resale market. According to price ranges published in May 2026 by Caladan Semiconductor, a used-GPU dealer, used Nvidia H100 SXM5 cards were trading at roughly $18,000 to $24,000 against $25,000 to $32,000 new, or about 72 to 75 percent of the new price. Used 80GB A100s fetched about 55 to 65 percent of new, down from more than 80 percent in 2022, according to the same source. Caladan reported that most used H100s were coming from failed startups and distressed sales rather than enterprise upgrades, while A100 supply was growing as companies that bought clusters in 2021 and 2022 moved to newer chips. The figures come from a dealer, and the article does not say whether they are asking or completed-sale prices, so they are a rough guide and not audited market data. Companies such as Reconext, a remarketer, now sell processes that wipe drives and test GPUs in place so racks can be resold without being pulled apart.

What does not get resold becomes electronic waste, a category already growing faster than the world's ability to handle it. The UN's Global E-waste Monitor 2024 counted a record 62 million tonnes generated in 2022, up 82 percent since 2010 and on track for about 82 million tonnes by 2030. Only 22.3 percent was documented as properly collected and recycled, and the report found generation rising five times faster than documented recycling. Researchers writing in Nature Computational Science in 2024 estimated that servers built for generative AI alone could add up to 5 million tons of e-waste cumulatively through 2030, and as much as 2.5 million tons a year by 2030 if nothing is done to reduce it, about 3 percent of the UN's projected 2030 global total. They also estimated that reusing and remanufacturing components, extending equipment life, and better recycling could cut that output by up to 86 percent.

The building itself is a separate problem. Ashwin Murthy of Columbia Law School's Sabin Center for Climate Change Law wrote in August that older abandoned data centers often sat vacant for long stretches, kept up by a skeleton crew while owners looked for buyers, and that modern hyperscale facilities are harder to repurpose because they can span millions of square feet and are specialized by design. Idle sites, he warned, risk becoming abandoned warehouses that waste land and leave hazardous material behind. There is also a cost that can fall on others: because data centers require heavy grid upgrades, he wrote, ratepayers may end up paying for infrastructure built for a company that fails and leaves a facility idle.

Some local governments are writing cleanup rules in advance. Murthy points to Susquehanna County, Pennsylvania, which requires decommissioning within twelve months after a data center's useful life ends, including removal of all hazardous materials and contents, backed by bonding, and notes that Smithfield Township has a draft ordinance. He argues that states could go further by requiring detailed decommissioning plans before permitting that bind any future owner, and bonds or similar financial assurances that cover the full cost, modeled on oil and gas bonding rules. Whether a locality can impose such rules varies by state, and he notes that rules aimed only at data centers could face equal-protection challenges, though he expects those to fail.

For now, how long a data center lasts depends on assumptions in annual reports, a resale market priced by dealers, and a patchwork of local ordinances. Who ends up paying when facilities are retired, whether shareholders, customers, or the town that hosted them, will depend on which of those holds up.

A Data Center Digest Original Story
Policy

Massena's 635-Megawatt Data Center May Skirt New York State Moratorium

New York Digital Investment Group and its affiliate North Country Collocation Services want to build a 1.5 million-square-foot, 635-megawatt data center on the former Alcoa East/Reynolds industrial site near the St. Lawrence River in Massena, a project that would draw roughly 70% of the output from the nearby Moses Saunders hydroelectric dam. Gov. Kathy Hochul's executive order restricts new hyperscale data centers exceeding 50 megawatts by barring the DEC from issuing new discretionary permits, but NCCS argues it applied for all necessary permits back in 2020 and requires no additional state approvals. Massena's Town Board voted 3-2 against a local moratorium, though the project still faces a SEQRA review and DEC environmental assessment. The St. Regis Mohawk Tribe formally opposed the project, while some Massena residents support it based on the company's promise of 200 full-time jobs, roughly 2,000 construction jobs, and millions in community investments.

Why this matters

The Massena case tests whether permit grandfathering can exempt a 635-megawatt facility from a statewide moratorium, a precedent that could allow other large projects to bypass similar restrictions simply by having applied for permits before such orders take effect. At a scale representing 70% of a major hydroelectric dam's output, the outcome will shape how New York and other states structure future moratorium language to close such loopholes.

Why the Digest selected this story

New York state action to pause large data center approvals is significant policy news, and the exemption angle for a massive Massena project adds a distinct local dimension not covered in prior published items about Texas or Chicago moratoriums.

Read the full story at North Country Public Radio →
North Country Public Radio · 12 hours ago
Power

Seven States Enact Ratepayer Protections as Data Center Power Costs Surge

Maryland, South Dakota, Alabama, Minnesota, Texas, Virginia, and Nebraska have each enacted distinct measures requiring data centers to bear more of the grid infrastructure costs their facilities create, rather than passing those costs to residential and business ratepayers. Alabama's law, effective October 1, 2026, covers data centers with electricity demands of at least 150 megawatts; Minnesota imposes annual fees ranging from $2 million for facilities drawing 100 to 250 megawatts up to $5 million for those requiring at least 750 megawatts. Maryland's framework, established under an executive order signed by Gov. Wes Moore in September 2026, requires state review of projects needing at least 25 megawatts and mandates a public dashboard disclosing electricity demand and water use. Virginia's State Corporation Commission issued an order in August requiring data centers to pay for dedicated transmission infrastructure following a request from Gov. Abigail Spanberger's administration.

Why this matters

The simultaneous passage of cost-allocation laws in seven states marks a measurable shift in how legislatures are treating large electricity customers, moving away from socializing grid upgrade costs across all ratepayers. The range of mechanisms adopted, from direct cost reimbursement requirements to annual fees and curtailment obligations, will serve as models or cautionary examples for the remaining states that have not yet acted.

Why the Digest selected this story

Newsweek coverage of multiple states taking legislative or regulatory action to protect ratepayers from data center electricity cost burden is highly newsworthy given ongoing national debate; ratepayer cost protection is a distinct angle not covered in already-published items.

Read the full story at Newsweek →
Newsweek · 8 hours ago
Opposition

Memphis Council Delays Data Center Moratorium Vote, Protests Erupt in City Hall

Memphis City Council voted 7-5 to postpone a decision on a proposed 12-month moratorium on new data center projects, triggering protests inside City Hall, a police clearing of the chamber, and a fight between two women in the lobby. The next vote is scheduled for October 20, according to the report; no one was detained or injured in the lobby altercation, a Memphis Police Department spokesperson said. Councilman J.B. Smiley Jr., the moratorium's lead sponsor, pushed for an immediate vote, while Councilman Jeff Warren warned that a pause could drive companies to neighboring DeSoto, Crittenden, Fayette, or Tipton counties. Elon Musk's SpaceXAI operates its Colossus supercomputer facilities in the Memphis area, and Mayor Paul Young has called for new rules on water and electricity use, air quality, noise, and siting while noting the tax revenue data centers generate.

Why this matters

The Memphis dispute shows how the presence of a single major operator, SpaceXAI's Colossus facility, can transform a local zoning question into a flashpoint severe enough to produce physical confrontations at a city council meeting. A moratorium covering one of the South's larger cities would affect a region that competing jurisdictions in multiple neighboring counties are openly positioning to absorb, illustrating the economic and regulatory competition shaping data center location decisions.

Why the Digest selected this story

A city council devolving into chaos over a data center moratorium vote is a vivid, high-profile community opposition moment; Memphis has not appeared in already-published items and the political disruption signals escalating local tensions.

Read the full story at Fox News →
Fox News · 6 hours ago
Power

Demand Response Could Cut Data Center Peak Loads Up to 30 Percent

Technology companies and utilities are expanding demand response programs at AI data centers, allowing facilities to temporarily reduce or shift electricity use during periods of grid stress. The Electric Power Research Institute projects U.S. data center electricity consumption could rise from 177 to 192 terawatt-hours in 2024 to between 383 and 793 terawatt-hours by 2030, and EPRI technical leader Arin Kaye said surveyed facilities reported peak power reduction potential of 10% to 30%, with some hyperscalers going higher. A Duke University Nicholas Institute study estimated that greater data center flexibility could avoid $40 billion to $150 billion in capital investments over the next decade. OpenAI recently agreed to cut electricity draws by up to 1 gigawatt from a planned 3.2-gigawatt Georgia facility during grid stress events, and Alphabet's Google, NVIDIA, and Emerald AI launched the AI Energy Management Alliance to advance flexible data center deployment.

Why this matters

The $40 billion to $150 billion in potential avoided grid investment is large enough to reshape how utilities and regulators approach interconnection timelines and infrastructure planning for the next decade. Federal regulators ordering grid operators in June to consider faster connection pathways for demand-responsive facilities means the policy framework is now moving in parallel with industry pilots, accelerating the likelihood that curtailment agreements become a standard condition for new large-load interconnections.

Why the Digest selected this story

Finance & Commerce coverage of data centers adopting flexible or demand-response power strategies to alleviate grid pressure is timely and distinct from the already-published $150 billion demand-response savings study; this appears to focus on operational approaches by operators rather than a single study's projections.

Read the full story at Finance & Commerce →
Finance & Commerce · 10 hours ago
Opposition

PR Firms Race to Defuse Anti-Data Center Movements Blocking Billions in Projects

During the first three months of 2026, data center opponents stopped or delayed 75 projects worth almost $130 billion, roughly equal to those blocked in all of 2025, according to figures cited at an industry panel convened by Data Center Frontier in Northern Virginia in early August. A Gallup survey conducted in March found that 71% of Americans would oppose a new AI data center in their area, a statistic industry executives described as their biggest barrier to growth. In response, a growing number of PR consultants are marketing specialized services including community outreach, landowner recruitment, and preemptive relationship-building before opposition groups can form. David McCall, described as a brand evangelist at QTS Data Centers, told the panel that community opposition is a card that can stop every project: 'It can stop and will stop everything.'

Why this matters

The pace at which opposition is accelerating,$130 billion in blocked or delayed projects in a single quarter matching the full prior year's total, signals that community resistance has become a primary constraint on data center supply, not capital or land availability. The emergence of a specialized PR industry targeting local permitting processes reflects how consequential municipal and county-level decisions have become for the national AI infrastructure buildout.

Why the Digest selected this story

A Slate feature on industry efforts to counter widespread public opposition to data centers — framed as 'the most evil innovation in America' — is notable for capturing the national cultural and PR battle over data center expansion, a perspective distinct from local opposition stories.

Read the full story at Slate →
Slate · 9 hours ago