--- 格式版本: 2 标题: "Texas AI Data Centers: Power, Policy, and Progress" 原文链接: "https://www.datacenterknowledge.com/regulations/texas-ai-data-centers-coverage-hub-power-policy-and-progress" 发布日期: "2026-06-25" 发布时间校准状态: "found" 发布时间需复核: "否" 发布时间来源: "llm:scrape:provider_published_at" 发布时间证据: "provider publishedAt: 2026-06-25" 发布时间校准原因: "优先级最高的候选c1为Firecrawl/provider publishedAt，且与datePublished（c2、c3）一致，均为2026-06-25，符合文章发布时间特征。c4、c5为'L…" 发布时间校准置信度: "1" 发布时间候选数量: 5 发布时间严格候选数量: 2 发布时间原页读取状态: "原页面已读取" 发布时间未找到原因: "" 发布时间校准时间: "2026-07-26T01:52:15+08:00" 发布时间仲裁状态: "confirmed" 发布时间仲裁尝试次数: 1 发布时间仲裁耗时毫秒: 4805 发现时间: "2026-07-26T01:44:03+08:00" 入库时间: "2026-07-25T17:52:20.858Z" 来源平台: "固定入口" 搜索渠道: "fixed_url" 搜索词: "https://www.datacenterknowledge.com/latest-news" 匹配关键词: - "roadmap" - "deployment" - "performance" - "bandwidth" 相关厂家: - "NVIDIA" - "Meta" - "Microsoft" - "Google" 相关专家: [] 内容类型: "网页" 抓取工具: "Free Fetch + Defuddle" 清洗工具: "Defuddle Markdown + Defuddle/Readability 正文提取" 原始附件: [] AI优质: "否" AI打分: 43 AI分档: "非优质" AI质检状态: "不通过" AI打分理由: "内容为德州AI数据中心电力政策与项目进展的综合报道，未涉及超节点/AI Rack硬件架构、技术细节或关键部件，主题相关性弱。" AI质检模型: "deepseek-v4-flash" AI质检时间: "2026-07-27T11:15:27+08:00" AI主题相关性: 5 AI来源权威性: 12 AI新颖性: 12 AI技术细节: 0 AI商业部署信号: 8 AI完整性: 6 采集批次: "2026年7月25日22点32分16秒" 采集批次ID: "20260725-223216-273" 去重键: "https://www.datacenterknowledge.com/regulations/texas-ai-data-centers-coverage-hub-power-policy-and-progress" --- The future of AI infrastructure is taking shape in Texas, where policy reform, power-first strategies, and transmission constraints are determining which gigawatt-scale campuses move from announcement to actual operation. In Texas, the projects moving fastest are those that brought their own power to the table.Getty Images / Data Center Knowledge This is a continuously updated hub for Data Center Knowledge reporting on Texas and the Electricity Reliability Council of Texas (ERCOT), tracking how power-first strategies, regulatory changes, and transmission expansion are shaping AI-scale infrastructure. Last updated: July 6, 2026. --- Texas has emerged as the testbed for AI-scale data centers – large campuses built to train and run AI systems that draw round-the-clock, high-magnitude electricity. At the center is ERCOT, which operates most of the state’s grid and wholesale power market. To keep pace, ERCOT is changing how projects are queued for grid connections. In the past, [speculative interconnection queuing](https://www.datacenterknowledge.com/energy-power-supply/the-ai-demand-dilemma-utilities-confront-speculative-growth) allowed projects to hold spots with little evidence they would actually build, creating long backlogs. ERCOT is now moving to evidence-based queue management with readiness tests – such as site control, [permits in progress](https://www.datacenterknowledge.com/regulations/data-center-permits-how-long-they-take-and-what-speeds-approval), financing plans, equipment orders, and realistic power-use forecasts – to separate committed loads (projects with a high likelihood of being built and energized) from aspirational requests. The goal is to focus limited grid-planning resources on buildable projects. ERCOT warns that rapid AI demand will only land on schedule if supply expands in step. Both generation (power plants and storage that supply electricity – natural gas, wind, solar, batteries) and [transmission](https://www.datacenterknowledge.com/build-design/utilities-may-get-an-ai-boom-the-grid-wasn-t-built-for) (the high-voltage lines that move power long distances) must grow. These dependencies will influence when sites receive power for the first time, what they pay for it, and whether projects ultimately come online. In parallel, developers are adopting a power-first model. They are locking in long-term power contracts earlier and, in some cases, pairing data center sites with on-site or nearby generation to [shorten time-to-power](https://www.datacenterknowledge.com/energy-power-supply/speed-to-power-how-developers-are-restructuring-for-ai-demand) and reduce risk. Others are exploring off-grid and [behind-the-meter options](https://www.datacenterknowledge.com/energy-power-supply/why-data-centers-produce-their-own-power) to gain more control over timelines and cost, though those approaches carry their own regulatory and financing hurdles. Across the stories below, readers will find how Texas policy is evolving, what tactics projects are using to get power sooner, how companies are buying electricity differently, and how Texas stacks up against other markets. The aim is to show how the Lone Star State is turning plans into delivered megawatts and where execution risks remain. ## ERCOT Policy, Rules, and Market Signals [Texas Tests New Rules for AI Campuses Behind Existing Power Plants](https://www.datacenterknowledge.com/energy-power-supply/texas-tests-new-rules-for-ai-campuses-behind-existing-power-plants) (July 2026) – Texas regulators are deciding whether a 260 MW AI data center expansion sharing a wind farm's grid connection must fully curtail during emergencies, testing whether curtailment obligations should apply to total behind-the-meter load or just enough to restore the generator's capacity to the grid. [Texas Approves Batch Zero Study as Data Center Demand Soars](https://www.datacenterknowledge.com/build-design/texas-creates-a-credibility-test-for-gigawatt-scale-data-center-demand) (June 2026) – The Public Utility Commission of Texas (PUCT) approved ERCOT’s one-time “Batch Zero,” a statewide study to evaluate very large new power users, such as AI data centers, in a consistent way. Instead of each utility running its own review, ERCOT will run a centralized process that prioritizes projects with evidence they will actually be built over early-stage or speculative requests. [Gridlock or Growth? ERCOT Warns Texas AI Power Boom May Not Materialize](https://www.datacenterknowledge.com/energy-power-supply/gridlock-or-growth-ercot-warns-texas-ai-power-boom-may-not-materialize) (May 2026) – ERCOT now projects that electricity demand could approach 368 GW by 2032 as AI and data centers scale. Some developers are pushing ahead by adding on-site generation to avoid delays, while ERCOT works to separate committed projects from those that are unlikely to move forward. [Texas Gets Tough on Data Center Power – Who’s Next?](https://www.datacenterknowledge.com/energy-power-supply/texas-gets-tough-on-data-center-power-who-s-next-) (January 2026) – Texas’s Senate Bill 6, enacted in June 2025, shifts more grid-connection costs and reliability responsibilities onto very large ERCOT customers (roughly 75 MW and up). The policy aims to prevent household ratepayers from subsidizing large data center loads and to ensure these facilities can reduce demand or be curtailed in emergencies. It’s part of a broader national trend toward “pay your own way” as AI growth strains grid planning and capacity. [Comment: Texas’ New Large-Load Law Is an Opportunity for Data Centers](https://www.datacenterknowledge.com/energy-power-supply/texas-new-large-load-law-is-an-opportunity-for-data-centers) (June 2025) – In this opinion piece, CPower Energy Senior Regulatory Affairs Analyst Claire Swingle argues that higher costs and new curtailment rules under SB-6 can be turned to data centers' advantage. By using demand response (being paid to lower usage during grid stress), adding on-site power, and choosing the right utility tariffs, data centers can manage costs while supporting grid reliability. ## Building the Grid for AI: Technology and Transmission [Texas’ 765 kV Decision: Build the Wires, the AI Will Follow](https://www.datacenterknowledge.com/energy-power-supply/how-ercot-s-post-crez-bet-is-reshaping-ai-infrastructure) (July 2026) – Texas approved a $13.8 billion, 765 kV transmission network to accommodate AI data centers and large electrical loads, shifting from reactive planning to building infrastructure before knowing where future demand will emerge. [DOE Is Quietly Building the AI Grid Operating System](https://www.datacenterknowledge.com/energy-power-supply/doe-is-quietly-building-the-ai-grid-operating-system) (June 2026) – The US Department of Energy’s “Agora” platform models how hyperscale AI data centers pull power, often in fast-changing bursts, and how they should behave during grid events. It helps validate scenarios like [ride-throughs](https://www.datacenterknowledge.com/uptime/voltage-ride-through-a-key-ingredient-in-data-center-resilience) (staying online during brief disturbances), reconnection (coming back in a controlled way after an outage), and controllable load (adjusting power use on command). The insights inform ERCOT’s and Texas A&M’s work on interconnection and operating requirements tailored to large electronic loads. [Texas May Have Accidentally Built the Perfect Grid for AI](https://www.datacenterknowledge.com/build-design/texas-may-have-accidentally-built-the-perfect-grid-for-ai) (May 2026) – Texas’s 345-kilovolt (kV) CREZ transmission corridors, originally built to move West Texas wind power, are now becoming a critical enabler for AI data center development. However, while ERCOT’s large-load queue remains crowded and key equipment is back-ordered, a substantial gap exists between announced gigawatt-scale projects and actual physical deployment, forcing regulators and utilities to distinguish viable developments from speculative proposals. ## Texas vs. the Competition: Market Position and Momentum [Texas Powers Past Virginia in Global Data Center Rankings](https://www.datacenterknowledge.com/build-design/texas-powers-past-virginia-in-global-data-center-rankings) (May 2026) – Texas’s edge – [abundant power](https://www.datacenterknowledge.com/data-center-site-selection/hyperscale-growth-shifts-inland-as-ai-drives-power-demand), ample land, and fewer siting constraints – is enabling multi-gigawatt data center campuses to move faster than in congested legacy markets. This momentum is challenging Northern Virginia’s long-standing lead. ## Major Projects: Who’s Building What and How They’re Getting Power [Behind-the-Meter AI: Microsoft-Chevron’s West Texas Bet](https://www.datacenterknowledge.com/build-design/chevron-lands-20-year-microsoft-deal-to-power-west-texas-ai-campus) (June 2026) – Near Pecos, Project Kilby pairs a Microsoft-run AI campus with roughly 2.67 GW of co-located natural-gas generation developed by Chevron’s Energy Forge One and Engine No. 1. The plan targets first power in 2028 and builds toward about 2 GW of data center capacity over five to seven years. In effect, Microsoft and its partners are bringing their own power to the site to [speed up and de-risk buildout](https://www.datacenterknowledge.com/data-center-construction/building-data-centers-faster-plays-that-de-risk-delays). [Google Bets on a Power-First Model for AI Data Centers](https://www.datacenterknowledge.com/energy-power-supply/google-bets-on-a-power-first-model-for-ai-data-centers) (June 2026) – Google and energy developer Intersect are pursuing a Texas campus with approximately 1 GW of dedicated power: mostly wind, solar, and batteries, supplemented by on-site gas to ensure an always-available, lower-carbon supply. [Unconventional Texas Data Center Explores Off-Grid Power](https://www.datacenterknowledge.com/build-design/unconventional-texas-data-center-explores-off-grid-power) (May 2026) – After facing roughly $35 million in interconnection costs and potential delays to 2029, industrial developer BaRupOn is pivoting a 700-acre Liberty County site to an off-grid, behind-the-meter design. Options under review include gas-fired generation and [modular microreactors](https://www.datacenterknowledge.com/energy-power-supply/how-realistic-is-nuclear-power-for-ai-data-centers) to shorten time-to-power and control costs. [Nvidia Places Massive AI Infrastructure Bet on IREN’s 5 GW Pipeline](https://www.datacenterknowledge.com/deals/nvidia-places-massive-ai-infrastructure-bet-on-iren-s-5-gw-pipeline) (May 2026) – Nvidia and IREN partnered to make IREN’s 2 GW Sweetwater campus a flagship [AI factory](https://www.datacenterknowledge.com/data-center-chips/nvidia-says-vera-rubin-vera-cpu-on-track-launches-dsx-os-to-run-ai-factories), with a plan to deploy up to 5 GW of Nvidia-optimized infrastructure across IREN’s global pipeline. [Exowatt Expands to Austin as Power Constraints Reshape AI Infrastructure](https://www.datacenterknowledge.com/build-design/exowatt-expands-to-austin-as-power-constraints-reshape-ai-infrastructure) (March 2026) – Exowatt is establishing an Austin hub to deploy modular renewable power systems that deliver energy directly to data centers behind the meter. The model aims to bypass grid bottlenecks and support high-density AI computing. [Crusoe and Microsoft Expand Abilene AI Campus with New 900 MW ‘AI Factory’](https://www.datacenterknowledge.com/build-design/crusoe-expands-abilene-ai-campus-with-new-900mw-ai-factory-for-microsoft) (March 2026) – Crusoe is adding a 900 MW AI data center to its Abilene campus for Microsoft, bringing the campus’s total planned capacity to 2.1 GW. The project exemplifies a fundamental shift in hyperscale infrastructure toward energy-first site selection, with on-site power generation becoming essential to bypass grid constraints. [Logix Ramps Up Fiber Networks as Texas Data Centers Boom](https://www.datacenterknowledge.com/networking/logix-ramps-up-fiber-networks-as-texas-data-centers-boom) (February 2026) – Logix Fiber Networks is building high-capacity routes in South Dallas and the Austin–Bastrop corridor, linking new growth zones to Dallas-Fort Worth carrier hubs. More fiber means more bandwidth and path diversity, improving performance and resilience for large AI campuses. [NRC Intervention Tests the Data Center Case for SMRs in Texas](https://www.datacenterknowledge.com/energy-power-supply/nrc-intervention-tests-the-data-center-case-for-smrs-in-texas) (February 2026) – Dow and X‑energy’s proposal for four Xe‑100 small modular reactors at the Long Mott Generating Station in Calhoun County hit a regulatory speed bump. The Nuclear Regulatory Commission’s Atomic Safety and Licensing Board accepted a challenge to the project’s financial qualifications, highlighting financing and timeline risks that nuclear-backed strategies face. [Soluna Expands Texas Campus With 100 MW AI-Ready Data Center](https://www.datacenterknowledge.com/data-center-construction/soluna-expands-texas-campus-with-100-mw-ai-ready-data-center) (January 2026) – Soluna and Metrobloks are developing Project Kati 2 in Willacy County, an AI/high-performance computing data center next to Soluna’s wind-powered campus. The plan starts just above 100 MW with a roadmap to exceed 300 MW as the buildout proceeds. [GridFree Unveils First ‘Power Foundry’ Site for AI Data Center Workloads](https://www.datacenterknowledge.com/energy-power-supply/gridfree-unveils-first-power-foundry-site-for-ai-data-center-workloads) (December 2025) – GridFree AI launched South Dallas One, the first site in a planned 5 GW cluster of gas-powered, grid-independent campuses. The concept combines dedicated on-site power and industrial-scale cooling to support rapid deployment without drawing on public grid capacity. [Meta Building New Gigawatt-Sized Data Center in Texas](https://www.datacenterknowledge.com/data-center-construction/meta-building-new-gigawatt-sized-data-center-in-texas) (October 2025) – Meta plans to invest more than $1.5 billion in a 1 GW data center in El Paso, targeting service in 2028. The project is part of Meta’s broader push to expand AI infrastructure at hyperscale. ## What Industry Leaders Are Saying About Texas [Texas Data Center Potential Unveiled at Industry Power Forum](https://www.datacenterknowledge.com/energy-power-supply/texas-data-center-potential-unveiled-at-industry-power-forum) (October 2025) – At the inaugural Data Center World Power event in San Antonio, speakers highlighted Texas as a leading destination for AI-era data centers, citing available power, supportive infrastructure, and fast-moving policy. ## About the Authors Senior Editor, Informa TechTarget Spencer Smith is senior editor at Data Center Knowledge, the leading source of news and in-depth analysis for the data center industry. Prior to joining Data Center Knowledge in 2025, he served as senior editor at ITPro Today and was a site editor for TechTarget’s SearchITChannel.com. You can reach him at [\[email protected\]](https://www.datacenterknowledge.com/cdn-cgi/l/email-protection) and on LinkedIn.