CoreWeave

Recent news

• Jun 8, 2026

  The AI Demand Dilemma: Utilities Confront Speculative Growth

  Utilities across the US are rewriting tariffs, demanding financial guarantees, and altering transmission and procurement plans to avoid building infrastructure for speculative AI-related data center load requests.

  • Main action: Utilities (notably AEP and Duke Energy) are tightening large-load rules and requiring financial commitments to move projects forward: AEP winnowed more than 30 GW of preliminary requests to ~13 GW for formal studies and 5.6 GW with signed Electric Service Agreements; AEP proposed requiring customers to commit to paying for 90% of requested capacity for a decade before the utility builds supporting infrastructure. These measures include specialized large-load tariffs, collateral/minimum-usage guarantees, and phased energization schedules to limit ratepayer exposure.
  • Background and implementation details: Regulators and reliability bodies (NERC, ERCOT, FERC) are developing new categories and study frameworks (e.g., Computational Load Entity, batch study processes) and reliability guidance. Utilities are expanding financing and procurement: Duke extended a $10 billion master credit facility through 2031 and raised its five-year capital plan to $103 billion; AEP raised its five-year capital plan to $78 billion. Industry forecasts and planning estimates include Wood Mackenzie projecting the US data center electrical equipment market could grow to $65 billion by 2030, and Grid Strategies/ACEG estimating roughly 5,000 miles of new high-capacity transmission annually through 2035 (fewer than 1,000 miles built in 2024).
• Jun 1, 2026

  The Case For Pragmatism in the AI Infrastructure Boom

  The article argues that the AI boom is driving substantial data centre investment while creating financing, power, permitting and skills challenges for European data infrastructure.

  • Main point: The piece synthesises findings from reports (McKinsey, SoEDC 2026, Moody’s, Gartner) that AI workloads require far larger CapEx (McKinsey: $5.2 trillion for AI vs $1.5 trillion for traditional IT) and that hyperscale construction investment in Europe is expected at €7 billion (~$8.1 billion) per year to 2031; it emphasises power demands (1–5 GW campuses, double rack densities, expansion from 7.6 GW colocation IT power in 2024 with 27% CAGR to 2031) as central constraints.
  • Context and sourcing: This is an analytical commentary referencing multiple published reports and news items (SoEDC 2026, McKinsey, Moody’s, Gartner, ECB statements) rather than a single new corporate announcement; it highlights modular/phased builds, neocloud providers (CoreWeave, Global AI, Nebius), and regional power projections (southern Europe: 682 MW in 2024 to ~5.9 GW by 2031) as concrete implementation details.
• May 27, 2026

  Texas May Have Accidentally Built the Perfect Grid for AI

  ERCOT and industry filings show CREZ-era West Texas transmission corridors are increasingly guiding hyperscale AI campus siting as developers seek bulk power and expandable transmission capacity.

  • Main announcement/action: Galaxy Digital’s Helios campus (previously an Argo Blockchain site) has ERCOT approval for an additional 830 MW, bringing total approved load above 1.6 GW after completion of Large Load Interconnection Studies (LLIS); the project has service agreements with AEP Texas and transmission coordination with WETT, but current physical deployment is phased (initial lease: 133 MW to CoreWeave).
  • Background and details: Filing documents indicate a 327.2 MW self-generator registration (backup power) including 121 Caterpillar diesel generators tied to roughly 252 MW of emergency generation, a currently identified 200 MW co-located load, and broader system context where ERCOT’s large-load queue has ballooned above ~230 GW and industry reports cite transformer shortages (~30%), multi-year lead times, and average global power-delivery timelines of ~4.4 years.
• May 19, 2026

  Does Google’s $5B TPU Deal Signal a New Neocloud Era?

  Blackstone and Google have announced a new US-based AI infrastructure venture built around Google’s Tensor Processing Units (TPUs).

  • Main announcement: The venture will provide data center capacity, networking, operations, and Google Cloud TPUs as a compute-as-a-service offering; Blackstone committed $5 billion in initial equity, and the partners plan to bring 500 MW of capacity online in 2027. Benjamin Treynor Sloss (longtime Google infrastructure executive) will lead the company as CEO.
  • Background and details: The deal structure separates TPU infrastructure from the traditional hyperscale cloud model and echoes the neocloud trend (players like CoreWeave, Crusoe, Lambda). Reuters reported the broader venture could involve up to $25 billion including leveraged financing. Locations and power sourcing for facilities were not disclosed.
• May 15, 2026

  Earnings Roundup: Neoclouds Shift From GPU Race to Power Wars

  CoreWeave and Nebius announced Q1 earnings showing rapid revenue growth and massive capex increases to industrialize AI infrastructure.

  • Main announcement:CoreWeave reported Q1 revenue of $2.08 billion, adjusted EBITDA $1.16 billion, crossed 1 GW of active power, expanded contracted power to >3.5 GW, posted $6.8 billion Q1 capex and disclosed a $21 billion Meta commitment; Nebius reported Q1 revenue $399 million (AI revenue $390 million), adjusted EBITDA $129.5 million, Q1 capex $2.5 billion, and raised FY capex guidance to $20–$25 billion, planning a Pennsylvania AI campus up to 1.2 GW and >4 GW contracted power by year-end.
  • Background and implementation details:Both firms raised multi-year capex plans (CoreWeave flagged up to $35 billion 2026 capex low end), shifting competition from GPU procurement to power, cooling, networking, and deployment velocity; the article cites utility interconnection queues, transformer shortages, transmission access, and substation construction as concrete execution bottlenecks.
• May 14, 2026

  Land and Expand: NVIDIA, IREN, Coatue, Microsoft, Switch, Cerebras, Core Scientific

  NVIDIA announced two major partnerships to accelerate industrial-scale AI infrastructure deployment with IREN and Corning Incorporated.

  • Main announcement: NVIDIA partnered with IREN to target deployment of up to 5 gigawatts of NVIDIA DSX-aligned AI infrastructure (focus on IREN’s 2-gigawatt Sweetwater campus in Texas) and separately partnered with Corning Incorporated to expand U.S. optical connectivity manufacturing (10x optical connectivity capacity increase; >50% domestic fiber production increase; construction of three new advanced manufacturing facilities in North Carolina and Texas). The IREN deal includes a five-year right for IREN to sell NVIDIA up to 30 million ordinary shares at $70 per share (potential consideration up to $2.1 billion).
  • Background and details: The article details additional industry moves into powered land, gigawatt campuses, crypto-to-AI conversions, and domestic supply-chain expansion, including Coatue/Next Frontier & Fluidstack’s 430 MW Indiana campus backed by $5.7 billion in senior secured notes (first 65 MW online by July 2027), Digi Power X’s 10-year MSA with Cerebras for a 40 MW Columbiana, AL campus (initial contract ~$1.1 billion, potential $2.5 billion, Phase 1 ready-for-service targeted Dec. 15, 2026), CloudBurst’s Texas campus ($14.5 billion investment; 1.2 GW planned), and Core Scientific’s acquisitions and campus expansions (e.g., $421 million cash acquisition of Polaris DS LLC; Muskogee and Pecos expansions to ~1.5 GW gross power).
• May 14, 2026

  Gridlock or Growth? ERCOT Warns Texas AI Power Boom May Not Materialize

  The Electric Reliability Council of Texas (ERCOT) filed a long-term load forecast to the Public Utility Commission of Texas projecting very large future demand but warning much of it may not materialize on schedule.

  • Main announcement: ERCOT submitted an April 2026 long-term load forecast projecting statewide peak demand could reach nearly 368 GW by 2032 (compared with the current record 85.5 GW), while explicitly stating concerns with using preliminary load forecast values and reserving the right to adjust forecasts based on “actual historical realization rates or other independent information.”
  • Background and detail: The filing reports ERCOT applied an internal realization factor (average peak consumption per site = 49.8% of requested MW) to some non-crypto data center additions, projects non-crypto data center load of ~228 GW by 2032, and expects summer 2026 peak between ~90.5 GW and 98 GW (versus a preliminary 112 GW figure). The filing is a regulatory submission (filing) and functions as a capacity/planning signal rather than a confirmation that proposed projects will be energized on forecast timelines.
• Apr 30, 2026

  Microsoft AI Surge Exposes Data Center Capacity Gap

  Microsoft reported quarterly results and highlighted a widening gap between AI-driven cloud bookings and physical delivery of capacity.

  • Main announcement: Microsoft reported Azure revenue growth of 40% YoY, the AI business at a $37 billion annual revenue run rate (up 123%), and a commercial RPO of $627 billion (up 99%), indicating significant contracted but undelivered cloud services and demand outstripping current delivery capacity.
  • Background/details: The release is an earnings report for the Intelligent Cloud segment ($34.7 billion quarterly, up 30%) and company-wide revenue ($82.9 billion, up 18%) and net income ($31.8 billion, up 23%). Independent analysis from Steven Dickens at HyperFrame Research cites delivery windows stretching from ~6 months to 18+ months, constraints across power, cooling (liquid cooling), transformer/local utility capacity, and skills, and mentions specialized GPU providers like CoreWeave acting as an “overflow valve”. Microsoft did not disclose capital expenditure or expansion timelines in the release.
• Apr 17, 2026

  AI Infrastructure Brief: Power, Capital, and the Feeling That Something Is Tightening

  Matt Vincent (Data Center Frontier) summarized the week’s announcements showing an accelerating AI data-center buildout paired with mounting power and coordination constraints.

  • Main observation: The industry is prioritizing power and speed: major deals and project announcements include Bloom Energy and Oracle planning up to 2.8 GW of deployment, Aligned Data Centers breaking ground on a 540 MW Project Caprock, an EdgeConneX affiliate proposing a 430 MW natural gas plant in New Albany, Ohio, proposals for 2 GW in New Mexico and 1.2 GW in Irwin County, Georgia, and Microsoft expanding datacenter operations in Cheyenne. The Maine legislature passed a temporary, exemption-inclusive ban on data centers, signaling emerging social-license constraints.
  • Capital and implementation details: Financial moves include Switch raising $768 million via ABS, Fluidstack reported in talks for a $1 billion round at an $18 billion valuation, and Jane Street signing a $6 billion AI cloud agreement with CoreWeave; CoreWeave also expanded a multi-year relationship with Anthropic. Utilities are signing long-term power agreements (e.g., NiSource with Alphabet and expanded ties with Amazon). AWS has launched “Project Houdini” to accelerate construction timelines. All items are factual recaps of announcements and reports from the week (no speculative outcomes included).
• Apr 14, 2026

  Data centers are moving inland, away from some traditional locations

  Synergy Research Group and Sightline Climate reported a geographic shift and widescale delays in U.S. data center construction.

  • Main announcement:Synergy Research Group finds the planning and build “center of gravity” for data centers is moving inland to Texas and Midwestern states (Wisconsin, Indiana, Michigan, Missouri). Sightline Climate reports 16 GW of data centers slated to open in the U.S. this year but only 5 GW are under construction now and expects 30–50% of projects to be delayed; 25 GW are announced for 2027 with only 6 GW under construction.
  • Background and details: Delays are driven by component shortages (memory, storage, batteries, electrical transformers, circuit breakers) and local opposition (e.g., the Seminole Nation banning data centers on tribal lands). Major cloud and AI firms named as project sponsors include Amazon, Google, Meta, Microsoft, OpenAI, and CoreWeave. The article also references Pennsylvania’s $70 billion push for data centers and notes many 2028–2032 projects have not broken ground.
• Apr 9, 2026

  Hyperscalers will own two-thirds of data center capacity by 2031

  Synergy Research Group reported that hyperscalers will account for 67% of all data center capacity by 2031.

  • Main announcement: Synergy Research Group says hyperscalers (Google, Microsoft, AWS) will reach 67% of global data center capacity by 2031, with enterprise on-prem data centers dropping from 56% in 2018 to 19% by 2031; the report also notes almost 60% of hyperscale capacity is in own-built facilities and non-hyperscale colocation accounts for ~20%.
  • Background & details: The article cites planned > $500 billion in capex by Google/Microsoft/AWS for AI infrastructure in fiscal year 2026, cites hyperscalers operating ~1,297 large data centers in Q3 2025 (1,360 by end-2025), references commitments such as the Ratepayer Protection Pledge (Google, Oracle, xAI, Meta, Microsoft, OpenAI, Amazon) and highlights electricity demand concerns (EIA: price hikes up to 79% in areas like Texas by 2027); it references expanded compute partnerships (Anthropic–Google/Broadcom; OpenAI–AMD) with multi-gigawatt capacity starting 2027.
• Apr 8, 2026

  Pa. data centers: How lawmakers are responding, from electricity and water use to tax breaks

  Pennsylvania lawmakers are proposing and debating multiple bills to regulate data center development, including requirements on payments for grid upgrades, reporting of energy and water use, and conditions on tax incentives.

  • Main action: The Pennsylvania House passed a bill (sponsored by Rep. Robert Matzie) requiring the state Public Utility Commission to develop rules forcing data centers to pay for grid upgrades, post deposits to cover stranded costs, provide at least 10% of their electricity from new clean power starting in 2027 (or pay into a clean energy/efficiency fund), and agree to curtail power use during emergencies; the measure must still pass the Republican-controlled Senate.
  • Background and other measures: Other proposed bills include HB2246 (reporting for data centers using >100,000 gallons/day of water), HB2150 (annual energy and water reporting), HB2198 (would repeal a 2022 sales-tax exemption for data center equipment), HB2153 (directs increased tax revenue from new data centers to homeowner/farm tax relief), and Senate proposals SB991 / SB939 to fast-track permitting (preapproved sites and 120-day permit timeline for large centers bringing their own power); the Data Center Coalition (representing AWS, Google, Microsoft, Anthropic, CoreWeave, OpenAI) opposes data-center-specific mandates and spent $19,632 lobbying in Pennsylvania in late 2025.
• Apr 6, 2026

  Neocloud Storm Gathers as Data Center Deals Stall Over Credit Risk

  A well-funded AI neocloud provider sought an initial 2 MW deployment (scaling to 12 MW within 18 months) on a 15-year term with a six-month prepaid deposit and millions in liquid-cooling capex, but the colocation deal failed.

  • Main announcement/action: The neocloud sought 2 MW initially, scaling to 12 MW within 18 months, offered $155-$160 per kW, a 15-year term, prepaid six months of recurring charges, and committed to millions in upfront liquid cooling infrastructure, yet colocation providers declined to close the deal.
  • Background and context: Market participants (cited by Robert West of TRG Datacenters and datacenterHawk research) report providers now require investment-grade credit, visibility into end-customer demand, and balance-sheet durability; wholesale pricing ranges of $140-$155 per kW are reported and liquid cooling adds ~$1.5M-$1.6M per MW in upfront capital.
• Apr 1, 2026

  OpenAI Raises $122B to Expand AI Infrastructure, Broadens Cloud and Chip Strategy

  OpenAI has announced it raised $122 billion in a funding round valuing the company at $852 billion to expand compute capacity, cloud partnerships, and data center infrastructure.

  • Funding and purpose:$122 billion raised at an $852 billion valuation to support expanded compute capacity, cloud partnerships, and data center expansion; OpenAI also expanded its credit facility to $4.7 billion to provide additional flexibility for capacity investments.
  • Infrastructure and partners: OpenAI is working across Microsoft, Oracle, AWS, CoreWeave, and Google Cloud and using a mix of silicon platforms including Nvidia GPUs, AMD chips, AWS Trainium, Cerebras systems, and custom Broadcom silicon; the platform now processes more than 15 billion tokens per minute, and analysts named include Holger Mueller, Daniel Newman, and Matt Kimball.