Getting your news
Attempting to reconnect
Finding the latest in Climate
Hang in there while we load your news feed
CoreWeave
Data center news, project activity, and monthly briefings for CoreWeave.
CoreWeave · Construction & power moves · 1
full tracker →Where CoreWeave is securing land and power — each traced to primary filings.
Editor's picks
-
Why Performance per Watt Is the Ultimate Metric for AI Infrastructure Efficiency
NVIDIA has published a blog post arguing that power efficiency is the defining constraint for AI factories and highlighting the rack-scale performance of its Blackwell and Vera Rubin platforms.
- The post says NVIDIA GB300 NVL72 delivers up to 25x performance per watt versus Hopper on leading open models, with additional examples of up to 20x on GLM5.1 and 10x on Kimi K2.6.
- It also says NVIDIA DSX MaxLPS can help operators run up to 40% more GPUs within the same power budget, and cites production use by Anthropic, OpenAI, CoreWeave, Perplexity, and Fireworks AI.
- This is a commentary-style NVIDIA blog, not a standalone product launch announcement, though it references existing deployments and platform capabilities.
-
Cloud Hidden, Rationale Unknown: The DMA’s Foggy Attack on AWS and Azure
The article is an opinion/commentary criticizing the European Commission’s preliminary decision to designate AWS and Azure under the Digital Markets Act.
- The author argues the Commission’s preliminary decision would extend the DMA to cloud computing services offered by Amazon’s AWS and Microsoft’s Azure, despite the services not meeting the statute’s quantitative thresholds.
- It says the DMA’s thresholds include €7.5 billion in EU revenue or €75 billion market capitalization for gatekeepers, and 45 million monthly active end users plus 10,000 yearly active business users for core platform services; it also compares AWS and Azure with Spotify and SAP as examples of EU firms allegedly overlooked.
-
NVIDIA DSX Air Boosts Time to Token With Accelerated Simulation for AI Factories
NVIDIA announced NVIDIA DSX Air at GTC 2026, a SaaS simulation platform to build full-stack digital twins of AI factories and accelerate deployments.
- Main announcement:NVIDIA DSX Air is a software-as-a-service platform that provides high-fidelity digital simulations of NVIDIA hardware (GPUs, SuperNICs, DPUs, switches) and integrates with partner solutions via open, API-based connectivity, enabling customers to validate compute, networking, storage, orchestration and security and reduce time to first token from weeks/months to days/hours.
- Background & implementation details: Partners and customers (examples: CoreWeave, Siam.AI, Hydra Host) are already using DSX Air; GTC demos included a multi-tenant RTX PRO Server simulation with Netris (network orchestration), Rafay (host orchestration), NVIDIA Run:ai (GPU allocation), a VAST AI Operating System video RAG pipeline demo, and security validations with Check Point, TrendAI Vision One, and Keysight running on simulated BlueField DPUs.
-
Commercio e finanza in un mondo frammentato
Fabio Panetta, Governor of Banca d’Italia, delivered a keynote speech at the 32nd ASSIOM FOREX Congress (19 February 2026, Venice).
- Main announcement/action: Governor Panetta highlighted that artificial intelligence is a central driver of recent global growth, with data center construction singled out as a core element of the technological transformation; he also called for stronger European coordination on energy, financial integration, and digital currency initiatives (projects Pontes and Appia) and flagged the need to govern changes in the international monetary system. Key facts: global GDP +3.3% (2025); international trade +4% (2025); dollar ≈60% of official reserves; Pontes experimentation by end-2026.
- Background and details: The speech documented trade reorganisation due to US tariffs (effective tariff rate reduced from 20% to 16%, border collections about 11%), noted the US–China reorientation of trade flows toward third countries (Mexico, Vietnam, Taiwan), reported the €2,000 billion (USD 2 trillion) nominal exposure of European investors to US government securities, and described monetary-policy and market developments (e.g., dollar depreciation up to 8% vs euro in April–May 2025).
Recent news
-
The AI Infrastructure Split Screen: Capital Rush Meets Community Resistance
Data Center Frontier has published an analysis of the mid-July AI infrastructure market, combining public-market activity, project announcements, financing moves, and local permitting disputes.
- Capital markets: Csquare priced 50 million shares at $21 to raise about $1.05 billion and implied an equity value of roughly $3.2 billion; Switch was linked to a potential $10 billion IPO and near $80 billion valuation including debt; Databricks signed a term sheet for a financing at a $188 billion valuation with the Wall Street Journal reporting about $3 billion in new investment.
- Project and policy developments: Meta said its Hyperion expansion in Louisiana could reach 5 GW and more than $50 billion; Google was tied to a 2.7 GW Project Tembo in Wyoming; New York Gov. Kathy Hochul ordered an up-to-one-year pause on discretionary environmental permitting for certain hyperscale data centers; Palm Beach County commissioners rejected Project Tango by 5–1 after a roughly 12-hour hearing.
- The article is a broad market commentary and synthesis, not a single original company announcement. It references multiple first-reported deals and filings, while also citing local opposition, power constraints, and community consent as central themes.
-
Google is customer behind 2.7GW data center campus near Cheyenne, Wyoming
Google has revealed itself as the company behind the Project Jade data center campus near Cheyenne, Wyoming, and the project has been renamed Project Tembo.
- The campus is planned on a 716-acre parcel eight miles south of Cheyenne in the Switchgrass Industrial Park and is described as a 2.7GW data center development.
- The application says four data halls, an office hub, logistics and network buildings are planned, with construction slated to be completed by 2031; the site was previously associated with Crusoe, which said in June it had paused work at the request of its customer.
- Google’s subsidiary Jupiter Star Holdings, LLC is now listed as the developer, while Tallgrass Energy said it was still building a bring-your-own power generation hub and Google’s application says it will also use energy resources from Black Hills Energy.
- The article also notes other large Cheyenne projects from Microsoft, Meta, and Related Digital/CoreWeave for local context.
-
How Yotta Is Building Its AI Infra Playbook To Serve Global Enterprises
Yotta Data Services has announced plans to scale its AI cloud and financing for AI infrastructure while preparing for an IPO.
- The company expects its cumulative AI infrastructure investment to exceed $7 Bn by FY27 and says about $4 Bn has already been committed to AI infrastructure.
- Yotta has raised $150 Mn from non-institutional investors at a valuation of ₹37,000 Cr (about $3.9 Bn), plans to file its DRHP in the coming months, and is targeting about $1.5 Bn across pre-IPO rounds and the IPO.
- Yotta expects its AI cloud to scale to 40,000+ NVIDIA Blackwell GPUs in four months and 85,000 GPUs by the end of the financial year; it also said an April partnership with Gorilla added $2.8 Bn to an agreement initially valued at $500 Mn.
-
Cerebras to deploy 200MW of compute in Europe by end of 2027
Cerebras Systems has announced it will bring its first data center capacity online in Europe by the end of 2026 and aims to deploy 200MW across Europe by the end of 2027.
- Europe expansion will focus on France and the Nordics, with data centers slated for Norway and Finland; CEO Andrew Feldman said the company is contracting significant capacity for 2027.
- The article also notes prior and parallel capacity deals: 40MW leased from a Digi Power X data center in Alabama in May, plus deployments in the US, Canada, and other regions; no European operator partners were named.
-
Galaxy Digital completes first phase of pivoting cryptomine to AI hosting
Galaxy Digital has announced completion of the first phase of its Helios campus pivot toward AI and HPC use, delivering power to CoreWeave.
- Phase I at the Helios data center campus in West Texas is complete and has delivered about 200MW gross power and 133MW critical IT load to CoreWeave; Phase II greenfield development is underway with 260MW planned.
- CoreWeave has committed to 526MW critical IT load (800MW gross power) across three phases under 15-year leases; Galaxy says the campus total approved power capacity has expanded to 1.63GW, with potential to reach 3.6GW. Galaxy acquired the campus in late 2022 from Argo Blockchain for about $65 million and later announced the pivot to AI use.
-
Anthropic signs $19bn, 20-year lease for Kentucky data center with TeraWulf
TeraWulf has announced a long-term lease with Anthropic for its Justified Data campus in Hawesville, Kentucky, and separately said it is selling its 50.1% ownership interest in the Abernathy Joint Venture.
- Anthropic lease: the agreement runs for 20 years and is expected to generate about $19 billion in revenue for TeraWulf; the campus is planned for up to 401MW of IT capacity, with initial capacity in the second half of 2027 and full operations by early 2028.
- Abernathy JV sale: TeraWulf said it is divesting its entire 50.1% ownership in the JV with Fluidstack in Abernathy, Texas; the JV was formed in October 2025 for a 168MW data center on a 120-acre campus, and TeraWulf said the deal helps it realize value from its $450m investment.
-
Nvidia acts as backstop for customer GPUs in return for cut of cloud revenue
Nvidia has announced a new credit-support and revenue-sharing model for AI cloud customers purchasing its AI chips.
- Nvidia said the model lets AI clouds procure Nvidia infrastructure for AI-native, enterprise, and ISV customers through a revenue-sharing and credit-support model.
- Under the partnership, Nvidia will rent back unused GPUs at a fixed rate and earn both standard product revenue and a share of cloud revenue on supported capacity; Firmus and Sharon AI are among the first adopters.
- Firmus is deploying 170,000 GPUs in Batam, Indonesia; Sharon AI is deploying 40,000 GB300 GPUs.
- The article also references earlier 2025 deals with CoreWeave and Lambda, including a September 2025 CoreWeave agreement valued at $6.3bn and a Lambda deal valued at $1.5bn.
-
Meta pushes ahead with plans to launch cloud business - report
Meta is reportedly planning a new cloud computing business unit to sell access to AI computing power and models.
- People familiar with the matter told Bloomberg that Meta is developing plans for a cloud business; Mark Zuckerberg had previously said the idea was “definitely on the table” but not an active plan.
- Meta CFO Susan Li said 2026 capex expectations were raised to $120 billion-$135 billion, reflecting higher component pricing and additional data center costs; Meta Compute was created earlier this year to expand data center capacity, with Zuckerberg saying it is planning “tens of gigawatts this decade” and “hundreds of gigawatts or more over time.”
-
Bulk buys land in Arendal, Norway, for data center development
Bulk Infrastructure has announced the acquisition of a plot in Arendal, Norway, where it plans to develop a data center with a 150MW grid connection by the end of 2029.
- Arendal officials approved the sale of plot no. 25, block no. 121 on Longum Nord for NOK 600 million ($60.47m); the land is located in the Eyde Material Park and will be acquired via Bulk Data Centers Arendal AS.
- The municipality bought the plot in 2024 for NOK 200 million ($20.16m) and had planned a recycling facility, but those plans were dropped; the park totals 250 hectares and claims 800MW of power availability, while the Longum Nord portion totals 82 hectares.
-
Galaxy Digital eyes second Texas data center site, buys land outside Waco
Galaxy Digital has announced an agreement to acquire land in McGregor, Texas (Project Merlin) to develop a new data center campus.
- Main announcement: Galaxy Digital aims to acquire 500 acres in McGregor’s industrial park (Project Merlin) and invest some $400 million to develop up to eight buildings, with the site to use closed-loop cooling systems to reduce water usage and Galaxy covering infrastructure upgrade costs.
- Background and other details: The move would be Galaxy’s second Texas campus after its 160-acre Helios campus (acquired from Argo Blockchain for approximately $65 million in late 2022); Helios has an 800MW approved capacity (fully leased to CoreWeave) and a lease projected to generate average annual revenue of more than $1 billion over a 15-year term.
-
CoreWeave signs colocation agreement with Conapto in Sweden
CoreWeave has signed a colocation agreement to lease capacity from Conapto in Sweden.
- Colocation agreement signed: CoreWeave will lease capacity from Conapto’s Stockholm campuses, with initial capacity already online at Stockholm 4 South; CoreWeave will deploy Nvidia Blackwell and Vera Rubin GPUs connected via Nvidia Quantum-X800 InfiniBand, and the deal brings CoreWeave’s European presence to eight sites.
- Site and operational details: Stockholm 4 South is Conapto’s newest and largest data center with 32MW IT capacity across 3,000 sqm; Conapto provides 100% renewable energy and heat recovery to Stockholm’s district heating network, with power supplied via Vattenfall 24/7 matching; CoreWeave reported operating in 49 data centers globally with >1GW active power and 3.5GW contracted as of March 31, 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).
-
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.
-
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.