Impact on assets
Theme Impact 7
Telecom Towers, Fiber & ColocationCloud & Digital Infrastructure
AI Data Center & Build-outArtificial Intelligence
Hyperscale Cloud (IaaS / PaaS)Cloud & Digital Infrastructure
Build-out, Construction & EngineeringArtificial Intelligence
Colocation & Hyperscale REITsArtificial Intelligence
AI Compute Cloud & NeocloudsArtificial Intelligence
Grid, Transmission & Power EquipmentEnergy Transition & Power Demand
Off-coverage companies
Related news
▼impact 4
Oracle has served a force majeure notice on its Project Jupiter AI data center development in New Mexico, which is being developed by Stack Infrastructure and owned by Blue Owl Capital. Oracle says there is no delay and that things are going as planned, but Reuters is reporting a potential one-year delay on the buildout. The notice has renewed concerns about Oracle's leverage and what one guest called potential over-financing, and it raises the broader question of whether slowing projects could threaten the AI investment case. Goldman Sachs estimates the major US hyperscalers will spend about $800 billion in 2026, rising to $1.1 trillion in 2027, and says roughly $300 billion in annual AI revenue will be needed in the next few years to break even on those CAPEX numbers. Goldman also estimates that AI users would need to spend roughly $1 trillion annually on AI applications for the hyperscalers to generate solid returns on investment and for the application layer to generate strong profit margins on their compute expenses.
Wison New Energies Unveils Floating Data Center Concept at Gastech 2026
Wison New Energies unveiled a floating data center concept at the Gastech 2026 conference in Bangkok, pairing its Floating Storage Regasification to Power technology with modular data-center infrastructure on a single offshore platform. The Floating Power Data Center, or FPDC, stores and regasifies LNG onboard and converts it directly to electricity for onsite computing racks, and the company is examining whether cold energy released during regasification could help cool the servers. The concept responds to mounting constraints on land-based data centers in the U.S., including multi-year grid interconnection queues, water-cooling objections and scarce land near metro and fiber hubs, as well as political backlash such as Texas Governor Greg Abbott's August directive to pause the state's data center permitting process pending a review of water and energy usage. Wison's most recent proof point in offshore execution came this summer, when it signed the EPCIC contract for the Baleine Phase 3 FPSO serving Eni's offshore field in Côte d'Ivoire, a nearly 308-meter facility designed to process 90,000 barrels of oil a day. At Gastech, Wison also signed a collaboration agreement with Shell to integrate Shell's Dual Mixed Refrigerant liquefaction technology into its floating LNG portfolio, the first time the process will be available across the wider FLNG market, and a strategic memorandum of understanding with Houston-based KBR covering engineering, advisory services and integrated project solutions. Norway's DNV granted the world's first FLNG ABATE Notation FEED Approval for Wison's Low Emission FLNG, and the Houston-headquartered American Bureau of Shipping granted Approval in Principle for Wison's Floating Storage and Regasification Unit design for ammonia.
▲2
MeetKai Rolls Out Sovereign AI on NVIDIA Across Six Countries
MeetKai announced a six-country rollout of its sovereign AI stack, built on NVIDIA AI infrastructure, across Brazil, Ukraine, Pakistan, Kazakhstan, Uzbekistan and Bangladesh, six countries home to more than 700 million people. In Brazil, MeetKai Brasil won SERPRO's public tender for the national sovereign AI platform with the highest overall score and is now launching it with SERPRO, including a sovereign coding agent for secure workflows. In Ukraine, MeetKai and Kyivstar, VEON's majority owned subsidiary and leading Ukrainian digital operator, will build a sovereign AI data center, the first national-scale AI factory in the country, with first capacity planned to come online in 2027 and a design expected to expand from 15 to 100 MW as demand from the public sector and Ukrainian industry grows. Across VEON's five markets, home to more than 150 million mobile customers and over 225 million digital customers, MeetKai acts as VEON's AI lab-as-a-service, beginning in Ukraine followed by Pakistan, Kazakhstan, Uzbekistan and Bangladesh, with each site operated by MeetKai together with VEON's local operator. MeetKai, an NVIDIA Inception member, is also expanding its work in Egypt with Smart Africa, building its AI stack on NVIDIA technology.