Showing 21–40 of 206 articles in data-centres
As hyperscale AI campuses scale up, water and wastewater capacity are emerging as siting gatekeepers, reshaping cooling choices, municipal planning, and project approvals.
Hyperscalers are adopting EV-style high-voltage power systems to reduce copper, cooling strain, and conversion losses.
Denmark halted new large-load grid agreements as AI, Power-to-X, and electrification demand overwhelm capacity, forcing power-access triage.
Enterprise AI workloads are moving from experimental pilots into persistent operational infrastructure, reshaping hyperscale compute demand.
A new platform from CoreWeave combines inference, reinforcement learning, and observability to continuously optimize AI agents using live production data.
Power constraints, permitting timelines and grid access are redefining Europe’s data center market country by country, pushing growth toward projects that can move forward.
Data Center Knowledge toured the TeraWulf and Schneider Electric campus in Buffalo, N.Y., to see what it takes to power – and cool – the data centers of the future.
Utah’s proposed 9 GW Stratos campus reflects a broader evolution toward AI infrastructure built around dedicated energy systems, accelerated permitting, and direct control over power.
Broadcom and FuriosaAI are building a rack-scale inference platform aimed at shifting AI infrastructure toward Ethernet fabrics, chiplets, and power-efficient token generation.
Texas built CREZ to move wind power. The same transmission corridors now increasingly attract hyperscale AI infrastructure.
The EPA’s plan to fast-track data center construction may cut red tape but also ignite legal battles and intensify environmental concerns.
AI infrastructure developers are no longer just buying cooling equipment. They are beginning to reserve industrial production capacity years in advance.
The Ratepayer Protection Act would require long‑term contracts, minimum billing, and closed‑loop cooling at large data centers.
Utilities are planning around giant AI training campuses, but distributed inference workloads could alter where and how future power demand emerges across the grid.
AI-scale computing is driving data centers to adopt advanced power electronics that can handle voltage fluctuations and load swings electrically, relegating diesel generators to a backup role for extreme conditions only.
As hyperscale data center timelines tighten, companies must assess how building material decisions could introduce schedule, sequencing, and rework risks.
Dallas, West Texas, and Austin-San Antonio climbed Cushman’s global rankings as AI infrastructure chases power and land.
Mathpix’s Brooklyn GPU deployment highlights how production AI workloads are driving new demand for urban colocation infrastructure.
A 2025 George Mason University study found homes near data centers had higher values. But does this trend hold true everywhere?
Explosive networking growth and new optics partnerships suggest AI infrastructure spending is spreading far beyond GPU clusters.