#ai-infrastructure

Google engineers say a new compression method called TurboQuant can cut AI working-memory needs by up to six times without sacrificing model performance, potentially easing one of the infrastructure burdens of large chat

Amazon says AWS revenue rose 28% year over year to $37.6 billion, its fastest growth rate in 15 quarters, even as the company signals continued heavy capital spending.

OpenAI says its Stargate effort has already moved beyond the 10-gigawatt U.S. infrastructure target it originally set for 2029, with more than 3 gigawatts added in the last 90 days alone.

Developers behind the huge Digital Gateway proposal in Prince William County are pulling back after years of legal and community resistance, marking a major moment for local campaigns challenging large-scale data center 

A London-based data-center developer has paused Middle East investments after war damage to one of its facilities, underscoring how conflict risk is colliding with the tech industry’s push to build major AI and cloud hub

Large data center projects are increasingly stalling before construction begins, as power shortages, public resistance and proposed state restrictions collide with the industry’s rush toward gigawatt-scale development.

Google is spotlighting its Tensor Processing Units as demand for larger AI workloads accelerates, framing custom chips as a core advantage in the race to train and run modern models at scale.

OpenAI has released Privacy Filter, an open-weight model for detecting and redacting personally identifiable information in text, positioning it as practical infrastructure for safer AI pipelines

Maine Governor Janet Mills has vetoed a bill that would have paused permits for new data centers until November 2027, rejecting what would have been the first statewide moratorium of its kind in the United 

Google is reportedly committing an initial $10 billion to Anthropic, with another $30 billion tied to performance targets, while also expanding the AI company’s access to large-scale Google Cloud compute.

Meta has reserved up to 1GW and 100GWh of energy storage capacity from Noon Energy in a deal aimed at supplying more reliable renewable power for data centers.

A novel medium-voltage UPS is being put through simulated grid and data-center conditions to see whether it can blunt power spikes, ride through disturbances, and make large computing loads easier to absorb.

Google says its new TPU 8i and TPU 8t chips are specialized for distinct AI workloads, with one focused on fast agent performance and the other on large-scale model training.

North American grid watchdog NERC is preparing a Level 3 essential actions alert after multiple large computational loads, including data centers, unexpectedly dropped 1,000 MW or more from the bulk power system in recen

A prolonged memory shortage may continue to squeeze electronics makers for years, with production expected to meet only 60 percent of demand by the end of 2027 and new capacity largely aimed at AI-focused memory.

The Federal Energy Regulatory Commission says it will act in June on reforms related to connecting data centers and other large loads to the transmission system, putting jurisdiction and cost allocation at the center of

The NAACP has sued xAI over the gas turbines powering its Colossus data center, arguing the units are operating without permits and placing pollution burdens on a nearby Black working-class community.

Span says it is working with Nvidia and partners on XFRA, a network that would place high-powered compute nodes at homes and small businesses, using smart panels to turn spare electrical capacity into distributed AI infr

OpenAI has expanded its Agents SDK with native sandbox execution and a model-native harness, a move aimed at making long-running, tool-using agents easier to build and safer to operate.

The Energy Information Administration plans a mandatory nationwide survey of data center electricity use, creating a new federal window into one of AI’s fastest-growing infrastructure demands.