The Vertiv UIG acquisition adds microgrid, storage and behind-the-meter power controls for up to $2.6B, attacking AI data center grid delays.
Broadcom AI chip revenue climbed 221% to $16.7B in fiscal Q3 as the company set a $230B target for 2028. Inside the numbers and what they mean for Nvidia.
Adobe for Slack brings 70+ creative tools such as Firefly, Photoshop and Acrobat into Slackbot, so teams can generate and edit content directly in chat.
PayPal AI layoffs are underway: the first waves remove hundreds of roles in India, Ireland, and Israel as part of a 4,760-job AI restructuring.
Google DeepMind's WeatherNext 3 issues hourly global forecasts up to 5 km sharp and now powers weather in Search, Gemini, Maps and Google Cloud.
Read more →
Equinix Inference Exchange gives enterprises an inference-first way around the AI compute scarcity premium exposed by Anthropic's $35B Lambda pact.
Muse Spark 1.3, Meta's most powerful model yet, posts a frontier-level benchmark score of 62, but developers pay for the API while consumers get the model free.
The Nvidia Hugging Face acquisition, a $12.93B bet near 86x revenue, puts the open-weight model hub inside the dominant AI chipmaker. We analyze what changes.
Brussels' EU AI Act enforcement reaches US labs: information requests to 30+ AI providers test extraterritorial reach as Washington pushes hands-off G20 rules.
The SB Energy IPO filing reveals heavy OpenAI dependence behind a $439B backlog, $3.2B H1 losses, and Nvidia guarantees for its AI buildout.
OpenAI's Astra is its first model to clear the critical cybersecurity threshold, yet autonomous zero-day tools launch restricted to alpha testers.
CrowdStrike SafeMind: NVIDIA Nemotron 3 agents on Falcon claim 29% higher detection, 6x faster response, and 99% lower cost than frontier models.
The Google zinc battery storage project pairs solar with zinc and lithium batteries to keep West Virginia AI data centers running around the clock.
Google's Gemini 3.8 Flash and Flash Cyber arrive as the third Flash release in six weeks, holding token prices flat while opening a gated agentic-security front.
Google's agentic video understanding for Gemini 3.7 Flash, 3.6 Flash and 3.5 Flash-Lite cuts token use by up to 88% and cost by 66%.
