AI Infrastructure news and analysis on Web Pulse: 34372 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
PyTorch 2.14 shipped on September 15 with 2,995 commits from 487 contributors, adding fault-tolerant distributed training as a first-class framework concept via a rewritten NCCL backend (nccl2) with in-place process-grou…
TypeSafe AI, a startup founded by former OpenAI researcher and RLHF co-inventor Diogo Almeida, launched Jev, an LLM that returns a defined decision plus its probability in a concise response instead of generating verbose…
Lucius AI, a solo-founder tender-intelligence startup, cut semantic-search query latency 47x — from 1.14 seconds to 24 milliseconds — by migrating to a ScaNN index on AlloyDB for PostgreSQL and managing database operatio…
Meta Platforms CEO Mark Zuckerberg, SpaceX CEO Elon Musk, and NVIDIA CEO Jensen Huang each spoke separately with President Donald Trump last week and persuaded him to abandon a proposed federal AI oversight body, accordi…
Amazon Bedrock Knowledge Bases supports three customer-managed vector store backends — Amazon OpenSearch Service, Amazon Aurora PostgreSQL with pgvector, and Amazon S3 Vectors — according to an AWS Machine Learning Blog …
AWS published a serverless Git metrics dashboard that automates collection from GitHub and GitLab APIs and visualizes the data in Amazon Quick Sight, using six core AWS services including EventBridge Scheduler, Step Func…
Wood Mackenzie built APEX (Agentic Platform for Energy eXperience), a shared agentic platform on Amazon Bedrock AgentCore, after finding that 88 percent of its AI proofs-of-concept never reach widescale deployment. The p…
MRH Trowe, one of Germany's leading commercial and industrial insurance brokers, rolled out secure self-service AI agents to approximately 400 employees in the first month of production, combining Strands Agents, Amazon …
Amazon published a walkthrough for implementing a multi-gate, defense-in-depth authorization pattern for Model Context Protocol (MCP) tool invocations on Amazon Quick, evaluating OpenID Connect (OIDC) JSON Web Token (JWT…
A synthetic data augmentation pipeline built on Amazon SageMaker AI and Amazon Rekognition improved person-detection mAP50 by up to 160 percent for industrial safety models without manual annotation or hazardous photogra…
Asus unveiled the Ascent QN10, which it calls the "world's first" mini PC built for agentic AI, powered by an 18-core Qualcomm Snapdragon X2 Elite chip with an 80 TOPS NPU, priced at $1,350 for 16GB of RAM and $1,700 for…
Anthropic signed its first Australian data center lease for the proposed Western Downs Digital Park near Dalby, Queensland, a campus planned to reach 2.16 gigawatts of capacity that could begin coming online in 2027, acc…
Eco Router, an MCP server that ranks AWS, Google Cloud and Azure regions by grid carbon intensity and latency, was published to the official MCP registry on 2026-09-17 as version 0.1.0 under the npm package eco-router-mc…
Compu Dynamics Modular Vice President Ron Mann will appear in an upcoming Data Center Dynamics virtual broadcast on scaling AI infrastructure, citybiz reported. The discussion will examine how data center operators are e…
The digital workspace is shifting from an application-delivery model to an orchestration layer that connects people, AI agents, applications, enterprise data and security policies, according to Forbes Councils member Ele…
OpenAI's chatgpt.com serves roughly 1 billion people, making it one of the most used web apps in the world, according to a reverse-engineering analysis of the site's HTML, JavaScript, CSS, and network requests. The analy…
Bart Smith, CEO of Avalanche Treasury Company (AVAT), warned on the July 9 episode of the Layer One podcast that proliferating AI agents will strain shared blockchains, arguing Avalanche's subnet architecture gives it an…
Forbes Councils member Brian Wilson writes that hidden token costs are making AI prohibitively expensive, and that FinOps practices can trim the bill. Wilson argues that when organizations can see and control token spend…
Fabric.AI (Nasdaq: FABC) has retained Haynes and Boone LLP as intellectual property counsel for Neural I/o, its MicroLED-based optical interconnect technology developed with Kopin. The appointment covers IP protection fo…
Apple is reportedly developing an enterprise AI server built around M8 Ultra chips, targeting 2029, according to The Information, with systems under consideration carrying either two or four M8 Ultra processors. The proj…