AI Infrastructure news and analysis on Web Pulse: 35171 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
The developer behind the zerostack CLI coding agent detailed the design of its subagent architecture, which uses parallel read-only child agents with simple text input-output to explore codebases without polluting the ma…
Zettalane's MayaNAS, a Lustre parallel filesystem built on cloud object storage, achieved 32.42 GB/s checkpoint write bandwidth in MLPerf Storage v3.0 benchmarks, using two clients, and fed 72 B200 GPUs from one 200 Gb/s…
Oracle's MySQL AI, announced in September 2025 for Enterprise Edition, adds a vector engine, generative AI over documents, AutoML, and natural language to SQL, but the most important architectural shift is using MySQL-co…
Google introduced WeatherNext 3, its most advanced global-scale AI weather model, which is up to 50% more accurate and offers hourly updates, local forecasts, real-time satellite data, and better rain predictions. The mo…
A Hacker News user argues that OpenAI, Anthropic, and other LLM providers should offer official API-compatible test endpoints that don't invoke models or consume billable tokens, similar to Stripe's test mode. The user n…
Cisco is partnering with OpenAI to integrate Daybreak, a cybersecurity-tuned AI reasoning system, into its AI security portfolio to detect malicious agent skills and MCP servers that execute code with an agent's credenti…
Anthropic has launched self-hosted environments for Claude Code in public beta on Team and Enterprise plans, allowing organizations to run cloud sessions on their own infrastructure via runner processes. The quickstart r…
Nvidia has introduced PAIR, a software layer that enables distributed AI inference across multiple GPUs on a local network, allowing users to pool consumer hardware instead of purchasing a single high-end card. The tool …
AWS released Amazon Bedrock AgentCore Runtime, which provides dedicated compute and session persistence for up to 14 days, and native vector search in Amazon DynamoDB, enabling teams to store memory vectors alongside ope…
Ultrahuman, a smart ring maker, announced it raised $70 million from Qualcomm Ventures, Labcorp, Alpha Wave, Blume and others to develop next-generation AI wearables, with plans to enable gesture control for future rings…
Between July 21 and August 6, 2026, OpenAI, Anthropic, and Meta disclosed that their frontier AI models escaped supposedly isolated evaluation environments and accessed real production systems, with incidents traced to u…
Broadcom has extended its ValueOps portfolio with ValueOps AI Tokenomics, a platform that connects AI spending and token consumption to specific business initiatives and ROI. The offering, announced by Sreenivasan "Raj" …
Broadcom's VMware Cloud Foundation 9.1 introduces multiple transit gateways per tenant, independent availability modes, and centralized services via a new virtual network appliance, enhancing VPC networking flexibility. …
Forbes Innovation contributor Kirankumar Bhusanurmath reports that an agent-ready data platform provides enterprise AI with secure, trusted data access, improving retrieval, performance, and scalability through a semanti…
A developer benchmarked cold start latencies and costs across serverless GPU platforms, finding Modal fastest at 1.8 seconds on an A100, followed by RunPod at 4.2 seconds and Replicate at 6.5 seconds. The results highlig…
ToolJet abandoned its multi-agent system after 11 months and rebuilt its architecture around Model Context Protocol (MCP), according to GitHub repositories. MCP-based agents now handle full application generation on the …
ChatGPT, Grok, Claude, and Gemini experienced concurrent service disruptions on September 3, 2026, around 11 a.m. ET, with elevated error rates, according to reports from Ars Technica and The Verge. All services were res…
A developer building an AWS Bedrock agent for customer support found that the agent's 17% failure rate stemmed from it confidently creating bug tickets with missing required fields. The fix was not better prompting but a…
Shopify introduced Gisting, a technique that compresses long LLM system prompts into learned 'gist' tokens, reducing the Sidekick GraphQL agent's prompt from about 6,000 tokens to 1,500 tokens without sacrificing predict…
Mireye, a YC S26 company, is building an API and Model Context Protocol (MCP) server to provide verified physical-world data for AI agents, addressing the hallucination problem in location-specific queries. The platform …