AI Infrastructure news and analysis on Web Pulse: 32135 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
LG Energy Solution held its Battery Tech Conference in Chicago to recruit top global research talent in next-generation battery technologies, energy storage systems, and artificial intelligence. CEO Kim Dong-myung emphas…
PingCAP's research team introduced TINE, a system that binds every AI agent iteration to a Git commit, a TiDB branch, and a sandbox runtime to solve schema drift, irreproducible state, and rollback gaps in agentic app bu…
AgentBridge, an open-source Python project, provides a neutral mesh for translating, routing, verifying, and governing calls between different AI agent protocols, including MCP, A2A, ACP, OpenAI function-calling, Gemini …
Schneider Electric and Foxconn have formed an alliance to accelerate AI data center deployment, combining expertise in power infrastructure and hardware at scale. The partnership aims to meet surging demand for AI-ready …
BridgeX API argues that AI agents require native messaging infrastructure rather than traditional SMS APIs, proposing the Model Context Protocol (MCP) to make messaging AI-native.
A developer built The Gauntlet, an open-source Next.js 16 app that connects seven MCP servers through a LangChain multi-agent pipeline and lets users toggle eight failure modes live during execution. The tool is designed…
A Hacker News user asked the community to share their local large language model setups, including models, hardware, purposes, challenges, and tips, to help newcomers.
Researchers propose HiLo-Token, an input-adaptive token compression framework that reduces latency in Diffusion Transformers for image editing by allocating more tokens to high-frequency regions and fewer to low-frequenc…
Researchers propose Gefen, a memory-efficient optimizer that reduces AdamW's memory footprint by ~8x while maintaining performance, enabling larger microbatches and improved throughput in deep learning training. Gefen au…
Researchers introduced a fully GPU-based workflow for building physics emulators of hypersonic flows, integrating accelerated data generation with neural emulator training augmented by uncertainty quantification and phys…
Researchers introduced D2H-AD, a hybrid anomaly detection framework using Hyperdimensional Computing that outperforms five baselines including One-Class SVM and Autoencoders across benchmark datasets. The lightweight mod…
Researchers introduced llada.cpp, the first NPU-aware inference framework for accelerating diffusion large language models on smartphones, achieving 17x-42x latency reduction over CPU baselines while preserving generatio…
AI agents can fail in subtle, non-obvious ways that resemble progress, including goal drift, tool misuse, state loss, hallucination, and premature task completion. These hidden failure modes make agent reliability challe…
Amazon is offering the ASUS TUF Gaming F16 laptop with an RTX 5050 GPU and 16GB DDR5 RAM at 21% off, priced at $1,029.99. The deal targets gamers seeking affordable FHD performance with Blackwell architecture and a 165Hz…
Innocean, the marketing arm of Hyundai Motor Group, partnered with venture capital firm SBVA to launch UP 2026, a growth platform for about 10 AI and technology startups including Ohou, Blind, and Kream. The program aims…
ByteDance, TikTok's parent company, is negotiating to buy at least 50,000 AI inference GPUs from Shanghai-based Iluvatar CoreX, as US export controls push Chinese tech firms to diversify away from Nvidia. The company is …
QuantaMind built a local AI evaluation engine using Tauri and Rust, enforcing strict engineering invariants for reproducibility. The team separated the inference core into pure Rust modules independent of the UI, mandate…
Google revealed that speed is a critical factor for its web search service, as even minor delays can significantly impact user satisfaction and search behavior. The company found that a 400-millisecond delay reduces sear…
Nvidia has become the central authority in the AI industry, analogous to the Federal Reserve in finance, due to its dominant position in supplying the hardware necessary for AI computation. The company's GPUs are essenti…
Apple's macOS 27 ships with a pre-installed CLI tool called 'fm' that includes 'fm serve', which turns a Mac into a local OpenAI-compatible inference server for free, requiring no API key or cloud billing. The tool suppo…