AI Infrastructure news and analysis on Web Pulse: 35548 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
NVIDIA CUDA and TensorRT can significantly accelerate Physical AI workloads on NVIDIA hardware, according to a developer from v-modal. The developer explains that optimizing the complete data path, rather than focusing o…
A new Axios report reveals that China-linked bots are amplifying America's data center revolt, boosting local opposition to AI infrastructure projects. The report highlights how coordinated online activity is shaping pub…
Firecrawl has released pdf-inspector, an open-source library that classifies PDFs as text-based, scanned, image-based, or mixed, and routes only the pages that need OCR to the OCR engine. The tool, available in Python, N…
A new AI model with 125 billion total parameters but only 6 billion active per token has sparked curiosity about its architecture. The model uses a Mixture-of-Experts (MoE) design, where a router activates only a subset …
Researchers from UC Berkeley and MIT released FreeToken, an open-source inference engine that enables frontier Mixture-of-Experts (MoE) models to run on consumer hardware via dynamic co-execution, achieving 3–4x faster d…
Binance introduced Agent OS, a framework announced on August 27 to connect AI applications with digital-asset financial infrastructure, providing structured access to trading, payments, and other crypto services while ma…
X said Thursday night that roughly 200 accounts from a suspected Chinese bot farm have tried to influence Americans to oppose AI data centers on social media, part of a broader probe that identified 200,000 suspected Chi…
A computer science student at Vardhaman College of Engineering completed the Google Cloud Gen AI Academy APAC Edition Cohort 3, building a system of AI agents for a coffee shop business deployed on Cloud Run. The project…
A forum user comparing CUDA and ROCm for local AI inference reports that ROCm issues have diminished and is considering an all-AMD build with dual R9700 GPUs by 2027, citing AMD's lower cost. Another user notes AMD's ROC…
Mcpindex.ai published version 1.35.32 of com.astranl/mcp, a hosted remote MCP server for AI agents that handles physical errands with funds held until agreed proof, on 2026-08-29. The index's semantic screen found no man…
Io.github.jamhimself/robinx-mcp, a Robinhood Chain (4663) caller accuracy and token intel server paid via x402, was published as version 0.6.1 to the official MCP registry on 2026-08-29, distributed as robinx-mcp on npm …
OpenRouter retired and delisted monitoring of the deepseek/deepseek-v3.2 endpoint at alibaba/fp8 on 2026-09-04 after it dropped out of the B3IT selection for 30 days, according to the tracking page. The endpoint returned…
An engineer explains the Model Context Protocol (MCP), a universal standard that enables AI agents to connect to external tools, databases, and APIs, likening it to USB for AI. The protocol uses a client-server architect…
Substructure AI released Subs, a pre-1.0 cloud-native agent harness that runs an unprivileged agent loop with no system access, using MCP servers for tools and supporting local or client-server operation. The tool manage…
ChimeraForge, a local-first LLM deployment planner for GPU/VRAM sizing, cost, and latency with provenance, was published as v0.30.9 to the official MCP registry on 2026-08-29 and is distributed as chimeraforge on PyPI. T…
Io.github.l33tdawg/sage, a persistent, consensus-validated institutional memory tool for AI agents, released version 11.19.9 to the official MCP registry on 2026-08-29, distributed as Docker image ghcr.io/l33tdawg/sage:1…
ServiceNow disclosed three CVSS 10.0 vulnerabilities in its AI Platform on August 27, 2025, including code injection (CVE-2026-18885), access control bypass (CVE-2026-18886), and SQL injection (CVE-2026-74820), all remot…
TabiToken, an AI API gateway and model routing service, is offering developers $120 in free signup credits through a referral link, along with a $20 reward per successful referral. The service supports multiple AI models…
A developer team at an unnamed company corrected a widely cited measurement of Claude Code subagent spawn costs, revealing that the actual cost is approximately 54,000 tokens, not the previously reported 436,000 tokens. …
A developer built Attest, an ADK-based agent that audits what AI models say about SEC-registered advisers against their Form ADV filings, for the All Things Agentic Hackathon. The agent runs a monthly battery of factual …