AI Infrastructure news and analysis on Web Pulse: 34123 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
A developer has open-sourced the Garza Global Graviton (GGG) Sovereign Edge Daemon, a self-contained local background daemon that runs an OpenAI-compatible loopback proxy endpoint at http://127.0.0.1:11834. The project c…
Garza Global Graviton LLC released v1.4.0 of its GGG sovereign proxy daemon, a local-first, air-gapped AI tool that runs entirely on local hardware with zero cloud telemetry and rejects any Ollama bridge endpoint other t…
Abacus.AI launched the Smaug line of three open-weight models on September 10, fine-tuned for long-running agentic loops and deployable inside a customer's own VPC at open-source rates. Abacus.AI claims the models are 10…
Huawei is accelerating development of an indigenous AI-computing platform intended to reduce China's reliance on U.S.-supplied accelerators, networking technologies, and associated software ecosystems, according to a Com…
A developer built Alca, a read-only, anonymous site that compares AI model access, pricing, free quotas and rate limits across providers, with a dated reset log linking each announced quota or limit change back to its or…
Meta plans to deploy its third-generation custom AI processor, MTIA 450, code-named Arke, in data centers in the first half of 2027, targeting lower cost per dollar and per watt for AI inference, according to Bloomberg. …
A developer documented five defects in a PNG validation service, including a decompression routine that inflated a 5 MB RGBA image to roughly 40 MB of memory per request, an unused QUEUE_MAX_SIZE constant that left inges…
Prox (Aiproxy), a closed AI API relay based on New-API that admits users only via OAuth from the external forum dc.hhhl.cc, was marked down in its latest status probe 52 minutes ago with 6001ms latency, after showing up …
Modeloc, an evaluation and compute-sharing platform for LLM relays, was flagged as high risk by Uprouter, which cited a decentralized compute-pool model that requires trust in unknown peers for data handling. Uprouter re…
Honor has revealed the Tiangong AXB35, an AI workstation platform for enterprise use in a 2 L small-form-factor chassis measuring 70 × 192 × 203 mm, offered in Standard, Pro, and Ultra variants. The AXB35 Pro ships with …
Honor has revealed the Tiangong AXB35, a 2-liter small-form-factor AI workstation for enterprise use that ships in three variants: a Standard model with Intel's Core Ultra X7 358H CPU and Arc B390 iGPU supporting up to 9…
Agentgit launched as a Git hosting service for AI agents that requires no account, token, or key, creating a repository automatically on the first push to a chosen name. The service, which uses a Signer List written to r…
An 8-GPU tensor-parallel (TP8) rig should prioritize PCIe topology over CPU choice, according to guidance recommending that GPUs communicate directly through a Gen5 switch rather than routing peer-to-peer traffic through…
Venice.ai, the privacy-focused AI platform founded by Erik Voorhees, has been listed as a service provider in Circle's Agent Marketplace, letting AI agents pay for its image generation, video creation, and library of ove…
A security analysis found that LiteLLM, an AI gateway tied to CVE-2026-59822 (CVSS 8.8), had 34,412 internet-facing deployments, while workflow orchestration platform Kestra, affected by CVE-2026-49869 (CVSS 10.0), had 1…
A user running llama.cpp in tensor split-mode on an ASUS ProArt X870E motherboard with two AMD R9700 GPUs reports that system suspend hangs instead of entering sleep mode, requiring a power cycle, even when no AI workloa…
WSO2 released Agent Manager as generally available this week, an Apache 2.0 open-source control plane that governs AI agents across frameworks, models, and deployment environments. The product gives every agent a verifia…
A forum post speculates that the 12V-3 connector, an alternative to the 12V-2 connector, is intended for AI-server GPUs and advises readers to stop buying any hardware that uses it. The post notes that datacenter equipme…
Livio_dev released Celestium Intel Cleaner V4 AUTO Background, a Windows optimization utility for Intel Arc, Iris Xe, and UHD GPUs that performs automatic background cleaning, shader cache reset, VRAM monitoring, and gam…
A developer deployed Google's Gemma 4 E2B to a single Tesla T4 GPU on a Compute Engine VM using vLLM 0.29.0, finding that the QAT int4-weight checkpoint decodes at 72.31 tok/s per stream versus 40.44 tok/s for bf16, a 1.…