AI Infrastructure news and analysis on Web Pulse: 34514 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
OpenAI will shut down the Sora API on September 24, 2026, after the product reportedly burned close to $1 million a day against just over $2 million in total lifetime revenue, according to a cited OpenAI help article. Th…
GNOME 51 shipped on September 16, six months after GNOME 50, removing support for legacy NVIDIA driver interfaces so the desktop communicates with the driver exclusively through modern, standard graphics interfaces. The …
Wiz Research found that a one-character Bearer token ("Bearer a") bypassed authentication in LiteLLM's Model Context Protocol gateway, tracked as CVE-2026-59822, after 90 days of honeypots mimicking AI infrastructure inc…
OmniRoute, an MIT-licensed self-hosted AI gateway from developer diegosouzapw, routes coding-agent sessions across 352 providers — including 150+ free ones — from a single OpenAI-compatible endpoint at localhost:20128 an…
A tutorial published on Modal Cloud's platform shows how to fine-tune the 135M-parameter HuggingFaceTB/SmolLM2-135M-Instruct model on the b-mc2/sql-create-context dataset using a single YAML configuration file and no Pyt…
An independent developer released Jev Logs, an MIT-licensed open-source layer that scores and routes OpenTelemetry logs with TypeSafe's Jev model before they reach a more expensive reasoning LLM. The tool assigns each re…
A developer scanned the official MCP registry's 32,380 unique servers and found that of the 2,967 advertising OAuth, only 8% meet all of the July 2026 authorization spec requirements. While older OAuth mechanisms pass at…
TypeSafe AI launched Jev, a "System One Model" that returns typed, structured outputs with calibrated confidence scores instead of generating text, priced at $0.042 per million tokens and trained with a method TypeSafe c…
SK hynix showcased its next-generation AI memory portfolio under the slogan 'New Spectrum' at the AI Infra Summit 2026, held September 15th-17th in Santa Clara, California, where the company doubled the size of its booth…
The House passed the Ratepayer Protection Act by a 417-3 vote on September 16, sending to the Senate a bill that would require data centers and hyperscalers drawing more than 100 megawatts to cover the full incremental c…
KT said Thursday it has joined a Ministry of Science and ICT-led research project to develop design and construction solutions for "dark factories" — fully automated, unmanned plants that run around the clock using artif…
Huawei Technologies rotating chairman David Wang announced on Thursday that the company will release two new AI chips in 2027 — the Ascend 960DT in the first quarter and the Ascend 960PR in the third quarter — built arou…
The U.S. House of Representatives passed H.R. 9340, the Ratepayer Protection Act, by a 417-3 vote on Wednesday, requiring state utility regulators to consider charging large-load electricity consumers at least 100 megawa…
A developer's analysis of LLM agent orchestration found that idle parent agents on Codex burn tokens polling for worker completion because the runtime lacks event-driven wake-ups, with one run issuing 89 waits that consu…
Meta Vice President of Infrastructure Manish Kadakia said at the AI Infra Summit in Santa Clara on Sept. 16, 2026, that all four core data center systems — servers, networks, electricity and cooling — are changing simult…
A developer worked around DynamoDB's TopK=100 cap on vector search results by sharding the vector index partition key into five hash-derived groups and querying them in parallel, then merging and deduplicating the per-sh…
A study measuring 54 configurations of Qwen2.5-7B-Instruct on vLLM 0.12 across L4, A100, and H100 GPUs found that 18 of 36 configurations reach the cost, quality, and latency Pareto frontier, with combined optimization m…
A training-free routing policy called the Tri-Metric Router achieved 0% out-of-memory failures and 88.5 ± 4.4% oracle alignment on out-of-distribution holdouts for retrieval-augmented generation on commodity GPUs, accord…
Researchers introduced Fathom, a per-query read-depth key scan for sparse decoding over offloaded KV caches, which at one million tokens on Qwen3-8B runs a decode step 1.67x faster in GPU time than the 136-bit scans of D…
Researchers released Zing-0.5, a 5B autoregressive world model that supports joint keyboard and online text control for real-time playable generated worlds, according to the arXiv paper 2609.17909v1. Zing-0.5 scores 81.0…