AI Infrastructure news and analysis on Web Pulse: 34756 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
An analysis of 2,218 instrumented production calls over 30 days found that phi4:14b delivered only 25.3 tokens per second to callers versus its 124.7 tok/s Ollama-reported decode speed, losing 79.7% of throughput with a …
A user running hybrid GPU+CPU inference on 4x Nvidia V100 GPUs reported beating an Nvidia RTX 5090 in token generation (TG) for Qwen 3.8 27B and outperforming two DGX Spark systems with Qwen 3.8 Flash Next, according to …
Databricks' marketing department uses data 3X more often in decisions after deploying Marge, a conversational analytics assistant built on Genie Agents and grounded in a governed Marketing Lakehouse, according to Liz Dob…
E2B sandboxes can be monitored and traced with OpenTelemetry by instrumenting the application that creates them, according to a technical guide from SigNoz. The guide details five telemetry paths — application instrument…
Google's Firebase launched spend caps in Public Preview on Sept. 14, 2026, pausing Firebase AI Logic, Cloud Functions for Firebase and Firebase App Hosting at 100% of a project's budget, according to a Firebase Blog post…
LM-Kit released LM-Kit One, a self-hosted AI server that bundles models, agents, search, and document intelligence into a single product with signed packages now available for Windows, Linux, and macOS. The company said …
OpenFaaS Ltd purchased 4 Nvidia DGX Spark systems, connecting the first two with a high-speed Connect-X cable, to run local AI models including Qwen 3.5-3.8 27B and larger models such as DeepSeek and GLM for customer sup…
ARIMLABS now runs roughly 500,000 agent executions per day on Islo's cloud computers, up from a small pilot, according to CEO Mykyta Mudryi. The company moved its Harbor-based long-horizon agent evaluations off local mac…
A two-stage agentic pipeline combining GLM-4.7-Flash, a 30B tool-calling model, with Poro2, a 70-billion-parameter Finnish-language model, enriches medical terminology via Model Context Protocol tools querying a medical …
Stripe data shows AI startups on its network faced up to 4.3x higher attempted transaction fraud rates than startups overall in Q3 2025, falling to 2.6x by Q1 2026, while attempted multi-account abuse at AI subscription …
Edge0 released Edge0-35B-A3B-preview, a 35-billion-parameter mixture-of-experts language model built on Qwen3.6-35B-A3B that runs with under 3 GiB of active memory by streaming experts from SSD instead of loading all wei…
An independent research group in Austria has developed two quantization techniques, GSQ and RCO, that compress Z.ai's 320-billion-parameter GLM 5.3 Flash vision-language model from roughly 320GB at full precision to unde…
Researchers published VQ-bench, an open-source framework that unifies vector quantization algorithms by decomposing them into 7 common conceptual primitives that can be composed arbitrarily. The authors re-expressed 25 c…
Anthropic's continuous integration jobs grew 25 times in six months as coding agents wrote 80 percent of the company's code, an Anthropic engineer reported on September 14, 2026, with each interim fix lasting less time t…
Senate Majority Leader John Thune (R-S.D.) said Congress has a role in regulating AI and that "guardrails there make sense to me" for frontier models, putting him at odds with President Trump, who posted on Truth Social …
Databricks' Lakebase Postgres is a serverless managed Postgres service that automates patching, scaling, failover, and backups, with autoscaling that Databricks says enables 5x faster Postgres writes than standard Postgr…
Vstorm released AgenticOS, an open-source, self-hosted control plane for building, running, and governing company AI agents on the user's own Postgres and Docker infrastructure under their own domain. The tool installs v…
A developer who spent a month building tools that combine deterministic scripts with LLM-driven agents found that the economics of probabilistic inference, not capability, became the limiting factor as systems grew more …
Nvidia CEO Jensen Huang put President Donald Trump on speakerphone during a live appearance at the All-In Summit in Los Angeles on Monday, telling Trump "We're not going to let that happen, sir" after Trump called effort…
DeepSeek's chat app and API went down for hundreds of users on September 14, 2026, while the company's own status page at status.deepseek.com showed every listed service as operational with no incident banner or acknowle…