AI Infrastructure news and analysis on Web Pulse: 34481 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
Weekly token consumption on OpenRouter surged more than 25,000 percent since January 2025, rising from 0.5 trillion to 126.2 trillion tokens, according to The Decoder. The publication attributes the increase less to actu…
Ontos-AI released Knowhere 2.0, a document parsing and retrieval system that reconstructs document hierarchy for AI agents through dual Vision and Text tracks, with the full stack open-sourced on May 7, 2026. The Septemb…
The Cloud Native Computing Foundation announced on September 7, 2026 that Karmada, a multi-cluster and multi-cloud Kubernetes orchestration project, has graduated to CNCF's highest maturity tier, coinciding with its v1.1…
Vinod Bijlani, a Forbes Councils Member, argues in a Forbes Innovation piece that dollars per FLOP is becoming a critical AI cost metric, and that tech leaders need to understand how efficiently capital, energy and physi…
Huawei's Intelligent World 2035: From Vision to Action report forecasts that global annual token consumption will rise 100,000-fold by 2035, with AI-agent traffic accounting for more than 90% of the total. The report als…
Speculative decoding, an inference optimization introduced independently by Chen et al. at DeepMind and Leviathan et al. at Google in 2023, cuts LLM decoding latency without changing output quality by having a small draf…
IDCA research indicates that as much as 13% of US cloud usage comes from zombie workloads — abandoned jobs, instances, and storage volumes that keep consuming power, rack space, and budget — according to Roger Strukhoff,…
Runpod found that improving the data foundation, not the underlying large language model, was what fundamentally changed the quality of its conversational Slack data agent, according to the company's account of building …
Zhipu AI founder Tang Jie announced that the company's GLM-5.3-driven 'Infra Agent' delivered a 3.2x end-to-end throughput improvement on a production-grade inference system running on a cluster of more than 100,000 dome…
TypeSafe AI released Jev, a "System One" model that answers structured choice, score, and boolean questions instead of generating text, claiming 193.6x faster and 444.6x cheaper performance than frontier LLMs on agent ro…
A developer built Don Watch, a browser-based real-time flood digital twin of Sheffield's River Don that runs entirely client-side using WebGPU, three.js, and an on-device language model loaded via WebLLM. The tool recons…
Z.ai published a blog post titled "GLM Built Its Own Inference Infrastructure," detailing the GLM model family's in-house inference stack. The post drew 2 points and 0 comments on Hacker News.
Huawei launched the Atlas 960 SuperPoD computing cluster at its annual conference in Shanghai on Thursday, a successor to the Atlas 950 SuperPod it released late last year, and said it plans to launch the Ascend 970 and …
Sysdig researchers coined the term LLMjacking in 2024 after observing attackers using stolen cloud credentials to access cloud-hosted LLM services, with one worst-case configuration involving unauthorized Claude usage th…
Telxius, the Telefónica-owned subsea cable business, is deploying Nokia's 800 Gb/s ICE-X coherent pluggable optics across its terrestrial infrastructure in Europe, the U.S., and Latin America to support transmissions ove…
Running inference on locally owned hardware is a hedge against frontier AI providers that can raise prices, deprecate models, throttle access, or cut off customers, according to a blog post by Herlein. The author argues …
Ripple released version 1.1 of its XRPL AI Starter Kit, adding support for the Machine Payments Protocol (MPP) co-authored by Stripe and Tempo, along with the Open Wallet Standard, according to a RippleX developer post. …
Imversion Technologies Pvt Ltd has published a production AI checklist covering reliability, latency, cost, security, UX, monitoring, testing, and launch controls for teams deploying AI features. The guidance stresses SL…
A developer documented a step-by-step process for installing Kubeflow on Amazon EKS, using eksctl to provision the cluster, VPC, subnets, and a managed node group in a single command. The guide notes that Kubeflow's cont…
A developer outlined a state-machine approach for handling timeouts in asynchronous AI image generation APIs, arguing that a lost HTTP response cannot be treated as a confirmed failure because the provider may have alrea…