AI Infrastructure news and analysis on Web Pulse: 34328 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
Bonsai 2 27B, a 27-billion-parameter ternary language model built on the Qwen3.8-27B hybrid-attention architecture, ships at 8.60GB total through the MLX 2-bit pack — a 7.67GB ternary language model plus a 0.92GB unquant…
The long-term AI competition may hinge less on which country builds the most capable frontier models than on two competing delivery models — centralized cloud intelligence versus intelligence embedded at the edge, accord…
A global fintech moved its AI coding-assistant workload to GLM-5.2 served through Together's Dedicated Model Inference, giving its engineers self-service endpoint scaling instead of capacity requests routed through Toget…
Local AI and private AI are not synonyms, according to a post from LM-Kit, which argues that running a model on hardware you control says nothing about where documents, embeddings, logs, or retrieval data travel. The pos…
Vercel added GLM 5.3 FlashX, Z.ai's multimodal coding model, to its AI Gateway, offering inference at roughly 200 tokens per second. The high-speed serving option is aimed at coding agents, tool loops, and interactive ap…
AMD published a white paper evaluating its 6th Gen EPYC 9006 "Venice" server CPUs across general-purpose, enterprise, cloud-native, AI and high-performance computing workloads, claiming the AMD EPYC 9996 delivers 1.2x th…
Hacktron reported a remote code execution vulnerability in August 2026 that traced to libheif, an AVIF image decoder used by Next.js, ImageMagick, WordPress, and sharp, rather than to Next.js itself. Vercel applied a pla…
Inco AI released Splash, an open-source inference engine for running Qwen3.6-35B-A3B and Qwen3.8-27B locally on Apple silicon, which in Inco's tests on a 48 GB M5 Pro delivered 74 tokens per second on short prompts and 5…
Neither "local AI" nor "private AI" has a formal definition, and no standard defines either term, according to an analysis that cites the NIST AI Risk Management Framework released January 26, 2023, which treats privacy …
Invisible Technologies is hiring a remote Software Engineer, Agentic AI based in Brazil with 2+ years of experience and equity compensation, but the posting does not publish a salary. The company reported $134M in revenu…
Cerebras Systems posted a Senior Staff AI Accelerator Performance Architect role in Sunnyvale, CA, with a salary range of $175,000 to $275,000 per year and a requirement of 7+ years of experience plus a Master's or PhD o…
Luminal co-founder and CEO Joe Fioti and founding compiler engineer Austin Glover will present how equality saturation drives Luminal's production tensor compiler at a talk scheduled for 2026-09-17. Luminal uses equality…
Marvell is pressing GlobalFoundries to ramp up wafer production to support its supply needs, according to The Register. The report notes Marvell could be the next trillion-dollar company but will need supply-side help to…
OpenRouter is hosting Pareto, a multimodal composite model built for research, coding, and agentic workflows, at $2.50 per million input tokens and $7.50 per million output tokens, with cache reads priced at $0.25 per mi…
A developer launched Scry, a search service built on a 500 TB NVMe internet index stored in ClickHouse that lets users run arbitrary read-only SQL and some Datalog queries, with congestion-based micro-auction pricing to …
Asus and Poesis.ai are reportedly partnering to build autonomous trading agents that combine Asus hardware with Poesis's AI-driven financial modeling and Nvidia technology, according to Crypto Briefing. Poesis.ai, founde…
AWS launched its R9g and R9gd memory-optimized EC2 instances on August 31, the first tier of the Graviton5 family to reach general availability. AWS claims the new chips deliver 25% more compute per vCPU, DDR5-8800 memor…
Prime Intellect is hiring a Member of Technical Staff for Storage Infrastructure at $150,000–$300,000 per year plus equity, on-site in San Francisco, requiring 3+ years of production distributed storage experience. The r…
At the AWS Global Meeting on September 17, AWS Chief AI and Technology Officer Matt Wood said AWS is building foundational AI tools the same way it built cloud computing, and Odyssey CEO Oliver Cameron unveiled Odyssey-3…
PrismML released Ternary Bonsai 2 27B on September 17, an open-weight model built on Alibaba's Qwen3.8 27B that compresses 27.8 billion parameters to roughly 1.76 bits per weight and a 5.9-gigabyte file size, more than n…