AI Infrastructure news and analysis on Web Pulse: 23076 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
Microsoft is replacing costly AI models from OpenAI and Anthropic with its own MAI models in Excel and Outlook to save money, as tens of thousands of weekly prompts are now handled in-house. The shift comes amid rising t…
At GTC 2026, Nvidia CEO Jensen Huang declared the token the unit of production in the AI economy, presenting a formula linking revenue to tokens per watt and available gigawatts. Between 2026 and 2030, IT is reorganizing…
CoreWeave has deployed NVIDIA Vera Rubin NVL72 at rack scale, becoming the first cloud provider to validate the system. The company integrated training, continuous inference, observability, and autonomous improvement int…
SK Hynix projects $231 billion in revenue for 2026, up from $67 billion in 2025, driven by surging demand for high-bandwidth memory used in AI accelerators. The company dominates HBM shipments with over 60% market share,…
OpenAI has taken an equity stake in Thrive Holdings, a permanent capital vehicle launched by Thrive Capital, to acquire and rewire accounting and IT firms with AI. The deal embeds OpenAI research teams inside portfolio c…
DeepSeek is developing a custom AI chip to bypass U.S. export restrictions and reduce its reliance on Nvidia, according to Yahoo Finance's Dan Howley. The move aims to secure supply amid geopolitical tensions.
Llama 3.1 8B Instruct requires 32 GB of memory at 128K context due to a 16 GB KV cache, not just the 15 GB of weights, contradicting earlier reports that it fits in 24 GB. Users can reduce memory by quantizing the cache,…
Abhijay Jain, maintainer of the open-source AI agent goose, detailed at the MCP Dev Summit Bengaluru how the platform evolved from a single-user bot into a distributed ecosystem for collaborative software development. Th…
High Bandwidth Flash (HBF), a NAND-based memory stack with read bandwidth comparable to HBM4 but 10x the capacity, is set to sample from Sandisk in late 2026 and appear in AI inference devices by early 2027. By stacking …
AWS engineers demonstrated troubleshooting agentic AI using OpenTelemetry and OpenSearch to gain insights into system performance and data before and during production.
Wall Street analysts issued bullish calls on SpaceX, with an average price target of $236 per share, representing 47% growth from Monday's close. More than a dozen brokers including Morgan Stanley, JPMorgan and Goldman S…
OpenComputer launched a Slack integration that turns channel threads into durable AI agent sessions, allowing teams to start and steer agents via @mentions. The product is built on persistent virtual machines that mainta…
Amazon secured $62 billion in investor demand for its $25 billion bond sale, earmarked for AI infrastructure including data centers and proprietary chips. The oversubscription allows Amazon to negotiate tighter spreads, …
Wall Street banks issued bullish research notes on SpaceX's stock, forecasting prices above $200, but shares trade around $152 as investors remain cautious. Analysts highlight SpaceX's dominance in space transportation a…
Japanese Prime Minister Sanae Takaichi unveiled a ¥370 trillion ($2.3 trillion) growth strategy targeting 17 strategic sectors by 2040, with major investments in AI and semiconductors. The plan aims to revive Japan's chi…
Anthropic signed a $19 billion, 20-year lease with TeraWulf for 401 megawatts of data center capacity in Kentucky to address chronic Claude API outages, but relief won't come until H2 2027 at the earliest. The deal, whic…
Presiden Prabowo Subianto dan Perdana Menteri India Narendra Modi mendorong kerja sama praktis antara Indonesia dan India di berbagai bidang, termasuk perdagangan, pangan, kesehatan, pertahanan, maritim, kecerdasan buata…
Decagon CEO Jesse Zhang argues that open-source AI models are not hurting frontier labs like Anthropic, as they serve different lifecycle phases—frontier models for discovery and open-source for production. Data from Ver…
Microsoft announced it will cut 4,800 jobs globally, including a 20% reduction at its Xbox division, as the company shifts investment toward artificial intelligence. Xbox CEO Asha Sharma cited declining revenue and thin …
Momentic, an AI-driven software testing platform, migrated its caching system from PostgreSQL to ClickHouse to handle over 2 million queries per day across 20 billion entries with 250 ms average latency. The move address…