Alibaba unveils the Zhenwu M890 AI chip with triple the performance of its predecessor, announces a multi-year chip roadmap, and launches a new AI model and server system, backed by a massive 380 billion yuan infrastructure investment as China races to reduce reliance on U.S. chips.
NVIDIA shatters records with $81.6 billion in Q1 fiscal 2027 revenue, surging 85% year-over-year, as explosive AI demand drives Data Center sales to $75.2 billion while the company projects an even bigger $91 billion quarter ahead.
Jeff Bezos calls space-based data centers 'very realistic' but pushes back on Elon Musk's aggressive 2-3 year timeline, as Blue Origin targets a 2027 launch for 'Project Sunrise,' a bold plan to deploy 51,600 data center satellites into low Earth orbit, while space stocks surge amid SpaceX's anticipated IPO at …
OpenAI launches a 'Guaranteed Capacity' program letting customers lock in one-to-three-year AI compute commitments at increasing discounts, with CEO Sam Altman calling it a 'big win-win' as the company targets $600 billion in total compute spend by 2030 and eyes a potential IPO.
Blackstone and Google are launching a $5 billion AI computing joint venture powered by Google's Tensor Processing Units, with 500 MW of U.S.-based data center capacity expected online by 2027 and Google veteran Benjamin Treynor Sloss named as CEO.
NVIDIA is hand-delivering its first custom CPU, Vera, to top AI labs including OpenAI, Anthropic, and SpaceXAI, featuring 88 custom Olympus cores and 50% faster per-core performance designed for agentic AI workloads, with Oracle Cloud Infrastructure set to deploy hundreds of thousands of units at hyperscale starting in 2026.
Google I/O 2026 kicks off with massive announcements including Gemini 4, AI-powered glasses built with Warby Parker featuring maps and real-time translation, and Project Aura, a spatial computing device developed with XREAL, signaling Google's boldest push yet into AI hardware and agents.
A new open-source CLI tool called 'whichllm' instantly scans your hardware and recommends the best local AI models you can actually run, ranked by real benchmark data from top leaderboards with a single command.
Unsloth Studio launches in beta, enabling users to train over 500 open-source AI models locally on Windows, Linux, and macOS up to 2x faster while using 70% less VRAM, with features including reinforcement learning, multi-GPU support, and API inference compatibility with tools like Claude Code.
A new 1B parameter AI model called HRM-Text is making foundation model pretraining accessible to nearly anyone, requiring up to 900x less data and 600x less compute than traditional methods, bringing full pretraining costs down to roughly $1000 using a hierarchical recurrent architecture with strong benchmark performance across GSM8k, MATH, …