#Nvidia

• Intel's 480GB D-Series GPU Beats Nvidia on Memory, Not Speed

  Intel's new D-series GPU carries 480GB of LPDDR5X, the memory in your laptop, and outguns Nvidia's 288GB Blackwell Ultra B300 on capacity. The catch: no price or launch date yet.

• Nvidia's Huang: Surface Laptop Ultra Took Four Years to Build

  Nvidia's Jensen Huang says the Microsoft Surface Laptop Ultra took four years and thousands of engineering years to build, calling it a work of love for local AI agents.

• Microsoft's Maia 300 Bets on Memory, Not Muscle, to Cut AI Costs

  Microsoft will make its own Maia 300 chip the primary engine for AI inference at multi-gigawatt scale in 2027. The bet is on memory bandwidth, not raw compute.

• RTX 5090 Vanishes From Nvidia's Store as AI Eats the GPU

  Nvidia's RTX 5090 is disappearing from its own store and retailer shelves, and a leaker says GB202 chips are going to AI cards. The RTX 5080 is now the flagship — and prices are…

• RTX Spark Debut: HP’s $3,200 OmniBook Ultra 16 Beats Snapdragon, Lags M5

  Nvidia’s RTX Spark platform makes its retail debut in HP’s OmniBook Ultra 16. Early benchmarks show strong multi-core and AI performance, but Apple’s M5 Pro and M5 Max still lead.

• Huang: OpenClaw Robotics Needs 'Zero Trust' Before It Ships

  Nvidia CEO Jensen Huang hails OpenClaw as a breakthrough for robotics, but warns agentic AI must run with zero trust and isolation before it ships.

• Nvidia: AI Data Centers Waste 40% of Power, Vertiv Deal Fixes It

  Nvidia claims AI data centers waste up to 40% of power on non-AI work. A new Vertiv partnership aims to fix cooling, failures, and grid bottlenecks.