Microsoft event debuts new AI-friendly hardware and Windows changes

AI-rewritten: This is a summary of an article from Ars Technica, rewritten by AI (Qwen, running locally) to make it easier to read. The facts come from the original article – read it for the full story.

Ars Technica •
Nick Indge
• October 8, 2026

Microsoft held its first live event in two years to announce the new Surface Laptop Ultra, which will be the company’s first device using the Nvidia RTX Spark SoC. Originally teased in May 2026, the laptop starts at $2,599 and features either a 5120-core or 6144-core Blackwell GPU with up to 128GB of LPDDR5x unified memory. It is available for pre-order now and will begin shipping on October 16. The device uses unified memory allocation instead of traditional physical VRAM limits, allowing it to handle modern gaming tasks and local AI development. During the presentation, Microsoft demonstrated that AAA titles like Gears of War: E-Day can run on the system without a dedicated gaming GPU.

On stage alongside Nvidia CEO Jensen Huang and Microsoft CEO Satya Nadella, President Pavan Davuluri also unveiled the Surface RTX Spark Dev Box priced at $5,999. This anodized aluminum box is designed for frontier developers and ships with a custom Windows 11 build optimized for AI. It includes 128GB of unified memory and an onboard Nvidia RTX GPU, claiming one petaflop of AI compute performance. The design resembles an Android TV box or game console rather than a traditional PC, featuring 1,000 chassis air vents on top as a nod to its processing power.

Microsoft also outlined significant updates for Windows 11 arriving later this year, including changes to the Copilot interface which will be divided into three tabs. The new update allows users to execute system-level tasks directly from the search bar without opening Control Panel. To address security concerns regarding AI agents accessing user data, Microsoft announced Microsoft Execution Containers (MXC). This framework acts as a policy layer that directs compute traffic to isolation backends like process containers or MicroVMs, ensuring agents operate within specific guardrails.

The company is promoting this approach as "Hybrid Intelligence," which moves away from cloud-backed AI toward local solutions. This allows users to allocate memory for specific applications and offers native support for local frameworks like llama.cpp and models from the Deepseek and Nvidia Nemotron families. While Microsoft is betting big on local AI, questions about security and user acceptance remain, especially given past AI agent mishaps. The Surface Laptop Ultra will hit shelves next week.

Source: Ars Technica •
Nick Indge
• October 8, 2026

Read the original article at Ars Technica →

Leave Comment

Your email address will not be published. Required fields are marked *