After months of teasers, Microsoft officially gave the Surface Laptop Ultra the two things it needed: a release date and a price. You’ll like one and hate the other. Let’s get into everything Microsoft announced at its Oct. 7 Windows and Surface event.

The Surface Laptop Ultra is anything but a surprise. Microsoft first unveiled the device this spring at Computex, and since then, it popped up at Microsoft’s Build software conference and IFA. The 15-inch laptop is powered by Nvidia’s RTX Spark chip and platform. This computer is a workhorse, meant for developers, creators and others who need a lot of horsepower to power through their work. It also comes with a magnetic USB-C charging cord, new for Microsoft.

The Surface Laptop Ultra is available for preorder now, and devices will begin shipping on Friday, Oct. 16.

Pricing for the Ultra is going to depend a lot on how much memory you need in your laptop. The base model with 24GB of RAM, a 512GB SSD and a lower-end RTX Spark N1X processor (the version with the 18-core CPU and 5,120-core GPU) starts at $2,600. But upgrade to the 32GB model, with 1TB SSD, and you’ll pay $3,300. The most expensive model, with 128GB of RAM, goes all the way up to $5,900. And that’s before shipping and taxes.

If the pricing didn’t give it away already, this laptop is not an affordable, everyday option. It’s not meant to be: It’s the flagship member of the Surface lineup, aimed at hardcore workers who need or want a device that can handle a lot of AI compute-intensive tasks or a lightweight system that can handle GPU-intensive work.

“We built it for developers and creators pushing the limits of performance, who want portability with exceptional craft,” Pavan Davuluri, Microsoft’s Windows chief, said at the event on Wednesday.

A big part of Microsoft’s sales pitch is the ability to run AI models and agents locally on device. There are some advantages to using local AI models. You keep all the information more secure within your hardware — no outsourcing to the cloud — and after the initial setup costs, it can be cheaper to run. Both of those points are appealing to businesses that need to keep their data private and avoid hefty bills from “AI tokenmaxxing” employees (people who use a lot of AI).

To truly get the most use out of local AI, you need a device that can handle your generative and agentic tasks without eating more memory than you’ve got or taking forever. That’s where Nvidia comes in with its RTX Spark. It’s a new Arm-based system-on-chip platform that brings Nvidia’s Blackwell GPU architecture — sometimes used in AI data centers — to Windows laptops.

We’ll have to wait and see just how well the Surface Laptop Ultra handles in our full testing. For now, check out CNET’s best laptops of 2026.

Katelyn Chedraoui

Reporter 2

Katelyn is a reporter with CNET covering artificial intelligence, including chatbots, image and video generators. Her work explores how new AI technology is infiltrating our lives, shaping the content we consume on social media and affecting the people behind the screens. She graduated from the University of North Carolina at Chapel Hill with a degree in media and journalism. You can reach her at [email protected]. See full bio