How Windows is being optimized for Nvidia’s RTX Spark laptops

PUBLISHER ARTICLE PREVIEW

From the original article

Microsoft and Nvidia held a launch party for the first wave of RTX Spark laptops last week in San Francisco, where a new wave of AI PCs took center stage amid a packed crowd and a lively hands-on area. A day after the media event, I was invited to a quieter Nvidia event to hear a lot more about the new RTX Spark computing platform.

While Nvidia’s new laptops might look like regular Windows machines, the chips inside are Arm-based, instead of the traditional x64 architecture used by AMD and Intel. Combined with a powerful GPU, RTX Spark is designed for the future of AI, and it means a lot of work has gone into getting these laptops ready for launch.

One of the biggest challenges was getting Windows ready for AI agents and local AI models. The big selling point of RTX Spark is the combination of Arm CPU, Blackwell GPU, and unified memory to enable powerful AI models to run on a laptop, so the experience of setting up an AI agent needs to be quick and easy.

“When users get into Windows for the first time, they’ll see a new Get Started experience that will launch on an RTX Spark laptop that allows them to find the agent they want, get started with that agent, install it, and get to that agent in only two clicks,” says Peter Dawoud, principal program manager at Microsoft.

Agents can be installed and updated through the Windows Store, so you don’t have to worry about finding updates or figuring out what dependencies each one needs. “We’ll continue to develop it and add more features so that agents can offer a whole new bunch of experiences coming through Windows,” Dawoud says.

Hermes, Perplexity, and OpenClaw are the first three officially supported agents in Windows at the launch of RTX Spark, but Microsoft is working on bringing others on board. That will presumably eventually include Muse, which Meta announced last week is coming to Windows at some point in the future.

Getting started with Hermes was mostly a two-click experience in my testing, but I did run into trouble getting OpenClaw set up properly. The app kept complaining that a local gateway hadn’t been installed properly, and it took an uninstall and reinstall to get everything working fully.

I’ve been using OpenClaw for a variety of tasks over the past week, but it’s going to take a lot more testing to really understand the value of this local AI support beyond obvious tasks like coding.