Nvidia Enters the PC Market with RTX Spark to Run AI Agents Locally

Microsoft and Nvidia are building a platform that allows autonomous AI agents to securely navigate and act within Windows.

Topics

  • [Image: PankajKirdatt/MITSMR Middle East]

    AI hardware company Nvidia is expanding beyond its data center stronghold with a new chip designed to bring advanced AI capabilities directly to personal computers, as the company bets that AI agents will increasingly run on users’ devices rather than in the cloud.

    The company on Monday launched RTX Spark, a new system-on-chip for laptops and desktops, co-developed with Taiwan’s MediaTek, at the Computex technology conference in Taipei. Devices powered by the chip are expected to be launched later this year from manufacturers including Dell, HP, Lenovo, Asus, MSI, and Microsoft Surface.

    Nvidia announced the platform alongside Microsoft, with the two companies introducing a joint architecture designed to support AI agents that can operate locally on Windows PCs. The partnership combines new Windows security primitives with Nvidia’s OpenShell runtime, allowing users to control which information AI agents can access and when data should remain on-device rather than be sent to cloud-based services.

    The move comes as technology companies increasingly shift their focus from AI chatbots to AI agents capable of carrying out tasks autonomously. Unlike traditional assistants that primarily respond to prompts, agents can navigate applications, manage files, and complete multi-step workflows on behalf of users.

    Nvidia said it has been working with Microsoft to enable agents that can interact with Windows computers in much the same way as a human user, operating a keyboard and mouse to perform actions across applications.

    “The PC is being reinvented,” said Jensen Huang, Nvidia’s founder and chief executive officer. “With RTX Spark and Microsoft Windows, you ask — and the PC does the work.”

    The hardware underpinning the platform combines Nvidia’s Blackwell RTX graphics processor with a 20-core Grace CPU and supports up to 128GB of unified memory. Nvidia said the system can run large language models with up to 120 billion parameters and context windows of up to one million tokens directly on a personal computer.

    Microsoft plans to integrate RTX Spark-powered agent capabilities into the Windows taskbar, while Nvidia said the same architecture will extend to enterprise deployments through its DGX Station systems.

    By enabling advanced AI workloads to run locally, Nvidia is targeting the growing demand for greater privacy, security, and control over personal data.

    Whether consumers will embrace AI-focused PCs remains unclear. Previous efforts to create a new category of AI-powered computers, including Microsoft’s Copilot+ PCs, have struggled to gain widespread adoption. Nvidia, however, is betting that increasingly capable AI agents will create demand for devices designed specifically to run them.

    Topics

    More Like This

    You must to post a comment.

    First time here? : Comment on articles and get access to many more articles.