Nvidia targets $200 billion AI PC market with Microsoft, Dell and HP

Nvidia unveiled RTX Spark, a new processor for personal computers that can run AI agents safely on the local machine, without a cloud connection. It is an attempt to capture a potential $200 billion AI PC market.

RTX Spark and local execution

RTX Spark is a "superchip" delivering 1 petaflop of performance. It is designed specifically to run complex AI models directly on a PC. The processor is powerful enough to handle large language models locally, without network latency.

But the main point is security. Together with Microsoft, Nvidia developed dedicated sandboxes for AI agents: they run in an isolated environment and cannot accidentally damage the system, access the user's files without permission, or perform dangerous operations. This makes local execution practical for ordinary users.

Local processing offers two big advantages: privacy (data does not go to OpenAI or Google servers) and speed (no network request delays).

Manufacturers and release timing

The first PCs with RTX Spark will arrive in fall 2026 from leading manufacturers:

• ASUS
• Dell
• HP
• Lenovo
• Microsoft Surface
• MSI

Acer and Gigabyte will join later. Each manufacturer will offer its own memory and storage configurations, but all systems will run on RTX Spark and an updated version of Windows with built-in support for AI agents.

The app ecosystem

The new PCs support more than 1000 existing games and applications out of the box. More than 100 software vendors have already announced their readiness:

• Adobe (Photoshop, Premiere Pro, After Effects)
• Blender (3D modeling)
• Riot Games (League of Legends, Valorant)
• Autodesk (design engineering)
• Figma (design)

Software makers are preparing RTX Spark optimizations so agents can assist users inside their applications. For example, in Photoshop an agent can help with photo editing, and in Figma with design versioning.

Strategic vision

Nvidia CEO Jensen Huang describes it as a revolution: "With RTX Spark and Windows, you just talk, and the computer does the work." Instead of clicking icons, the user gives the agent an instruction in natural language.

Nvidia sees a potential $200 billion market here. The company has already earned $20 billion on its Vera server processor, testing this technology in the enterprise sector.

History remembers a failure: in 2013, Nvidia tried to push Intel out with ARM processors for Windows, an investment that cost Microsoft $900 million. This time the stakes are higher: RTX Spark is significantly more powerful, the technology is more mature, and the industry agrees that AI agents are the next leap.

What this means

If Nvidia has managed to combine AI agents with reliability and simplicity, it could permanently change how people work with computers. Local processing instead of the cloud, privacy instead of data leaks, agents instead of apps.