Nvidia Sets Sights on PC Market With New AI-Focused Chip Backed by Microsoft, Dell, and HP

Nvidia kicked off Computex in Taipei with a literal spark on Sunday, unveiling a new PC processor called the RTX Spark. The company is calling it a “superchip,” and it has lined up a roster of major PC makers ready to build machines around it.

The chip, capable of 1 petaflop of performance, is designed to run AI agents such as OpenClaw or Hermes Agent securely. Systems powered by the RTX Spark will ship this fall from ASUS, Dell, HP, Lenovo, Microsoft Surface, and MSI, with Acer and Gigabyte models coming later.

These PCs include secure sandboxes developed with Microsoft to keep agents isolated, along with enough CPU, GPU, RAM, and Nvidia’s CUDA software to run large language models locally. Nvidia says the RTX technology will speed up AI tasks, improve image quality, and support AI features in more than 1,000 games and apps.

Nvidia is pitching the chip to creators making AI content and to gamers. More than 100 Windows software makers, including Adobe, Blender, ComfyUI, Riot Games, and Xbox, have signed on.

But CEO Jensen Huang has bigger plans. He wants to move past launching apps, clicking, and typing. “With RTX Spark and Microsoft Windows, you ask — and the PC does the work,” he said in a press release.

Last month, Huang told investors he sees a $200 billion market for Nvidia in selling CPUs for AI, not just GPUs. He noted that the company has already sold $20 billion worth of its Vera server CPU, released earlier this year.

Nvidia has tried ARM-based Windows PCs before. In 2013, Microsoft wrote off $900 million on the Surface RT. But this time, the chip is far more powerful. Microsoft is calling its RTX Spark model the Surface Laptop Ultra, billing it as “the most powerful Surface Laptop ever built.”

Pricing remains unclear. These systems resemble the DGX Spark mini-computer Nvidia sells to developers for about $4,800. It remains to be seen whether they will compete with the affordable Mac Mini or sit at the high end of the market.

If Nvidia has figured out how to bring AI agents to the masses safely and usefully, this could be a big deal.