With Nvidia rolling out RTX Spark, an integrated PC chip aimed at AI agents, PCs that had been sidelined as consumer attention shifted to smartphones have gained more momentum than ever.

Nvidia, Microsoft and Apple are also stepping up moves to reinvent PCs with AI. Attention is focused on whether AI will shake the view that in the PC market, “if you can’t use it, just buy something adequate and use it,” and become a meaningful incentive that draws consumers to buy PCs.

Companies have tried to shake up the PC market with AI before, but the reality is those efforts did not show the disruptive power many expected. Microsoft and Qualcomm introduced Copilot Plus PCs 2 years ago, highlighting how AI could make it easier to find documents and edit photos, but they did not deliver a big impact.

Against that backdrop, Nvidia is working with Microsoft and PC makers so that AI agents can autonomously use the mouse and keyboard like users on RTX Spark-based PCs.

Nvidia appears to see the situation as different from 2 years ago, citing timing as AI agents become increasingly widespread.

The New York Times reported on June 1 that Nvidia CEO Jensen Huang (젠슨 황) said at the Computex IT trade show in Taipei, Taiwan, “Just as home theaters, big TVs, lawn mowers and dishwashers are no longer special, in the future an AI supercomputer could become a common household appliance. It is entirely imaginable that someday there will be an AI supercomputer in the home.” He added, “An AI supercomputer will run every agent for a user, operate every AI assistant, and let them always handle all sorts of tasks for the user.”

Nvidia’s RTX Spark is focused on supporting people who use PCs for AI system development, playing video games and assisting computer graphics creators.

RTX Spark-based PCs can be as thin as 14 millimetres and weigh less than 3 pounds (1.36 kg).

To expand RTX Spark-based laptops, Nvidia will initially work with six companies: Dell Technologies, Lenovo, Microsoft, HP, Asus and MSI. It plans to release details on RTX Spark battery life and performance this autumn.

Nvidia’s unveiling of RTX Spark has also drawn attention to competition in the laptop market between Apple and an Nvidia-Microsoft alliance. Apple has established a presence in the AI PC market. The New York Times reported that the Mac Pro is popular among AI developers, and that Mac minis with large amounts of memory are selling briskly as many AI developers use them to run AI agents.

Max Weinbach (맥스 웨인바흐), a tech analyst at Creative Strategies, said, “Right now, Apple effectively dominates the AI PC market. Nvidia wants to build a Windows laptop ecosystem that can be an alternative.”

Apple also appears to be continuing to increase spending to strengthen on-device AI capabilities that run AI on devices. Apple unveiled the M5 chip last October with a neural accelerator built into every GPU core. By its own performance measurements, the M5 can run a 30 billion-parameter model in 3 seconds. Apple’s push to expand on-device AI appears to be widening beyond developing efficient chips to include AI models that run on devices and the software capabilities that support them.

The Information recently reported that as part of a contract with Google, Apple can use Google’s Gemini models to train small models that can run on Apple devices. Apple is also considering acquisitions of small companies that could help shrink AI models so they can run on devices. The Information cited sources as saying companies Apple is considering include LiquidAI, a startup specialising in running AI on devices.

Apple, Nvidia and Qualcomm all provide Arm architecture-based PC chips commonly used in mobile chips, unlike Intel and AMD, which are based on x86.

Windows Central reported that Kedar Kondap (케다르 콘답), senior vice president of Qualcomm’s computing division, met reporters shortly after Nvidia announced RTX Spark and said, “Welcome to the family. It’s good proof the ecosystem is growing outside x86.” He stressed that Qualcomm has led efforts for years to build the Arm PC ecosystem, including printer support, software apps, peripheral compatibility and support for more than 2,500 games.

Arm-based PCs may have an advantage over the x86 camp in areas such as battery life, but securing sufficient compatibility with existing applications based on x86 chips may be a challenge. It has not yet been confirmed whether Nvidia has resolved compatibility issues with x86-based applications. Windows Central reported that current Snapdragon-based Arm PCs run x86 apps through Microsoft’s Prism emulation layer, but there are limits with some older apps or the latest games.