**Title: Nvidia Targets $200 Billion CPU Sector with AI Agent PCs from Microsoft, Dell, and HP

Nvidia kicked off Taipei’s massive Computex trade show on Sunday with a literal spark, unveiling the RTX Spark, a new PC CPU the company describes as a “superchip.” Alongside the hardware announcement, Nvidia introduced a roster of prominent PC manufacturers poised to launch AI-enabled devices powered by this new technology. Designed to securely execute AI agents such as OpenClaw and Hermes Agent, the high-performance chip delivers speeds of one petaflop.

These RTX Spark-enabled Windows PCs are slated for release this autumn. Initial availability will include models from ASUS, Dell, HP, Lenovo, MSI, and Microsoft Surface, with Acer and Gigabyte expected to join the lineup subsequently. To ensure secure operation, the devices will feature sandboxes developed in partnership with Microsoft. Furthermore, the systems will be equipped with sufficient CPU, GPU, RAM, and underlying Nvidia CUDA software to support local instances of large language models. Nvidia asserts that its RTX technology will enhance AI performance, improve image fidelity, and enable AI capabilities in over 1,000 games and applications.

The company is positioning the technology as a vital tool for AI content creators while also offering a substantial upgrade for its traditional gaming audience. More than 100 Windows software developers, including Adobe, Blender, ComfyUI, Riot Games, and Xbox, have committed to supporting the new chip.

However, Nvidia founder and CEO Jensen Huang envisions a future that extends beyond these immediate applications. He aims to eliminate the traditional reliance on launching applications, typing, and clicking. “With RTX Spark and Microsoft Windows, you ask — and the PC does the work,” Huang stated in a press release. “Frontier models. Creative workflows. RTX games. All on a laptop.”

This launch follows Huang’s recent assertion to investors that Nvidia has identified a new $200 billion market for AI-focused CPUs, moving beyond its reliance on GPUs. He specifically highlighted the Vera, a high-end server CPU released earlier this year, noting that sales have already reached $20 billion. During May’s earnings call, Huang expanded on this vision, suggesting a future populated by billions of agents requiring tools analogous to human PCs. “We’re going to need a lot more CPUs,” he remarked.

Historically, attempts to bring ARM-based Windows devices to market have faced significant hurdles. In 2013, Microsoft incurred a $900 million write-off for its Nvidia ARM-based Surface RT, and partners such as Dell withdrew from the project. Nevertheless, Nvidia’s consistent record of quarterly revenue growth makes Huang’s renewed pursuit of the PC market difficult to dismiss. This new chip represents a fundamentally different approach, offering greater power rather than less. Microsoft is marketing its RTX Spark-powered device, the Surface Laptop Ultra, as “the most powerful Surface Laptop ever built.”

Despite the enthusiasm, PC manufacturers have yet to disclose detailed specifications or pricing for their respective offerings. The new systems resemble full-fledged Windows versions of Nvidia’s DGX Spark mini-computer, which is currently available to developers for approximately $4,800. It remains unclear whether these PCs will compete on price with the popular Mac Mini, which is frequently used to run OpenClaw, or if they will occupy the premium segment of the market, similar to Nvidia’s own agent-running hardware. Regardless of their positioning, if Nvidia successfully makes AI agents accessible, safe, and useful for the general public, the opportunity could be substantial.

Source: TechCrunch Generated at: 2026-06-01 21:35:00 UTC