Nvidia and Microsoft unveil RTX Spark chip for Windows AI PCs

By Splitfeed EditorialLast updated Oct 7, 2026, 19:00 UTC

On October 7, at a Microsoft event in San Francisco, Nvidia CEO Jensen Huang and Microsoft CEO Satya Nadella announced a co-engineering partnership centered on the Nvidia RTX Spark chip. The chip, built on Nvidia's Grace Blackwell architecture, marks Nvidia's entry into laptop system-on-chip design. The two companies said they are jointly developing hardware and software to allow AI agents to run locally on Windows machines. Leaked specifications for a Microsoft Surface Laptop Ultra, reported ahead of the event, pointed to configurations with up to 128 GB of unified memory and prices reaching $7,000, with a base model carrying an 18-core CPU and 24 GB of unified RAM. Separately, HP revealed pricing for its own RTX Spark laptops before an official announcement, according to The Verge. Huang framed the partnership as a return to roots, noting that Nvidia was founded because of Windows.

The announcement matters because it represents a meaningful shift in where demanding AI workloads are expected to run. Rather than relying on cloud servers, the RTX Spark design targets local inference on consumer and professional laptops, a direction that has real consequences for privacy, latency, connectivity dependence, and cost structure. The pricing, potentially reaching $7,000 for high-end Surface configurations, puts the hardware in workstation territory, raising genuine questions about who the intended customer actually is and whether the performance gains justify the premium over existing options from Apple, Qualcomm, and Intel.

The Pro-AI camp has responded with clear enthusiasm. Accounts including The Tech Drop highlighted that the chip can run 120-billion-parameter AI models directly on a laptop, treating that figure as a landmark. B&H Photo described the RTX Spark as a comprehensive workstation platform in a compact form factor, emphasizing capability density. A Japanese-language account framed the chip as a tool that would let creators turn ideas into finished work far more quickly, pointing to video editing, 3D production, and AI-assisted tasks. The Anti-AI camp has not published reactions to this story yet; that camp would typically raise concerns about energy consumption, the environmental cost of high-performance silicon, and whether the push toward AI integration in everyday hardware is driven by genuine user need or marketing momentum. The Middle Ground camp has also not weighed in; it would typically ask whether the performance claims hold up in real-world use and whether the price points limit meaningful access.

The key things to watch are the official Surface Laptop Ultra launch date and confirmed retail pricing, HP's formal RTX Spark product announcement, and early independent benchmarks testing whether 120-billion-parameter local inference performs as described under sustained workloads.