The Nvidia RTX Spark N1X is set to reach laptops in October 2026, and the company has now shared concrete details on the chip it once referred to simply as the RTX Spark. The processor will ship in two distinct configurations aimed at different device classes, giving hardware makers flexibility across notebooks and small desktops.
The higher-tier version pairs a 20-core Grace CPU with a Blackwell GPU featuring 6,144 CUDA cores. It supports between 24GB and 128GB of unified memory and is designed for both laptops and mini PCs. The second configuration uses an 18-core Grace CPU alongside a scaled-back Blackwell GPU with 5,120 CUDA cores and 24GB to 32GB of unified memory. That variant is limited to laptops only.
Two Configurations, One Platform
Beyond core counts and memory ceilings, Nvidia has not disclosed clock speeds, thermal targets, or pricing. With the launch scheduled for the coming month, additional specifications are expected from either Nvidia or its hardware partners in the near term. The public use of the N1X label is notable, as it had previously surfaced only in leaks. The naming also hints at the possibility of a lower-end N1 chip arriving later.
Nvidia timed the reveal to coincide with IFA in Berlin, where Lenovo and Acer introduced systems built around the new silicon. Lenovo showed the Yoga 9n 2-in-1 laptop, while Acer presented the SFF RTX Spark, a compact desktop. Neither device carried a price at announcement.
Broad OEM Support and AI Focus
More systems are expected from Asus, Dell, MSI, HP, and Microsoft. Microsoft previewed its flagship Surface Laptop Ultra at Computex earlier this year. These machines are being positioned primarily around agentic AI, though Nvidia has also demonstrated gaming performance in early showcases. The company has promised all-day battery life, a claim that will require independent testing to verify.
It remains unclear which devices will ship in October and which will follow later in the year. That timeline could firm up quickly now that Nvidia has committed to a launch window. A related open question is whether Microsoft’s Agent Framework, which allows agents to run in the background under operating system control, will be ready in Windows by then.
Unified Memory and the Windows on Arm Shift
Nvidia’s unified memory strategy mirrors the approach Apple popularized with its M-series chips, including the recently announced M6 and M5 Ultra, both of which also rely on an Arm instruction set. On the x86 side, AMD’s Strix Halo and the upcoming Gorgon Halo employ unified memory as well.
The arrival of the RTX Spark N1X also ends Qualcomm’s exclusivity on Windows on Arm, introducing a major new silicon supplier to the platform. That expansion could attract additional partnerships from application developers looking to build on Nvidia’s software tools. Both configurations are scheduled to begin appearing in shipping devices in October 2026.
Source
Image: tomshardware.com
