Nvidia has pinned October on its RTX Spark N1X launch, with laptop and mini PC chips aimed at local AI work.
According to Tom’s Hardware, the company has confirmed two N1X configurations, both pairing Grace CPU cores with Blackwell GPU hardware and a chunk of unified memory.
The bigger version uses a 20-core Grace CPU and a 6,144-CUDA-core Blackwell GPU, with 24GB to 128GB of unified memory.
That higher-end part is meant for laptops and mini PCs, giving vendors room to build compact systems for agentic AI and graphics work.
The smaller version keeps the same basic idea but drops to an 18-core Grace CPU and a 5,120-CUDA-core Blackwell GPU. That chip supports 24GB to 32GB of unified memory and will be limited to laptops, at least from what Nvidia has said.
Nvidia has not provided clocks, power limits or other awkward details, leaving plenty for OEMs and reviewers to dig through next month.
The announcement came ahead of IFA in Berlin, where Lenovo and Acer showed the first fresh RTX Spark systems using the new silicon. Lenovo revealed its Yoga 9n 2-in-1, while Acer showed an SFF RTX Spark mini desktop, with neither company announcing pricing.
More systems are expected from Asus, the Grey Box Shifter, Dell, MSI, HP and Microsoft, although launch timing remains a bit vague.
Microsoft has already shown a Surface Laptop Ultra prototype using Nvidia’s platform, with 128GB of RAM and a 20-core Arm CPU. Nvidia is pitching these machines as agentic AI machines, but it has shown gaming demos and promised all-day battery life for good measure.
The N1X name had appeared in leaks before, but this is the first time Nvidia has used the branding in public. That wording suggests a lower-end N1 chip could arrive later, although Nvidia has not confirmed what that part might look like.
It is still unclear which systems will ship in October, and whether Microsoft’s Agent Framework will be ready in Windows by then.
