Nvidia's RTX Spark chips arrive in premium laptops and compact desktops designed for local AI

Nvidia's RTX Spark "superchip" is finally making its way into consumer hardware. Following the company's announcement three months ago, hands-on systems are now on display at IFA 2026 in Berlin, offering a glimpse at how the chip will power a new generation of AI-capable personal computers.

These devices share a common goal: running sophisticated AI models locally rather than relying on cloud services. For privacy-conscious users working with sensitive information—financial data, emails, and personal documents—keeping AI processing on-device eliminates the need to transmit such content to external servers.

Premium laptops challenge the MacBook#

The standout device is Lenovo's Yoga 9n 2-in-1, a 16-inch convertible that positions itself as a direct MacBook Pro alternative. The machine packs impressive features across the board: a 2880 x 1800 OLED display refreshing at 120 Hz, a six-speaker audio system, a 9.2-megapixel camera, and metallic edges with minimal bezels. At just 0.69 inches thick, it's nearly as slim as Apple's comparable 16-inch model.

The real innovation lies beneath the surface. Nvidia's Arm-based SoC pairs a Grace CPU (up to 20 cores) with a Blackwell RTX GPU (up to 6,144 cores) on a single chip. GPU performance theoretically falls between an RTX 5070 Ti and RTX 5080 found in traditional Windows laptops. This integration approach mirrors Apple's silicon strategy and should deliver better battery efficiency than discrete graphics would.

Lenovo is also releasing a 15-inch variant, the Yoga Pro 9n, which trades the 2-in-1 hinge for a more traditional form factor. At 0.66 inches thick, it's even thinner and features a stylus-enabled touchpad and higher-resolution screen (2560 x 1600). The key distinction between the two models appears to be memory: the 16-inch maxes out at 64 GB, while the Pro model scales up to 128 GB.

Other RTX Spark laptops launching this fall include offerings from Dell (XPS 16), Asus (ProArt P16), Microsoft (Surface Laptop Ultra), and HP (OmniBook X 14), all targeting creators and AI enthusiasts.

Compact desktops join the race#

Beyond laptops, Acer and Asus have unveiled mini PC designs built around the RTX Spark chip. Roughly Mac Mini-sized, these systems pack up to 128 GB of memory and deliver what manufacturers describe as "petaflop AI performance"—enough to run demanding agentic workflows without cloud assistance.

The timing places these systems in direct competition with Apple's newly announced M6 Mac mini and M5 Ultra Mac Studio. Lenovo is also taking a different approach with its AMD-powered ThinkCentre X Ultra, featuring Ryzen Max+ Pro 495 processing and 128 GB RAM, priced from $3,699. AMD's alternative is gaining traction for local AI use cases, including the Framework Desktop, and the AMD Developer Kit is already available for purchase.