doge

News · 16 Oct 2026 · 3 min read

NVIDIA RTX Spark laptops go on sale, built to run AI agents on your own PC

NVIDIA says laptops from ASUS, Dell, HP, Lenovo, Microsoft and MSI are available now, starting at $2,599.

A slim silver laptop sits half open on a dark desk, its keyboard glowing in blue and green light.

NVIDIA said on October 16, 2026, that laptops built on its new RTX Spark chip are on sale now. NVIDIA's press release names ASUS, Dell, HP, Lenovo, Microsoft and MSI as the first makers, and says more systems will follow.

What is inside

RTX Spark puts NVIDIA's graphics chip and a processor on one piece of silicon. NVIDIA says MediaTek helped design the processor. NVIDIA says the chip shares up to 128GB of memory between the two and can do up to 1 petaflop of AI math. Those are NVIDIA's top-end figures.

The point, NVIDIA says, is to run big AI models and AI agents on the laptop itself instead of in the cloud. NVIDIA CEO Jensen Huang says the machines are built for creating, gaming and "running personal AI agents."

Agents out of the box

NVIDIA says guided installers set up agents such as GitHub Copilot, Hermes, OpenClaw, Perplexity, Pi and WorkBuddy without typing commands. It says security tools from NVIDIA and Microsoft let you control what an agent is allowed to do.

Games and creative apps

NVIDIA says the laptops can play big games at 1440p at over 100 frames per second. Its launch list includes Cyberpunk 2077, Fortnite and NBA 2K27. Adobe says Photoshop and Premiere have beta versions that support RTX Spark now, with Substance coming later.

Which laptops, and the price

NVIDIA lists the ASUS ProArt P14 and P16, Dell XPS 16, HP OmniBook Ultra and OmniBook X 14, Lenovo Yoga Pro 9n, Microsoft Surface Laptop Ultra, and MSI Prestige N16 Flip AI+ and Prestige N16. NVIDIA says prices start at $2,599, and that prices and configurations vary by region.

What you can try

If you are shopping, check how much memory a model has. NVIDIA's 128GB figure is the maximum, and the bigger AI models need the bigger configurations.

Source: NVIDIA Newsroom