Skip to content

NVIDIA RTX Spark Windows Laptops Reach Stores from Six PC Makers, Starting at $2,599

Nine launch models share one NVIDIA chip in two configurations, with up to 128GB of unified memory and guided installers for local AI agents.

NVIDIA RTX Spark laptops with see-through chassis showing the chip, above a glowing circuit pattern
NVIDIA RTX Spark · Credit: NVIDIA

NVIDIA RTX Spark laptops went on sale today from ASUS, Dell, HP, Lenovo, Microsoft and MSI, putting a Blackwell RTX graphics processor and a Grace CPU on a single chip inside slim Windows machines. Prices start at $2,599, and the chip can address up to 128GB of memory shared between processor and graphics.

NVIDIA lists nine launch models: 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 the MSI Prestige N16 Flip AI+ and Prestige N16. They ship globally through the makers' websites and major retailers, with pricing and configurations varying by region, and NVIDIA says more systems will follow. New Perplexity users get a limited-time promotion with a purchase.

The chip, the RTX Spark N1X, comes in two laptop configurations, per NVIDIA. The larger pairs a 6,144-core GPU with a 20-core CPU and up to 128GB of LPDDR5X. The smaller has 5,120 GPU cores, 18 CPU cores and up to 64GB. Both are rated at 45 to 80 W TDP.

NVIDIA rates the chip at up to 1 petaflop of AI compute in FP4, a low-precision format used for running models. For games, it quotes more than 100 frames per second in AAA titles at 1440p. Launch-day game support includes Fortnite, Marvel Rivals and NBA 2K27, which uses DLSS 5 with 3D-Guided Neural Rendering, according to NVIDIA. A new AI model for DLSS 5, available today, raises its frame rates by up to 1.4x.

NVIDIA also aims the machines at local AI agents. Guided installers set up GitHub Copilot, Hermes, OpenClaw, Perplexity, Pi and WorkBuddy without a command line, and NVIDIA's Nemotron 3.5 Bolt, a 74-billion-parameter model tuned for tool calling, runs on supported configurations. A laptop can also act as an inference server for other machines through NVIDIA PAIR, which routes requests across Windows, Linux and macOS computers on a home network and works with Ollama, llama.cpp and LM Studio.

For developers, NVIDIA and Microsoft say they have readied the first PyTorch-CUDA build for Windows on Arm, a native Python build and support for 97% of the top 15,000 Python packages. Adobe's Photoshop and Premiere run natively on RTX Spark in beta releases available at launch, while Substance 3D support is coming soon.

Share this story

Ravi Murthy

Ravi Murthy is the news-desk byline at techshooked for breaking stories across cybersecurity, AI, hardware and policy. A brief under this name gives the facts as the primary source states them, names who said what, marks what is still unconfirmed, and adds the one piece of context a reader needs to judge how much the story matters.