NVIDIA RTX Spark Laptops Are on Sale: Models, Prices and Who They Suit

Author TeraNews16.10.2026◉ 6

The first laptops built on NVIDIA’s RTX Spark platform reached shops on 16 October. It is the first time in a long while that NVIDIA has entered the laptop market not as a graphics card supplier but as the author of the whole core of the machine: CPU, graphics and memory in one chip package. Pre-orders opened on 7 October, and machines from Microsoft, ASUS, Dell and Lenovo have now started shipping to customers. HP models follow later, and compact desktops on the same chip are promised for November.

What is inside RTX Spark

RTX Spark is called a superchip, and the label is fair. A single package combines a Grace CPU on the Arm architecture with RTX graphics from the Blackwell generation. Both halves share one pool of LPDDR5X memory of up to 128 GB. There is no usual split between system RAM and video memory: whatever the CPU can see is immediately available to the graphics core too.

The chip comes in two versions:

Lower tier: 18 CPU cores and graphics with 5,120 cores.
Top tier (full die): 20 CPU cores and 6,144 graphics cores. This is the one used in the 128 GB configuration.

The laptops run Windows 11 for Arm. NVIDIA’s main selling point is local AI: according to the company, the platform can handle language models of up to 120 billion parameters and a context window of up to 1 million tokens. That is a manufacturer claim rather than the result of independent testing, but the idea is clear: a large pool of shared memory lets you load a model that would not fit into the graphics card of an ordinary gaming laptop.

Which laptops are out and what they cost

Eight manufacturers are launch partners: Acer, ASUS, Dell, Gigabyte, HP, Lenovo, Microsoft and MSI. The prices below are US prices before tax, as reported by Windows Central.

Microsoft Surface Laptop Ultra from $2,599.99. A 15-inch touchscreen mini-LED display at 3K resolution, up to 128 GB of memory and a reworked cooling system. On sale from 16 October.
ASUS ProArt P14 from $2,599.99. 14 inches, 0.55 inches thick, 3.26 lb (about 1.5 kg). ASUS calls it the lightest laptop in the ProArt line. On sale from 16 October.
ASUS ProArt P16 from $2,799.99. 16 inches, OLED, up to 128 GB of memory. On sale from 16 October.
Lenovo Yoga 9n from $3,099.99. A convertible with a 16-inch screen. On sale from 16 October.
Dell XPS 16 from $3,799.99. A 3.2K OLED with a touch layer, about 0.7 inches thick. On sale from 16 October.
HP OmniBook X 14 from $2,999.99 and HP OmniBook Ultra 16 from $3,199.99. Both are expected on 4 November.
MSI Prestige N16 Flip AI+ from $3,299.99. A 16-inch convertible with a Tandem OLED panel.

The $2,599 floor applies to base configurations. Versions with the top chip and 128 GB of memory cost considerably more, so the real budget for a “full” RTX Spark machine easily climbs past the price of a good gaming laptop.

What the first reviews say

Journalists received the hardware early, but until 16 October they could only discuss the chip itself, not verdicts on specific laptops. Even so, a fairly clear picture has emerged.

Windows Central calls RTX Spark graphics very powerful and notes that the platform shines most when running local AI models. HotHardware reports that the top N1X version delivers competitive multithreaded CPU performance against leading mobile platforms and best-in-class graphics performance. Tom’s Hardware praises the maturity of the platform for a first-generation product but says plainly that games are not its strong suit. Tom’s Guide advises against judging the chip by single-core CPU speed alone, because that hides its main strengths.

NVIDIA itself positions RTX Spark as a tool for developers, designers, video editors and people who run AI models on their own computer. There are no gaming laptops on this chip in the launch wave. Alongside the launch, the company is updating its software ecosystem: GitHub Copilot, Perplexity and Hermes will get local AI agents, and Microsoft is adding isolated containers to Windows so that agents can take actions in the system without unrestricted access to it.

Limitations to know about before you buy

First and most important: the memory is soldered to the board. The amount you choose at purchase stays with you for the life of the laptop, and you cannot expand it later. On a platform like this that matters even more, because memory capacity decides which AI models you will be able to run.

Second: the Arm architecture. Windows can run x86 programs through emulation, and according to reviewers most compatibility problems have already been solved. But programs only reach top speed in ARM64 builds. If your work depends on niche software, drivers for older hardware or corporate utilities, check in advance whether they have Arm versions.

Third: games. Reviewers report that major titles run on the chip, and DLSS 5 debuts on RTX Spark, but nobody calls these laptops gaming machines. On top of that, graphics performance in any given model will depend on the cooling and power settings the manufacturer chose, so the same chip can behave differently in different chassis.

What this means for buyers

RTX Spark opens a new niche: a thin premium laptop that can hold genuinely large AI models in memory without weighing as much as a workstation. Until now, a similar scenario in the Windows world required a bulky gaming laptop with a discrete graphics card, and those usually have several times less video memory.

Who should look now

Developers who test AI models locally rather than in the cloud. Video editors and 3D designers who want powerful graphics and a large pool of memory in a portable body. People who work with confidential data and do not want to send it to outside services. For these tasks, the combination of plenty of shared memory and strong graphics offers something classic x86 laptops do not.

Who should wait

Gamers: a gaming laptop with a discrete RTX card will make more sense for the same money. People who depend on rare programs with no Arm versions. And anyone buying a laptop for the office, studying and browsing: RTX Spark is simply overkill there, and the premium runs to thousands of dollars.

One more piece of advice concerns configuration. Since memory cannot be expanded, there is no point skimping on it with a laptop like this. If local AI is the goal, aim for the higher-end variants, otherwise the platform’s main advantage stays on paper.

The prices above are US prices, and first shipments traditionally go to the US market, so check local pricing and availability before planning a purchase. A sensible strategy for most buyers is to wait for full reviews of specific models, which start appearing from today, compare cooling and battery life, and see how quickly developers ship ARM64 versions of their software. The first generation already looks mature, but it is usually the second that goes properly mainstream.

What to pair with a new laptop

Like most thin modern laptops, the RTX Spark machines lean on USB-C for charging, displays and peripherals. That makes a few inexpensive accessories more useful than any sleeve or case.

This section contains affiliate links: if you buy through them, the site earns a small commission from AliExpress at no extra cost to you.

Baseus Mini 100W GaN charger

A spare 100 W charger lives in your bag so the original can stay on the desk. This compact GaN model from Baseus is one of the best sellers on AliExpress.