An RTX Spark laptop is a Windows laptop built on Nvidia’s new N1X chip, the company’s first complete laptop processor with the CPU, GPU and memory in one package. The first models went on sale on 16 October 2026, and they promise something no other Windows laptop offers: desktop class Nvidia graphics plus up to 128GB of shared memory for AI. However, these machines are expensive and run Windows on Arm, so they are not right for everyone. This guide explains the specs, the models, the prices and who should actually buy an RTX Spark laptop.
What Is Nvidia RTX Spark?
Nvidia announced RTX Spark at Computex on 1 June 2026, built together with MediaTek. Simply put, the chip pairs Arm CPU cores with a Blackwell RTX GPU, the same graphics generation used in RTX 50 series cards. In addition, it is based on the same design as Nvidia’s DGX Spark mini AI computer, shrunk into a laptop power budget of 45 to 80 watts.
Above all, the big idea is unified memory. Instead of a separate graphics card with its own small pool of video memory, the CPU and GPU share one large pool. As a result, an RTX Spark laptop can load AI models that would never fit on a normal laptop GPU.
RTX Spark Laptop Specs
Nvidia’s official RTX Spark specifications list two laptop versions of the N1X chip:
| Spec | Top laptop version | Lower laptop version |
|---|---|---|
| CPU | 20 Arm cores | 18 Arm cores |
| GPU | Blackwell, 6,144 CUDA cores | Blackwell, 5,120 CUDA cores |
| Unified memory | Up to 128GB LPDDR5X | Up to 32GB LPDDR5X |
| AI performance | Up to 1 petaflop | 827 teraflops |
| Power | 45 to 80W | 45 to 80W |
In simple terms, the smaller GPU appears only in 24GB and 32GB configurations. The full chip is offered from 24GB all the way up to 128GB, so check the exact GPU core count before you buy a low memory model.
RTX Spark Laptop Models and Prices
So far, six brands launched RTX Spark laptops in the first wave: ASUS, Dell, HP, Lenovo, Microsoft and MSI. Later, Acer and Gigabyte are expected to follow. US starting prices at launch looked like this:
| Laptop | Screen | US starting price |
|---|---|---|
| Microsoft Surface Laptop Ultra | 15 inch | $2,599.99 |
| ASUS ProArt P14 | 14 inch | $2,599.99 |
| ASUS ProArt P16 | 16 inch | $2,799.99 |
| HP OmniBook X 14 | 14 inch | $2,999.99 |
| HP OmniBook Ultra 16 | 16 inch | $3,199.99 |
| Dell XPS 16 Creator Edition | 16 inch | $3,799.99 |
Lenovo models are also on sale through retailers, and MSI has announced its Prestige N16 Flip AI+, with prices varying by memory size. In addition, some 128GB configurations sell for well over $4,000. Prices move often because of the global memory shortage, so check the maker’s store before buying. We cover Microsoft’s model in depth in our Surface Laptop Ultra guide.
What an RTX Spark Laptop Does Well
Local AI models
First of all, this is the main reason to buy one. With 64GB or 128GB of unified memory, you can run large open models on the laptop itself, with no cloud and no monthly fee. Nvidia’s CUDA tools also work on the chip, so many AI apps built for Nvidia GPUs can run here. Our guide on how to run AI locally explains the free tools to start with.
Creative work
Similarly, the Blackwell GPU speeds up video editing, 3D rendering and AI image tools. Early reviews also show strong gaming results for a thin laptop, although it does not match a full RTX 5090 gaming laptop.
Battery life for its power
Because the CPU, GPU and memory sit in one package, these laptops avoid the heavy power draw of a separate graphics card. As a result, many models are thinner than typical creator laptops.
The Trade Offs You Should Know
- Windows on Arm. The CPU uses Arm cores, like Snapdragon laptops. Most modern apps run natively or through Microsoft’s Prism emulation, but some older software, drivers and games may not work. Check your must have apps first.
- High price. In fact, even the cheapest models cost more than most premium laptops.
- No memory upgrades. The memory sits inside the chip package, so buy the amount you will need for years.
- Linux is not ready yet. Linux support for the laptop versions is still in progress, which matters for some developers.
RTX Spark Laptop vs Other AI Laptops
Compared with a high end MacBook Pro, an RTX Spark laptop offers Nvidia’s CUDA software and Windows, while Apple offers mature macOS support and long battery life. Our MacBook for AI guide covers Apple’s side.
Against an RTX 5090 laptop, RTX Spark wins on memory for AI models, while the RTX 5090 wins on raw gaming speed and x86 compatibility. Finally, Snapdragon X2 laptops are far cheaper, but their graphics are much weaker; see our Snapdragon X2 Elite laptop guide for details.
Who Should Buy an RTX Spark Laptop?
In short, buy one if you are an AI developer, researcher or creator who wants to run big models on the go and already works with Nvidia tools. In that case, it can replace a desktop workstation for many tasks.
On the other hand, skip it if you mainly browse, write and stream, or if you rely on older Windows software. A Ryzen AI or Intel Core Ultra laptop will serve you better for far less money; our Ryzen AI laptop guide is a good place to start.
FAQ
When did RTX Spark laptops go on sale?
The first RTX Spark laptops went on sale on 16 October 2026 in the US, including the Surface Laptop Ultra, the Dell XPS 16 Creator Edition and models from ASUS and Lenovo. HP’s OmniBook X 14 and OmniBook Ultra 16 ship from 4 November 2026.
Can an RTX Spark laptop play games?
Yes. The Blackwell GPU runs many games well for a thin laptop. However, some games with anti cheat software may not run on Windows on Arm, so check each game first.
How much memory should I choose?
Generally, choose 64GB if you plan to run mid size AI models and 128GB for the largest ones. For creative work without local AI, 32GB is enough.