Nvidia has announced a new computer chip called the RTX Spark that can run AI right on your laptop or desktop, instead of sending your requests off to the internet. Nvidia says it's working with Microsoft to remake the PC for the age of AI.
The chip will show up this fall in laptops and small desktops from Dell, HP, Lenovo, ASUS, Microsoft Surface, and MSI, with brands like Acer to follow. That puts Nvidia up against the companies that usually make the main chips in your computer, including Intel, AMD, and Apple.
Read More: Surface Laptop Ultra is Microsoft's NVIDIA-powered MacBook Pro rival
Why this chip is important
The RTX Spark is built on the same broad family of chip technology, called Arm, that Apple moved its MacBooks to a few years ago. That switch is a big reason newer MacBooks can be fast while still running cool and lasting all day on a charge, sort of the best of both worlds. Windows laptops have mostly missed out on that mix so far.
Over the last two years, compatibility has become a non-issue for most apps with one big exception: Gaming has continued to be hit or miss on Arm-based Windows machines. Nvidia says it has that solved. At the company's Computex press conference this week, Nvidia CEO Jensen Huang claimed the RTX Spark PCs can run "every application Windows has ever run." If that holds up, it would clear the one hurdle that has held these laptops back the most.
What "AI on your PC" actually means
Right now, when you ask an AI chatbot a question, your request usually travels to a giant computer somewhere far away, gets an answer, and sends it back. The RTX Spark is built to do that work on your own machine instead.
Your information can stay on your computer; it can keep working without an internet connection, and you don't have to wait on a far-off server. That's the part I like about the idea.
The catch is that the AI you can run on a regular laptop today, including ones powered by RTX Spark, still isn't as smart as the big online versions like ChatGPT. Local AI models are just not that good yet, although they are making solid progress. So the privacy and other offline perks you get from running local AI models are nice; it's just early days.
Nvidia is not alone in making promises about local AI. Qualcomm has been calling 2026 "the year of AI agents," meaning AI that can quietly handle tasks for you on its own. Maybe. But I've heard this kind of pitch before, and there still isn't an obvious reason for most people to pay extra for it.
Buyers seem to agree. Dell said earlier this year that people weren't buying these AI laptops as much as it had hoped, and a similar trend can be seen with other Windows laptop brands.
What makes me pay a bit more attention this time is Nvidia itself. It's the company behind the chips that power most of the AI you already hear about, so it knows this stuff better than almost anyone. Its chips are built to give AI access to a big shared pool of memory, which is exactly what running AI on your own machine needs. Nvidia has also made some solid Arm chips before, so the track record is as good as it can get.
Would I buy an RTX Spark-powered laptop over a MacBook right now? No, not yet. But if I needed a Windows laptop this fall, the RTX Spark would be on my list, and that's more than I can say for the AI PCs that came before it.
[Image credit: Nvidia, Microsoft]
