Fourteen years ago this month, Microsoft launched an Nvidia-powered Windows PC that would become one of the biggest technological failures in modern history. The original Microsoft Surface, which ran a stripped-down version of Windows on an Nvidia Tegra chip, failed so badly that Microsoft had to write off $900 million in inventory.
The problem was that Nvidia’s Arm chip couldn’t run existing Windows applications. Buyers had to leave behind their legacy programs.
This Friday, October 16, Nvidia will once again be at the heart of Arm-powered Windows PCs with the first wave of RTX Spark laptops. But now, the company says, they will execute all. Adobe. Office. Blender. Even games that require kernel-level anti-cheat.
“All your apps, all your games, everything will work great on Spark,” Nvidia senior product manager Justin Walker told reporters last week.
I’m willing to bet more Edge readers don’t have up to $7,000 burning a hole in their pockets to buy an untested local AI workstation with 128GB of RAM. Even if you do, I’m not sure you’d appreciate the tradeoffs of a thin laptop that maxes out its GPU to produce “free” tokens locally. As Tom points out, it’s much slower than cloud AI.
But it’s hard not to be excited for a future where Arm chips can do everything other things we expect from Windows PCs.
Nvidia held Editor’s Day in San Francisco last week to convey the idea that RTX Spark laptops can be everything. AI players, creators, developers and power users each got its own dedicated slide deck to explain the benefits of the chip.
In the end, Nvidia handed out laptops for review and I’m writing this story on one right now. After several days I am finding it. mostly indistinguishable from a traditional Windows machine, although we’ll have more thoughts on the early pros, cons and quirks when the full embargo is lifted tomorrow.
The RTX Spark “superchip” isn’t an entirely new piece of silicon: It’s the same chip the company installed in mini desktop computers for AI developers a year ago. But those boxes ran Linux; This one runs full Windows 11, with Microsoft’s Prism emulator that makes old Windows apps and games work on the Arm chip.
Nvidia isn’t the first company to make a big bet on extending PC software to a platform it wasn’t originally designed for. As of 2020, Apple moved all its computers from Intel x86 chips to its own Arm-based silicon with the help of its Rosetta 2 emulator. In 2022, Valve’s SteamOS demonstrated the potential of Windows gaming on Linux, and then on Arm, with the help of the Linux community. And in 2024, the Prism emulator made its first appearance when Qualcomm launched the first well-received Windows on Arm machines, including the 7th edition of the Microsoft Surface Laptop.
But Nvidia wants to surpass them all when it comes to AI and gaming.
Although Qualcomm originally suggested that most games should “just run” on Snapdragon-powered Windows laptops, it was a rude awakening. Some games failed completely, some had issues, and Qualcomm’s Snapdragon X Elite chips were aimed at relatively low-end games to begin with. Many of the world’s biggest multiplayer games, such as fortnitewould not run on Qualcomm or SteamOS devices because its anti-cheat software was incompatible with Arm and Linux respectively.
However, Nvidia’s RTX Spark has substantially more graphics power than a Steam Deck or a Qualcomm Snapdragon
Nvidia says it already supports the various anti-cheat programs used in fortniteRoblox, Oxide, rainbow six siegeand History of maple – all of which have been incompatible with Valve’s SteamOS – and Nvidia says that the various anti-cheat software that protects Obligations, PUBG, League of Legends, Valuing, Apex Legends, Battlefieldand many EA Sports titles will officially come to RTX Spark later.
The reality is that some “CPU-limited” games will run slower because Microsoft’s Prism has to convert its x86 instructions into Arm instructions in real time, adding a delay to everything. My colleague Tom Warren was unable to appear Horizon Zero Dawn Remastered 30+ fps no matter what you tried, probably because that game is pretty CPU limited. And when he tried Assassin’s Creed Shadowsmy colleague Antonio Di Benedetto couldn’t beat 30fps there either: a MacBook Pro M5 Max ran that game twice as fast.
But most games are still limited by their graphics rather than the CPU, and the Nvidia Blackwell GPU inside the RTX Spark is the same Blackwell you’d find in a traditional laptop, no translation required. So far, we’re seeing smooth 1080p performance in games like Cyberpunk 2077and we found that you can upscale that to the native 1800p or 2180p resolution of these laptops if you’re willing to settle for low or medium settings with plenty of DLSS upscaling.
(Officially, Nvidia claims that you can play 007 First light, cyberpunk, Gears of War: E-Day, Doom: The Middle Ages, Naraka: sword tipand pragmatic at 135 fps at 1440p, but that’s with DLSS and Generation of 3x frames, that is, two “fake” frames for each real one. Those results could also be due to using Nvidia’s new faster frame generation model with reduced image quality.)
In a normal year, when Nvidia was a normal company, it would have been enough for the RTX Spark laptops to “just work.” Instead, Nvidia is on top of the world, making hundreds of billions of dollars in profits selling shovels in the AI gold rush, which, in turn, has driven the price of memory and storage into the stratosphere.
The bad news is that this makes these RTX Spark laptops very expensive, especially if you want enough memory and GPU cores to do everything they say. For $2,600, the Surface Laptop Ultra actually starts with a stripped-down Nvidia N1X with 18 CPU cores and 5,120 GPU cores and 24GB of memory; You’ll pay $3,700 for the full complement of 20 CPU cores and 6,144 GPU cores, and up to $5,900 if you want 128GB of unified memory to accommodate larger local AI models.
The good news is that Windows PC makers, including Microsoft, seem more encouraged to pursue MacBook Pro-like quality now that they can justify asking that much money. And by the time the rest of us can afford an Nvidia machine (Nvidia says Spark is a family of products with lower-cost versions in the works), perhaps the early growing pains will have been ironed out.
The best news is that the future Nvidia is building doesn’t belong to Nvidia alone. Qualcomm helped make today’s Windows on Arm possible, and Nvidia should also help make tomorrow’s Qualcomm machines much better.
Arm’s rising tide lifts all his boats.
