Why Nvidia’s RTX Spark Matters: It Paves the Way for Arm Compatibility Across the Board

Fourteen years ago this month, Microsoft launched an Nvidia-powered Windows PC that turned out to be one of the biggest tech missteps in recent memory. The original Microsoft Surface, which operated on a simplified version of Windows using an Nvidia Tegra chip, flopped so spectacularly that Microsoft had to write off $900 million in inventory. The main issue was that Nvidia’s Arm chip couldn’t run existing Windows applications, forcing users to abandon their legacy software.

Fast forward to this Friday, October 16th, and Nvidia is back in the spotlight with the debut of RTX Spark laptops, promising a new era for Arm-powered Windows PCs. This time, the company assures users that they’ll be able to run everything from Adobe and Office to Blender and even games that require kernel-level anti-cheat measures. “All your apps, all your games, everything’s going to run great on Spark,” said Nvidia senior director of products Justin Walker.

While it’s unlikely that many gamers have $7,000 lying around for an untested local AI workstation with 128GB of RAM, the excitement is palpable for a future where Arm chips can handle all the tasks we expect from Windows PCs. Nvidia recently held an event in San Francisco to emphasize that RTX Spark laptops can cater to gamers, creators, developers, and AI enthusiasts alike, each receiving tailored presentations on the chip’s advantages.

I’m currently writing this on one of the laptops handed out for review, and after several days of use, it feels largely indistinguishable from a traditional Windows machine. We’ll share more insights on its early performance, pros, and cons once the full embargo lifts.

The RTX Spark “superchip” isn’t entirely new; it’s the same chip that powered mini desktops for AI developers a year ago. However, this version runs full Windows 11, utilizing Microsoft’s Prism emulator to make older Windows applications and games compatible with the Arm architecture.

Nvidia isn’t the first to attempt to bridge the gap between PC software and a platform it wasn’t originally designed for. Since 2020, Apple has transitioned its computers from Intel x86 chips to its own Arm-based silicon, aided by the Rosetta 2 emulator. In 2022, Valve’s SteamOS showcased the potential for Windows games on Linux, and in 2024, Qualcomm introduced its first well-received Windows on Arm devices, including the Microsoft Surface Laptop 7th Edition, using the Prism emulator.

Nvidia aims to surpass these efforts, particularly in the realms of AI and gaming. While Qualcomm initially claimed that most games would “just work” on Snapdragon-powered Windows laptops, the reality was disappointing. Many games encountered errors or performance issues, and Qualcomm’s Snapdragon X Elite chips were designed for lower-end gaming. Major multiplayer titles like Fortnite were unplayable on either Qualcomm or SteamOS devices due to incompatible anti-cheat software.

In contrast, Nvidia’s RTX Spark boasts significantly more graphical power than a Steam Deck or Qualcomm Snapdragon X Elite, comparable to a midrange discrete laptop GPU like an RTX 5060 in early tests. Nvidia is collaborating with numerous game developers to ensure compatibility with a wide range of anti-cheat solutions. The company claims support for various anti-cheat software used in popular games like Fortnite, Roblox, Rust, Rainbow Six Siege, and MapleStory, which have previously faced compatibility issues on Valve’s SteamOS. Additionally, support for anti-cheat software in titles like Call of Duty, PUBG, League of Legends, Valorant, Apex Legends, and many EA Sports games is on the way.

However, some “CPU-limited” games may experience slower performance due to Microsoft’s Prism needing to convert x86 instructions to Arm instructions in real time, which can introduce delays. For instance, one colleague struggled to run Horizon Zero Dawn Remastered above 30fps, likely due to its CPU demands. Another colleague faced similar issues with Assassin’s Creed Shadows, where performance lagged compared to a MacBook Pro M5 Max.

Most games, however, are more limited by their graphics than their CPU, and the Nvidia Blackwell GPU in the RTX Spark is the same as that found in traditional laptops, requiring no translation. Early tests show smooth 1080p performance in titles like Cyberpunk 2077, and with some adjustments, it’s possible to push the resolution to 1800p or 2180p at lower settings with DLSS upscaling.

Nvidia claims you can achieve impressive frame rates in games like 007 First Light, Cyberpunk, Gears of War: E-Day, Doom: The Dark Ages, Naraka: Bladepoint, and Pragmata at 135fps at 1440p, but this is with DLSS and frame generation, which creates additional “fake” frames. These results may also utilize Nvidia’s faster new frame generation model, which could compromise image quality.

In a typical year, it would have sufficed for RTX Spark laptops to simply function well. However, Nvidia is currently riding high, reaping massive profits from the AI boom, which has driven up the costs of memory and storage. This means that RTX Spark laptops come with a hefty price tag, especially if you want enough memory and GPU cores to fully utilize their capabilities. The Surface Laptop Ultra starts at $2,600 with a scaled-down Nvidia N1X, while the full version with 20 CPU cores and 6,144 GPU cores costs $3,700, and you’ll pay up to $5,900 for 128GB of unified memory for larger local AI models.

On the upside, Windows PC manufacturers, including Microsoft, seem more willing to pursue MacBook Pro-level quality now that they can justify higher prices. By the time more gamers can afford an Nvidia machine—Nvidia has indicated that lower-cost versions of Spark are in development—any initial growing pains may be resolved.

The most promising aspect is that the future Nvidia is shaping isn’t solely theirs. Qualcomm played a crucial role in making today’s Windows on Arm a reality, and Nvidia’s advancements should enhance future Qualcomm devices as well. The rising tide of Arm technology benefits everyone involved.