There’s a lot we still don’t know about NVIDIA’s RTX Spark AI chip — we’re still waiting on deeper technical details and pricing for the first batch of systems — but it has a decent shot of changing the way we think of Windows PCs entirely. RTX Spark should offer the raw performance that the first batch of Copilot+ systems lacked, and it will also push Microsoft to make the Windows experience even better for Arm CPUs (something I argued was the real highlight of the whole Copilot+ initiative, more so than AI support).
With Qualcomm’s latest Snapdragon chips powering low-end and mid-range Windows PCs, there’s room for NVIDIA to fill the gap at the high end, delivering powerful (and likely very expensive) PCs that will appeal to memory-hungry content creators, developers and AI enthusiasts.
Windows might finally get its Apple Silicon moment this year — a bigger push to optimize the aging OS for more efficient Arm chips, while also rethinking what’s possible when you can give GPUs direct access to insane gobs of memory. The MacBook Pro comparisons were immediately clear from the first batch of RTX Spark machines like the Surface Pro Ultra and ASUS’s new ProArt systems. Of course, powerful Windows laptops are nothing new, but they typically had tradeoffs like thick and heavy cases, or enormous battery-draining power demands. What if you could get a powerhouse PC that didn’t feel like a brick, and also lasted as long as a low-power ultraportable?
Microsoft
That’s the promise of RTX Spark systems, but there are plenty of details still up in the air. Based on NVIDIA’s description of the chip, RTX Spark appears to be the same hardware that’s in the DGX Spark AI workstation that launched last year for $3,999, but now sells for $4,699. That makes me think the initial RTX Spark systems could start around $4,000 or more. Its 20-core CPU is made up of 10 Cortex X-925 cores and 10 Cortex-A275, which are both slower than Arm’s newer C1-Ultra core, Engadget Editor-in-Chief, Aaron Souppouris, points out. That hardware is currently only found in the Mediatek Dimensity 9500 chip, powering the new Oppo Find X9 Pro. NVIDIA’s Arm cores are also slower than the latest Snapdragon Elite Oryon cores, as well as Apple’s M5 chips. If NVIDIA is charging a premium for RTX Spark systems, you’d expect them to at least be using CPU cores from 2026.
The real stars of the RTX Spark systems will be their 6,144 RTX Blackwell GPU cores, which is the same as the RTX 5070 desktop GPU, along with their support for up to 128GB of unified memory. Conceptually, that’s not far off from what Apple did with its M-series chips, which also combined Arm cores, beefy GPUs and tons of unified memory. Apple Silicon gave Macs a huge performance and efficiency boost over PCs, and the entire Windows ecosystem has been struggling to keep up.
The only competition for NVIDIA’s hardware on the PC side is AMD’s Ryzen AI Ultra lineup, which are also complex system-on-chip designs with powerful CPU cores, built-in graphics, a powerful NPU and integrated memory. Last year’s flagship Ryzen AI Max+ 395 (AKA “Strix Halo”) chip supported up to 128GB of RAM, and the newer AI Max 400 series brings that up to 192GB of unified memory.
The Ryzen AI Max lineup is still based on the x86-64 architecture, giving them total compatibility with legacy Windows software, but none of the power efficiency benefits of Arm cores. NVIDIA RTX Spark systems, just like the Snapdragon X Elite Copilopt+ machines, will have to run x86 software through Microsoft’s Prism emulator. I found that to be a pretty seamless experience while reviewing the Surface Pro Copilot+ in 2024, and Microsoft says it’s also been working together with NVIDIA to optimize Windows 11 even more for the RTX Spark.
Last year, Microsoft’s Copilot+ initiative for AI PCs seemed like a dud, but I argued at the time that it at least pushed PC makers to solidify new standards for their premium laptops. Finally, just about everyone started offering their PCs with at least 16GB of RAM. The real failure of Copilot+ was that it hinged on AI features few people wanted, and which were fraught with privacy issues, like Microsoft’s Recall.
Things are a bit different now, as many developers and power users are adopting AI agents like OpenClaw. NVIDIA is clearly trying to target that crowd with the RTX Spark systems — the people willing to spend a ton of money to get PCs with lots of RAM just so they can use the latest AI tools and models. (Those folks are also why the Mac Mini has been out of stock for months, and you can partially blame them for the death of the $599 model.)
Beyond AI enthusiasts, I could also see content creators and non-AI developers tempted by the possibility of an exorbitant amount of unified memory. RTX Spark systems theoretically would be able to handle massive video editing jobs far better than existing laptops. Their Blackwell GPUs also mean you’ll have full access to NVIDIA’s DLSS AI upscaling and frame generation for games (and you’ll likely have more than enough overhead for serious gaming while working on AI and rendering jobs in the background).
As powerful as RTX Spark systems may be, they’ll ultimately be limited by their pricing. Based on what we know so far, these will be incredibly expensive machines for deep-pocketed consumers. Still, there’s hope for the future: Rumors point to NVIDIA working on stripped down versions of the chip for cheaper systems, and potentially genuine ultraportables. And hopefully by that point, the RAMaggedon will be over, and more people will actually be able to afford RTX Spark systems.
