NVIDIA’s RTX Spark platform is arriving in Windows PCs aimed at local AI work, creative tasks and gaming, with configurations offering up to 128GB of unified memory. Engadget’s Devindra Hardawar reports that prices start at $2,600 for an 18-core system and rise to $3,700 for the 20-core model; the 128GB Surface Laptop Ultra is listed at $5,900.
NVIDIA Watch analysis
What happened
RTX Spark combines an Arm-based Grace CPU with NVIDIA’s Blackwell GPU on one chip. Hardawar reports configurations with either 18 CPU cores and 5,120 GPU cores or 20 CPU cores and 6,144 GPU cores, alongside up to 128GB of unified LPDDR5x memory. NVIDIA says the platform is twice as efficient as a laptop with an Intel Core Ultra 9 285H and RTX 5070 when gaming at 60 watts.
In a brief hands-on test, Hardawar says Gears of War: E-Day ran smoothly at 1440p on battery. NVIDIA is targeting the platform at developers and researchers who want to run large models locally, as well as people doing creative work or gaming. Adobe is reworking Photoshop and Premiere for the platform, while Microsoft’s Prism emulator is intended to run x86 software. The full Engadget report includes the author’s impressions and the configuration details.
Why it matters
The practical pitch is unusually clear: more memory for local AI models, with a powerful GPU and an Arm CPU in a laptop-sized system. That could suit people who want to experiment with larger models without sending every task to a cloud service. It also gives Windows PCs a new high-end option for creative work, not just AI workloads.
There are trade-offs. The prices put these machines well beyond casual upgrades, and Arm compatibility still matters: some x86 apps and games may depend on emulation or support from their developers. Hardawar argues that people who mainly use a laptop plugged in may see less benefit from the efficiency pitch.
Our read
The memory and battery-performance ambitions make RTX Spark more interesting than another round of AI branding on a laptop lid. But $2,600 is a hefty entry fee, and a short hands-on session is a promising first impression, not a verdict on sustained performance, software compatibility or local-model speed. Buyers should wait for testing that matches their actual apps and workloads.
What to watch
- Independent tests of performance, battery life and thermals under sustained workloads.
- How well the Prism emulator handles the apps and games buyers already use.
- Whether the local AI software ecosystem makes use of the available memory without awkward setup.
Discussion spark: Would you pay a premium for a Windows laptop that can run larger AI models locally, or would cloud access and cheaper hardware make more sense for your workload?
Sources and evidence
- NVIDIA's RTX Spark is good for Windows, even if you don't need local AI (15 October 2026, 13:00 UTC)
Independent WittyWires tracker for public updates about NVIDIA. Not affiliated with or endorsed by NVIDIA; this is not an official account.