The laptop has always been a machine of compromises. Workstation-class performance typically arrived in thick chassis with short battery life and fan noise audible from across a room. Getting genuine power in a form factor thin and light enough to carry without a second thought has been largely Apple’s territory, a problem it’s been solving with its own ARM-based chips while Windows machines played catch-up.
NVIDIA is changing that calculus for Windows with RTX Spark, an ARM-based superchip that fuses a 20-core Grace CPU with a Blackwell RTX GPU carrying 6,144 CUDA cores, connected by NVIDIA’s NVLink-C2C chip-to-chip interconnect. Microsoft built the Surface Laptop Ultra around it from the silicon up, designing the machine and the chip in concert, producing what it describes as the most powerful Surface Laptop ever built.
The reason ARM architecture matters for laptop design is power efficiency. Compared to x86 chips, ARM-based designs deliver significantly more performance per watt, and that ratio determines what’s physically possible in a chassis. RTX Spark laptops are engineered to be as slim as 14mm and as light as 3 pounds, proportions that previously excluded any serious dedicated GPU from the equation entirely.
The Surface Laptop Ultra lands at under 18mm thick and under 4.5 pounds, housed in CNC-machined aluminum in Platinum and Nightfall finishes. The 15-inch mini-LED PixelSense Ultra touchscreen reaches up to 2,000 nits of peak HDR brightness with a 3:2 aspect ratio and 262 pixels per inch, making it the brightest display Microsoft has ever shipped on a Surface. A full port set, including HDMI, USB-C, USB-A, SD card, and headphone jack, rounds out a machine designed for professional use.
RTX Spark’s most defining architectural choice is unified memory, where up to 128GB of RAM is shared dynamically between the CPU and GPU. A 3D rendering job, a video edit, and a locally running AI model can all draw from that same pool simultaneously, without the bottlenecks discrete memory architectures create. That arrangement enables 1 petaflop of AI compute, enough to run 120-billion-parameter models entirely on the device.
The full CUDA software stack runs natively on RTX Spark, which matters directly for creative professionals. Adobe is rebuilding Photoshop and Premiere from the ground up for the chip, targeting 2x faster AI and graphics performance. On the creative side, RTX Spark handles 12K video editing, renders 90GB-plus 3D scenes using NVIDIA OptiX, and generates 4K AI video, tasks that previously required a dedicated workstation to complete without serious compromise.
NVIDIA describes RTX Spark as the most efficient PC chip ever built, a statement aimed squarely at Apple Silicon’s grip on the high-end creative laptop market. That efficiency is also what allows the Surface Laptop Ultra’s all-new thermal system to sustain heavy workloads without the throttling and fan noise that defined previous Windows machines in this tier. Microsoft’s own engineers worked across mechanical, thermal, materials, and industrial design disciplines simultaneously, treating the chassis and the chip as a single system.
All-day battery life holds even while running on battery power, and the compact charger is small enough to fit in a jacket pocket. The Surface Laptop Ultra and additional RTX Spark-powered devices from ASUS, Dell, HP, Lenovo, and MSI are expected in fall 2026. For a platform that has long asked users to choose between portability and capability, the arrival of an ARM PC chip in NVIDIA’s hands changes the terms of that conversation considerably.
