Nvidia Debuts RTX Spark Superchip for Thin Windows PCs

At Computex in Taipei, Nvidia unveiled the RTX Spark superchip for thin Windows laptops and compact desktops, pairing a 20‑core Arm central processor with a Blackwell graphics engine and estimating up to 1 petaflop of local AI performance. The company presented the chip as a full computing platform for slimmer Windows systems.

The design supports up to 128GB of unified memory and is intended to run RTX‑class gaming, creative applications and developer tools on thin laptops and small desktops. Nvidia listed compatibility with its existing software stack, including CUDA, RTX and DLSS. The company said more than 100 software providers and game developers are backing the platform.

Nvidia’s entry into PC processors places it alongside Intel, AMD and Qualcomm in the Windows laptop market. Nvidia reported fiscal 2026 revenue of $215.9 billion, a 65% increase year over year; Intel’s full‑year 2025 revenue was $52.9 billion and broadly flat.

Intel executives have emphasized the continuing role of the CPU in AI computing. Intel CFO David Zinsner in April described the results as reflecting the “growing and essential role of the CPU in the AI era.”

Qualcomm has promoted Windows on Arm with chips such as the Snapdragon X Elite and has introduced Snapdragon C, an entry‑tier Windows platform priced from about $300. AMD’s laptop processors remain largely x86‑based.

Analyst Jason Tsai warned the platform could remain a niche, luxury product unless complete systems land around a $1,500 price point, where RTX Spark would compete with mainstream and high‑end notebooks.

Nvidia framed RTX Spark as a hardware and software platform that integrates CPU, GPU, memory and drivers for Windows devices, presenting it as a way to add local AI acceleration alongside gaming and creative workloads on thinner machines.