Nvidia just made its most aggressive move yet into the personal computing space. The company unveiled RTX Spark at GTC Taipei during Computex on June 1, 2026, a new superchip designed from the ground up to run AI agents locally on Windows laptops, no cloud required.
CEO Jensen Huang framed the technology as transforming PCs from mere tools into “collaborative teammates.”
What RTX Spark actually is
The RTX Spark is built on Nvidia’s Blackwell architecture and packs some genuinely absurd specifications for something meant to fit inside a thin-and-light notebook.
The chip delivers up to 1 petaflop of AI performance. For context, that’s a quadrillion floating-point operations per second.
