Jensen Huang Declares Entry into the AI PC Market, Moves Beyond GPUs to Build CPUs Too [Computex 2026]

[Financial News, Taipei (Taiwan) = Reporter Choi Hye-rim, Jeong Won-il] "It took 33 years just to develop this chip. A computer equipped with 'N1X' can run every piece of software ever created by humans and can also power AI agents."
On the 1st, NVIDIA CEO Jensen Huang unveiled the new 'N1X' chip for artificial intelligence (AI) PCs during his keynote at GTC Taipei, held at the Taipei Music Center in Taiwan. The chip is expected to be installed in the RTX Spark series of laptops, signaling NVIDIA's full-scale entry into the PC market. Huang also declared that the next-generation AI computing platform, 'Vera Rubin,' has entered mass production.
■ "Developing an AI PC chip optimized for Windows"
He described the N1X chip as "the most amazing chip in the world" and said it had been optimized through deep collaboration with Microsoft (MS). N1X is a System on Chip (SoC) that combines NVIDIA's own CPU and the Nvidia Blackwell graphics processing unit (GPU) for AI. Huang said this makes it powerful enough to run NVIDIA's entire software stack, including digital biology, earthquake processing, astrophysics, genomics, AI and computer graphics, as well as all Microsoft Windows applications, at 100%. According to major foreign media reports, early laptop partners for RTX Spark include Acer Inc., ASUSTeK Computer Inc. (ASUS), GIGABYTE, MSI, Dell, Hewlett-Packard, Lenovo and MS. The product is expected to launch this fall. Samsung Electronics is also seen as a strong potential future partner.
He also pushed back directly against claims that the AI boom is a bubble. Huang said, "AI is now a profit engine and a GDP engine." He explained that while data centers used to be infrastructure that only spent money, they are now evolving into revenue-generating facilities that directly convert AI-generated tokens into sales and profits. "Customers no longer want to buy computers. They want to build AI factories," Huang said, adding that NVIDIA aims to move beyond being a GPU maker and become an infrastructure company that builds the AI factory ecosystem.
■ "'Vera Rubin' is also entering mass production," with Samsung and SK hynix HBM expected to be installed
Huang's emphasis on the next-generation CPU 'Vera,' designed to power AI factories, was in the same context. He said, "In the past, CPUs were designed for humans, but now we need CPUs for agents," adding that "far more AI agents than humans will be running nonstop."
He also said the next-generation AI computing platform 'Vera Rubin,' scheduled for release in the second half of this year, has entered full-scale mass production. Huang said, "Vera Rubin was not built simply to run AI. It was built for agents," and added that it is now in full production. Vera Rubin is expected to be equipped with next-generation High Bandwidth Memory 4 (HBM4) supplied by Samsung Electronics, SK hynix and Micron.


kaya@fnnews.com Reporter Choi Hye-rim, Jeong Won-il Reporter