Jensen Huang just delivered the keynote at GTC Taipei 2026 and NVIDIA's next move is bigger than I expected. The Vera Rubin platform is officially detailed: 3.5 times the training performance of Blackwell, 5 times the inference performance, and inference costs drop to ONE-SEVENTH of what Blackwell costs today. That's not an upgrade. That's a generational leap.
And NVIDIA isn't stopping at GPUs. Jensen made it clear: the company is now in the AI factory business. Full-stack, data center-scale compute infrastructure. They also confirmed the N1X ARM laptop chip, which means NVIDIA is pushing into PC silicon directly. Qualcomm and Intel should be watching this closely.
Here's the bottom line: if you thought NVDA was expensive after the Blackwell cycle, the Rubin numbers suggest the market is still underpricing what this company is building. The chip export restriction headline is real, but it's a sideshow compared to the pace of what the product roadmap is doing. 📈
The copyright of this article belongs to the original author/organization.
The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.
