“`html
CNBC AI News, August 31 – Mark your calendars: NVIDIA has announced that its highly anticipated GTC 2026 will take place March 16-19, 2026, in San Jose, California.
GTC has long been a bellwether for advancements in GPU technology and data center innovation, consistently drawing the attention of developers, researchers, and investors worldwide.
At this year’s GTC 2025, NVIDIA unveiled its roadmap for future GPU hardware, confirming plans to upgrade to Blackwell Ultra (NVL72) and announcing the launch of the Vera Rubin architecture in 2026. This architecture is poised to scale to the NVL144 system, enabling massive-scale deployments.
Looking further ahead, NVIDIA also offered a sneak peek at Rubin Ultra (NVL576) slated for 2027, and the next-generation Feynman architecture due in 2028.
On the software front, GTC 2025 saw NVIDIA unveiling several key initiatives, including NVIDIA Dynamo (dubbed the operating system for AI factories), Isaac GR00T N1 (a foundational model for robotics), Cosmos AI model (for synthetic training data), and Newton (a new robotic simulation physics engine).
The upcoming GTC 2026 is expected to heavily feature the debut of the Rubin GPU and Vera CPU, with a keen focus on their deployment timelines, performance metrics, and their capabilities to push beyond the boundaries established by Blackwell. Industry insiders will be watching closely for details on how these advancements will shape NVIDIA’s competitive edge and influence the broader landscape of accelerated computing.
“`
Original article, Author: Tobias. If you wish to reprint this article, please indicate the source:https://aicnbc.com/8389.html
