Jensen Huang Promises a Chip "the World Has Never Seen" at GTC 2026 in March
Announcement
NVIDIA CEO Jensen Huang revealed on February 19, 2026, that the company will unveil "chips the world has never seen before" at the upcoming GTC 2026 conference in San Jose. The event runs March 16–19, with Huang's keynote scheduled for March 16.
What Industry Expects
No specific product details have been disclosed, but speculation centers on several possibilities:
- Rubin Ultra/Next-Gen refresh: An upgrade to the Rubin platform already in production, set for second-half 2026 deployment
- Feynman early preview: A look ahead at the post-Rubin architecture planned for around 2028
- AI SSD: Novel storage hardware reportedly developed in partnership with SK Hynix and Kioxia, targeting 100 million IOPS
Current Roadmap Status
NVIDIA confirmed Rubin is in full production. AWS, Google Cloud, Microsoft Azure, and Oracle Cloud are among the first to deploy Vera Rubin-based instances in the second half of 2026. The GTC announcement could introduce additional products or accelerate the existing roadmap.
Source: HotHardware
Related Articles
NVIDIA revealed detailed specs for Vera Rubin NVL72. Each Rubin GPU delivers 50 PFLOPS inference (5x Blackwell GB200), 22 TB/s HBM4 bandwidth (2.8x Blackwell), and cuts inference cost per million tokens by 10x. Ships H2 2026.
NVIDIA announced the Rubin platform at CES 2026 in January. Comprising six new chips, the Vera Rubin superchip delivers 5x improved inference performance over GB200. Major AI companies including OpenAI, Meta, and Microsoft plan to adopt it.
NVIDIA unveiled its next-gen AI platform Rubin, delivering 10x reduction in inference token cost and 4x fewer GPUs for MoE model training vs. Blackwell. Launch planned for H2 2026.
Comments (0)
No comments yet. Be the first to comment!