Jensen Huang used an unscheduled conversation with The Korea Economic Daily in Santa Clara, California to signal a major reveal at Nvidia’s upcoming GPU Technology Conference. Fresh from a casual meetup with Nvidia and SK hynix engineering teams, Huang responded to questions about the conference, which runs from March 16 to 19, by stating that “a chip that will surprise the world will be unveiled at GTC next month.” He invited extensive questioning with the remark “ask as much as you want,” suggesting strong confidence in the company’s roadmap and the significance of the planned announcements.
Huang is expected to maintain his yearly cadence of headline keynote presentations, with 2026 framed as a pivotal launch year for the “Vera Rubin” platform. Industry observers anticipate a VR200 NVL72 heavy showcase on stage, aligning the event with Nvidia’s next wave of data center and accelerator hardware. During the Q&A, Huang expanded on the scope of what is coming by saying “we’ve prepared several new chips the world has never seen before,” hinting that the surprise unveiling will sit alongside a broader portfolio refresh rather than a single flagship reveal.
The discussion also pointed to possible early appearances of “Feynman” Artificial Intelligence accelerator prototypes at GTC 2026, as well as potential “Vera GPU” derivatives that could extend the Vera Rubin architecture into new segments. Huang acknowledged the difficulty of pushing semiconductor technology forward, stating that “nothing is easy because all technologies are at their limits,” underscoring the engineering challenges behind Nvidia’s roadmap. The reporter highlighted Huang’s just concluded meal with “NVIDIA and SK Hynix memory engineers,” and in describing the HBM4 focused partnership with SK hynix, Huang emphasized its strategic importance by saying “with a team like this, nothing is impossible,” signaling deeper collaboration on next generation high bandwidth memory to support the forthcoming chips.
