英伟达GTC盛会明年3月16日举行 黄仁勋将释出Vera Rubin最新进展

Group 1 - Nvidia will hold its annual AI conference "GTC 2026" in San Jose, California, from March 16 to 19, 2026, with CEO Jensen Huang expected to deliver a keynote speech [1] - The next-generation AI platform, Vera Rubin, is anticipated to be a focal point, with strong order flow from Nvidia boosting its major server manufacturing partner, Hon Hai [1][2] - Huang is also expected to provide insights on the next-generation Feynman platform during the keynote, alongside showcasing new supply chain partners and clients [1] Group 2 - The industry expects Huang's keynote to focus on AI infrastructure, particularly the Vera Rubin platform, as well as advancements in CUDA technology and robotics [2] - The Vera Rubin NVL144 platform is projected to achieve 3.6 Exaflops of FP4 inference and 1.2 Exaflops of FP8 training performance, representing a 3.3 times improvement over the GB300 NVL72 [2] - Hon Hai is already developing servers for the Vera Rubin platform, planning to start mass production in the second half of next year, which is expected to drive significant growth from 2026 to 2027 [2]