Next-Generation AI Accelerator Production Unveiled

Samsung, SK hynix, and Micron Memory Highlighted

Based on TSMC's 3-Nanometer Process

Jensen Huang, CEO of Nvidia, announced that the next-generation artificial intelligence (AI) accelerator, Vera Rubin, has officially entered full-scale production.


During his keynote speech at the 'Nvidia GTC Taiwan 2026' held at the Taipei Music Center in Taiwan on June 1, 2026 (local time), CEO Huang stated, "Vera Rubin is in full production." He also added that the sixth-generation high-bandwidth memory (HBM4) from Samsung Electronics, SK hynix, and Micron will be integrated into the product.


Jensen Huang, CEO of NVIDIA, is delivering a keynote speech at the "NVIDIA GTC Taiwan 2026" held at the Taipei Music Center in Taiwan on the 1st (local time). Photo by Jin Young Kim.

Jensen Huang, CEO of NVIDIA, is delivering a keynote speech at the "NVIDIA GTC Taiwan 2026" held at the Taipei Music Center in Taiwan on the 1st (local time). Photo by Jin Young Kim.

View original image

According to CEO Huang, Vera Rubin is manufactured based on TSMC's 3-nanometer process. It consists of seven new chips, each produced through hundreds of semiconductor manufacturing steps, and utilizes CoWoS-L packaging and HBM4 memory.



The Vera Rubin system features approximately 6 trillion transistors and more than 18,000 components integrated into a single computing line. Nvidia explained that this design enables the simultaneous execution of large-scale AI computation and data processing.


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing