Jensen Huang Delivers Nvidia GTC Keynote
Crowds Exceed 6,000 Attendees
"From Generative AI to the Age of Agent AI"
'Vera Rubin' Supercomputer Ushers in a New Era

Nvidia has officially announced the start of mass production for its next-generation artificial intelligence (AI) accelerator, ‘Vera Rubin,’ emphasizing that the center of gravity in the AI industry is shifting from generative AI to ‘agent AI (AI assistants).’ Having started as a graphics processing unit (GPU) company, Nvidia declared its transformation into an ‘AI systems company’ that encompasses not only AI semiconductors but also data centers, networks, central processing units (CPUs), and software. Alongside this, Nvidia brought to the forefront its collaborations with major Korean partners such as Samsung Electronics, SK hynix, Naver, and Hyundai Motor Company, as well as its global clientele.


Jensen Huang, CEO of Nvidia, stated during his keynote speech at Nvidia’s annual AI conference ‘GTC Taipei 2026,’ held on June 1 at the Taipei Music Center in Taiwan, “Vera Rubin has now entered full production,” and added, “Vera Rubin is not a computer for AI, but a computer for agents.”


Jensen Huang, CEO of Nvidia, is delivering a keynote speech at Nvidia's annual AI conference 'GTC Taipei 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 Nvidia's annual AI conference 'GTC Taipei 2026,' held at the Taipei Music Center in Taiwan on the 1st (local time). Photo by Jin-Young Kim.

View original image

CEO Huang forecast that software architecture will shift from being centered on traditional applications (apps) to being agent-centric. He explained, “Agents observe, reason, plan, and use tools,” and emphasized, “To process agents, a new computing architecture is needed where models, tools, runtime, and memory systems are all integrated as one.”


Nvidia explained that it developed Vera Rubin to respond to this change. Vera Rubin, described as an ‘AI supercomputer’ with a multi-rack structure, integrates not only GPUs but also CPUs, networks, storage, and security processors, further underscoring its expanded capabilities.


"Beyond a GPU Company to a Systems Company"— Declaration of Ecosystem Expansion

On the 1st (local time), attendees are seated before Jensen Huang, CEO of Nvidia, delivers the keynote speech at Nvidia's annual AI conference 'GTC Taipei 2026' held at the Taipei Music Center in Taiwan. Photo by Jin-Young Kim.

On the 1st (local time), attendees are seated before Jensen Huang, CEO of Nvidia, delivers the keynote speech at Nvidia's annual AI conference 'GTC Taipei 2026' held at the Taipei Music Center in Taiwan. Photo by Jin-Young Kim.

View original image

More than 6,000 attendees filled the auditorium for the speech that day. From as early as 8:00 a.m., three hours before the event began, a diverse crowd of general audience members, businesspeople, and journalists lined up to enter the venue. Even before the speech began, the 6,000-seat hall was already at full capacity. Notable figures at the event included Madison Huang, Senior Director at Nvidia, and Tae-won Chey, Chairman of SK Group. As usual, CEO Huang appeared wearing a leather jacket and greeted the local audience by saying “Dàjiā hǎo (Hello, everyone),” as he began his speech.


On this day, he demonstrated how Nvidia’s business focus is expanding beyond semiconductor supply to encompass the entire infrastructure of AI data centers and AI factories. CEO Huang predicted, “Every company will become an agent-operating company in the future,” and added, “The next 10 years of the AI era will be led by agents.” He continued, “In the past, Nvidia was a GPU company, but now it has become a systems company. Customers are no longer trying to buy computers; they want to build AI factories. This is why Nvidia is once again beginning to transform itself.”


He particularly emphasized the economics of AI infrastructure, repeatedly stating, “Compute is revenue.” He explained that the core competitive edge of AI data centers lies not in the price of chips, but in throughput per watt (Token per Watt), power efficiency, and system reliability. CEO Huang stressed, “If the wrong architecture is chosen, profitability cannot be secured, even if the chips are inexpensive,” and underscored, “In the era of AI factories, computing performance translates directly into revenue and profit.”


On this day, Nvidia also unveiled ‘Vera,’ a CPU optimized for the agent AI era. CEO Huang said, “Until now, CPUs have been designed for people, but Vera is the first CPU designed for agents,” and predicted, “In the future, the number of agents will far exceed the number of humans.”


Shoutout to Korean Partners: Samsung, SK hynix, Hyundai Motor

NVIDIA CEO Jensen Huang is introducing collaboration with Naver Cloud during the 'NVIDIA GTC 2026' keynote speech held on the 1st at the Taipei Music Center in Taiwan. Screen capture from NVIDIA YouTube.

NVIDIA CEO Jensen Huang is introducing collaboration with Naver Cloud during the 'NVIDIA GTC 2026' keynote speech held on the 1st at the Taipei Music Center in Taiwan. Screen capture from NVIDIA YouTube.

View original image

CEO Huang highlighted Nvidia’s relationships with a variety of global partners, including Korean companies. In particular, he explained that the sixth-generation high-bandwidth memory (HBM) HBM4 used in Vera Rubin will be supplied by SK hynix, Samsung Electronics, and Micron. He added that the product will be manufactured based on TSMC’s 3nm process and CoWoS-L packaging technology, as well as HBM4.


HBM is a core memory that enables AI semiconductors to process data quickly and determines AI server performance. Notably, Samsung Electronics has recently become the first in the industry to pass Nvidia’s HBM4 quality certification and has begun mass production and supply. This announcement reconfirms that Samsung Electronics and SK hynix are key partners supplying semiconductors for next-generation AI accelerators.


Additionally, while presenting cases of collaboration with global cloud companies, Nvidia also displayed Naver Cloud on the screen along with U.S. Google, Japan’s Yotta, Indonesia’s Indosat, and Taiwan’s GMI. On the presentation screen, the words ‘Nvidia’ and ‘Naver Cloud’ appeared with a red heart image between them. With Naver Cloud introduced alongside a heart symbol among global cloud companies, it attracted significant attention from those in attendance.


Hyundai Motor Group also appeared during the latter part of the keynote. In a video of a robot performance revealed by CEO Huang, the Genesis model and the Ioniq electric vehicle from Hyundai Motor Group were featured. As Nvidia is increasingly focusing on robotics, autonomous driving, and physical AI, interest in its collaboration with Hyundai Motor Group is growing.



Meanwhile, on this day, Nvidia introduced its proprietary inference-based large language model (LLM), ‘Nemotron 3 Ultra.’ Nvidia plans to connect Nemotron to its AI software stack, enabling enterprises to more easily build their own AI agents. The company also unveiled its product roadmap for the next five years, including the next-generation accelerator ‘Feynman,’ scheduled for release in 2029, which will adopt next-generation HBM. Furthermore, the company announced that ‘Rosa Feynman Spark’ is planned for launch in 2030.


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