TMTPOST--Nvidia CEO Jensen Huang has announced that the company plans to introduce a new version of its flagship AI chips annually. As the world's most valuable chipmaker, Nvidia aims to capitalize on the $100 trillion worth of opportunities presented by what Huang describes as a "new Industrial Revolution."

Speaking at a keynote address in Taipei on Sunday, Huang revealed that a significant upgrade of Nvidia's next-generation Rubin GPU (graphics processing unit) is in development and will be available in 2026. Additionally, Nvidia recently announced its plans to begin shipping the new Blackwell GPU system in the coming months, with an upgraded version, Blackwell Ultra, scheduled for release next year.

"Our company is on a one-year rhythm," Huang said during his speech at the National Taiwan University sports stadium ahead of the Computex 2024 tech industry event. He emphasized Nvidia's commitment to pushing the boundaries of technological innovation in chipmaking and maintaining an annual release schedule for their customers.

“Accelerated computing is sustainable computing,” he emphasized, outlining how the combination of GPUs and CPUs can deliver up to a 100x speedup while only increasing power consumption by a factor of three, achieving 25x more performance per Watt over CPUs alone.

“The more you buy, the more you save,” Huang noted, highlighting this approach’s significant cost and energy savings.

At NVIDIA GTC 2024 in March, Huang said that in just eight years, NVIDIA has increased computational capacity by a thousandfold, a progress that far exceeds the benchmarks set during the heyday of Moore’s Law.

“In the past eight years, we’ve increased computation by 1000 times, and we have two more years to go. So that puts it into perspective [the fact that] the rate at which we’re advancing computing is insane. And it’s still not fast enough,” said Huang. 

In a presentation that blended artificial intelligence simulations with personal anecdotes, Huang highlighted the vast potential of Nvidia's chip solutions for AI factories, predicting they will unlock "$100 trillion" in opportunities across various sectors, including computing, healthcare, transportation, and manufacturing.

Nvidia's Omniverse graphic simulation platform, which the company has been promoting since 2011, played a significant role in Sunday’s presentation. The event featured a digitally simulated version of Jensen Huang for a portion of the presentation, showcasing the capabilities of the Omniverse platform.

“The era of robotics has arrived. Everything that moves will one day be autonomous,” said Huang. “We are working to accelerate generative physical AI by advancing the NVIDIA robotics stack, including Omniverse for simulation applications, Project GR00T humanoid foundation models, and the Jetson Thor robotics computer.”

In his keynote speech, Huang demonstrated robots used in transportation, healthcare and industrial manufacturing. In one demonstration, Foxconn, the world’s largest electronics manufacturer, showcases a fully simulated autonomous factory in NVIDIA Omniverse, featuring fleets of AI robots developed by NVIDIA robotics partners, based on NVIDIA Isaac.

"Everything is going to be robotic. All of the factories will be robotic. The factories will orchestrate robots. And those robots will be building products that are robotic robots interacting with robots, building robotic products," said Huang.

"Generative Physical AI can learn skills using reinforcement learning from physics feedback in a simulated world. In these simulation environments, robots learn to make decisions by performing actions in a virtual world that obeys the laws of physics,” he added.

"The IT industry is about to create something that can directly serve $100 trillion worth of industries. It's no longer just an instrument for information storage or data processing, but a factory for generating intelligence for every industry. This has never happened before," Huang said. "What started with accelerated computing led to AI, led to generative AI, and now an industrial revolution."

Huang also announced collaborations with major Taiwanese electronics suppliers, including Foxconn, Pegatron, Wistron, and Delta Electronics. These partnerships aim to integrate Nvidia's AI technologies, such as real-time 3D simulations and AI robotic development and deployment, into their manufacturing facilities.

Huang said that the computer industry is undergoing a significant reset. He emphasized that two core technologies, accelerated computing and artificial intelligence, are the transformative forces that "will reshape the computer industry."

Huang elaborated on Nvidia's role in addressing what he termed "computation inflation." He explained that while the performance of CPUs (central processing units), which handle most computer processing tasks, has significantly slowed, the demand for computation continues to grow exponentially.

In one presentation, he named the Chinese name of token as “词元”.

He stressed that this disparity necessitates innovative solutions like Nvidia's accelerated computing, which aims to bridge the gap and meet the rapidly increasing computational demands.

"AI factories" are now feasible given GPUs (graphic processing units) are integrated with CPUs, which means many calculations can be made simultaneously and makes AI processing faster and more efficient, he noted.

The computer "is no longer just an instrument for information storage or data processing, but a factory for generating intelligence for every industry," he said.

"This is going to be a manufacturing industry, not a manufacturing industry of computers, but using the computers in manufacturing, which has never happened before," Huang pointed out, saying that what started with accelerated computing, "led to AI, led to generative AI, and now an industrial revolution."

Huang also introduced a generative AI model for weather and climate research, as well as AI tools that will accelerate the next wave of "digital humans," and "pave the way for future GenAI avatars."