CES: Nvidia unveils next-generation family of gaming chips

He also introduced an AI model, called Cosmos, which he said could generate video that can be used to train robots and self-driving cars at a much lower cost than current methods.

By creating what is known in the industry as "synthetic" training data, the model can help robots and cars better understand the physical world.

Users will be able to give Cosmos a text description that can be used to generate video of a world that obeys the laws of physics.

"All of the enabling technologies that I've been talking about is going to make it possible for us in the next several years to see very rapid breakthroughs, surprising breakthroughs in general robotics," he predicted, though he added much more training data would be needed.

Mr Huang carried out a real-time demonstration of the new gaming chip that showed off highly detailed graphics featuring an array of textures and manoeuvres.

"It was awesome that they can do this in real time," said Gary Yang, a graduate student in robotics at the California Institute of Technology.

"Previously we'd think of these graphics as pre-rendered."

The new chips will start making their way to consumers starting in late January.

"I thought it was incredible," said Scott Epstein of technology start-up Agenovate AI.

"They are continuing to innovate."

Mr Yang and Mr Epstein were among thousands of people who watched the speech both in person and virtually on the eve of the official opening of CES.