Nvidia's Advances and Challenges in Robot Technology
On October 2nd, at the 2024 Computex Taipei, Nvidia CEO Jensen Huang, dressed in his signature black leather jacket, passionately introduced the new darling of future technology - humanoid robots.
He jokingly said that these robots are about the same size as him, and emphasized: Robot technology has arrived, physical artificial intelligence has arrived, this is not science fiction.
The robots Jensen Huang demonstrated were virtual, but the wheeled delivery robot on the stage gave a glimpse into the corner of future technology. He disclosed that in the future, two high-volume robot products will come out, among which self-driving cars and humanoid robots are highly anticipated. These technologies, thanks to the rapid development of machine learning, are gradually merging and maturing.
However, although Nvidia has achieved great success in the field of graphics processing units and its market capitalization once reached as high as 3 trillion US dollars, robot technology currently only accounts for a small part of its revenue. Jensen Huang knows well that to make robot technology a reality, many challenges need to be overcome.
In order to promote the development of robot technology, Nvidia has launched a series of tools and platforms. Omniverse is a simulation platform that developers can use to test the simulation of robots; Isaac is a gym built on Omniverse for training robots to perform tasks. In addition, Nvidia also provides the Jetson Thor chip for robots, and the Groot project known as the Moon Landing Project aiming to create a basic model for humanoid robots.
Jensen Huang believes that humanoid robots are easier to manufacture. He emphasized that since humanoid robots have roughly similar physiques as humans, there is more data to train these robots. However, collecting these data still requires time, effort and money.
Although Nvidia has made quite a few progress in the robot field, some experts said that the company has not yet lowered the barrier to entry for robot manufacturers like in the language and image AI fields. Competitors are joining in, trying to carve out an ideal stack before Nvidia takes the dominant position.
However, Nvidia also realizes the importance of cooperation with other companies. Jensen Huang said that Nvidia is cooperating with every robot and artificial intelligence company on the earth to help them deploy robots faster. In addition, the company also launched the humanoid robot developer program, providing developers with a series of tools and resources.
Despite many challenges, Nvidia is full of confidence in the future of robot technology. Jensen Huang said that self-driving cars and robots will become an important part of Nvidia's future development. He predicted that eventually everything will be automated, starting from robotic arms and vehicles, and then buildings and even cities.
With the rapid development of artificial intelligence technology and the active promotion of technology giants such as Nvidia, robot technology is ushering in an unprecedented development opportunity. However, to truly realize the widespread application and commercial landing of robots, many technical, legal and ethical issues still need to be overcome.