The Nvidia booth at the Nvidia GTC global AI conference in San Jose, Calif., on March 17.Carlos Barria/Reuters
Cadence Design Systems CDNS-Q and Nvidia NVDA-Q are partnering to further the development of artificial intelligence for robots, the chief executive officers of the two companies said on Wednesday.
Cadence, which is one of the major suppliers of the software used in designing advanced computing chips, is working with Nvidia to integrate its physics engines, which predict how real-world materials interact, with Nvidia AI models designed to train robots inside computer simulations.
“We’re working with you across the board on robotic systems,” Nvidia CEO Jensen Huang said at a conference hosted by Cadence in Santa Clara, Calif.
Training robots inside such simulations can be faster than training them in the real world, but the training data for doing so is not readily available and must be generated by software such as Cadence’s physics engines. The goal of the collaboration, the two CEOs said, is to shrink the time needed to get robots to carry out useful tasks.
“The more accurate [generated training data] is, the better the model will be,” said Cadence CEO Anirudh Devgan.