Nvidia CEO unveils robotics training technology, Toyota deal and new gaming chips
By Max A. Cherney and Stephen Nellis
(Reuters) – AI to better train robots and cars, and new gaming chips dominated Nvidia CEO Jensen Huang’s keynote speech at the CES 2025 conference on Monday as the world’s second-largest company outlined its potential to expand its business.
Nvidia has introduced what it calls Cosmos ‘baseline’ models that generate realistic video that can be used to train robots and self-driving cars at a much lower cost than using traditional data.
Using what’s known in the tech industry as “artificial” training data, the models can help robots and cars understand the virtual world in the same way that large language models have helped chatbots generate responses in natural language.
Users will be able to provide Cosmos with a text description that can be used to generate a video of the world that obeys the laws of physics.
This promises to be much cheaper than collecting data as it is done today. To train self-driving cars, for example, companies have dozens of cars roaming the streets to collect data, and humanoid robots are often trained by having real people repeat tasks over and over.
Huang, however, warned that the Cosmos models will need more data before hitting their ‘ChatGPT moment’.
Cosmos will be made available under an “open license,” similar to the Meta Platforms Llama3 languages widely used in the technology industry.
“We really hope (Cosmos) will do for the world of robotics and industrial AI what Lama3 did for business AI,” Huang said.
New gaming chips use Nvidia’s ‘Blackwell’ AI technology to give video games movie-like graphics, particularly in a field known as ‘shaders,’ which can help images like a ceramic teapot look more realistic by adding imperfections and fingerprint smudges to its surface. . .
The new chips also feature AI technology to help game developers produce more accurate human faces, an area where players are more likely to notice less realistic features. The chips, which Nvidia calls its RTX 50 series, will range in price from $549 to $1,999, with the top models arriving in Jan. 30 and lower-end models arriving in February.
Nvidia said the $549 mid-grade gaming chips will be compatible with the company’s previous chip, the RTX 4090 which sold for $1,600.
Nvidia also said Toyota Motor will use its Orin chips and automotive operating system to enable advanced driver assistance in several models. It did not provide details about the models.
Huang expects automotive hardware and software revenue of $5 billion in fiscal 2026, up from $4 billion expected this year.
Source link
