Jensen Huang, the head of Nvidia, delivered a presentation at a consumer-electronics exhibition in Nevada on Monday, highlighting advancements in AI agents, autonomous vehicles, and robotics on a day when his firm’s market value peaked.
Speaking at the annual CES event in Las Vegas, Huang showcased “physical AI” instruments that he claimed would assist robots in learning through simulated environments closely resembling the real world. Such advancements could increase automation in warehouses and production facilities and elevate a humanoid-robot market projected to reach $38 billion in the coming decades.
“The ChatGPT moment for general robotics is imminent,” he remarked, surrounded by over a dozen humanoid robots. This statement referred to the surge of interest in AI following the debut of OpenAI’s chatbot more than two years ago.
In his speech, which came after a trading session that propelled Nvidia’s worth to $3.66 trillion, Huang emphasized the company’s rapid progress within the automotive sector.
He announced that Nvidia secured a contract to supply driver-assistance chips and software for forthcoming Toyota automobiles, supplementing a roster of automotive alliances that includes Mercedes and Volvo. He predicted that the company’s automotive earnings would reach $5 billion in the upcoming fiscal year, starting late this month, up from approximately $4 billion this year.
The organization accounts for automotive earnings in its financial disclosures, recording $1.1 billion in its last complete fiscal year, though Huang’s $5 billion estimate also encompassed sales of additional products and services to the sector.
Among Huang’s other declarations:
- Nvidia will introduce a personal AI supercomputer named Project DIGITS. This will be a desktop unit featuring a version of its latest Blackwell AI chip, starting at $3,000. These computers target AI researchers and data scientists, enabling them to work on AI models without relying on Nvidia’s premier AI chips located in data centers.
- New AI “blueprints” designed to simplify the creation and deployment of AI agents for tasks such as video feed analysis and blog post generation. One of Nvidia’s blueprints allows users to input multiple PDF documents, from which it produces a podcast “narrated in a natural voice,” according to a company announcement.
- A new generation of graphics chips tailored for gamers. This hardware, with a price tag reaching about $2,000, enhances resolution and frame rates for demanding games, in part through AI integration, as stated by executives. The chips are slated for desktop computers this month, with laptop availability expected in March.
Contact Asa Fitch at asa.fitch@wsj.com