Unlock Editor's Digest at no cost
Roula Khalaf, editor of the FT, picks her favorite stories on this weekly newsletter.
Nvidia is on the verge of revolutionizing robotics through artificial intelligence, CEO Jensen Huang said Monday as he laid out his vision for the following stage of the corporate's stunning growth.
Huang announced quite a few latest products and partnerships in “physical AI” during his keynote address on the annual Consumer Electronics Show in Las, including AI models for humanoid robots and a significant cope with Toyota to make use of Nvidia Vegas' self-driving automotive technology.
Nvidia has surpassed a $3 trillion market cap and develop into one of the vital precious corporations on the earth on account of demand for its AI chips. Huang, for his part, has develop into a household name greater than 30 years after he founded Nvidia as a video game graphics chip company.
Long before the keynote began, huge lines had formed outside the Mandalay Bay Convention Center, and a few people were still standing in line when Huang got here on stage in a shiny version of his signature leather jacket and quipped, “I'm in Las Vegas, in spite of everything.”
Outside the semiconductor industry, Nvidia has developed the software that permits corporations to coach and deploy robots, from those utilized in smart factories and warehouses to self-driving cars and humanoids, and is pushing to expand the use cases for AI that come up its chips are running, to expand.
Overcoming the technological challenges of deploying robots on a big scale will pave the option to “the biggest technology industry the world has ever seen,” Huang said.
Nvidia said the sector of robotics has reached a technological tipping point as AI accelerates and refines the technique of simulating the physical world and generating the huge amounts of knowledge needed to coach robots. The marketplace for humanoid robots alone is predicted to achieve $38 billion over the following 20 years, the corporate said.
On Monday, Nvidia announced a set of foundational AI models on its latest Cosmos platform, allowing developers to generate data and construct their very own models at no cost.
Nvidia said the bottom models, which it said were trained on 20 million hours of video data, were as fundamental a technological development as the massive language models that underpin apps like OpenAI's ChatGPT. It is coupled with Nvidia's Omniverse platform, which runs simulations of the physical world.
“What (these models) do for language, we will now do for understanding the physical world,” Rev Lebaredian, Nvidia’s vp of omniverse and simulation technology, told the Financial Times. While data concerning the physical world is far harder to gather and process than text, Lebaredian said it’s “a essential part” of the corporate's mission.
“The big takeaway (from Huang's CES speech) is that this moment will likely be special,” he added. “I feel this 12 months is a turning point where we’re going to see this acceleration of physical AI and robotics.”
The Omniverse platform and robotics currently represent only a small portion of the corporate's total revenue. In Nvidia's quarter ending October, “skilled virtualization” accounted for $486 million in revenue, while automotive and robotics corporations totaled $449 million.
This is only a fraction of total revenue, as the corporate generated $30.8 billion in revenue in the course of the same period from selling chips for the info centers that run AI models.
Nvidia's search for brand spanking new markets comes amid growing pressure from its biggest customers, including Amazon and Microsoft, who’re rushing to construct their very own chips for AI data centers.
Analysts at Bank of America said Nvidia's decision to ramp up “physical AI” was the “next logical step.” The challenge could be to “make the products reliable enough, low cost enough and widespread enough to provide credible business models,” they added.
At CES, Nvidia also unveiled a set of base models for humanoid robots called the “GR00T Blueprint” that it said would “stimulate” robot development, in addition to latest tools for developing and testing fleets of factory and warehouse robots and training autonomous vehicles.
Toyota announced it will construct its next generation of autonomous vehicles on Nvidia's hardware and software, often called Drive AGX. Self-driving automotive company Aurora and auto parts maker Continental will use Nvidia's hardware and software to power 1000’s of driverless trucks as a part of their long-term strategic partnerships with the chipmaker.
Nvidia expects its automotive business to grow to $6 billion in fiscal 2026. Autonomous vehicles “will likely be the primary multi-billion dollar robotics industry,” Huang told the CES audience.
Separately, Nvidia announced it’s going to launch a “personal AI supercomputer” with its newest and strongest AI chip, Blackwell, allowing researchers and students to run AI models with billions of parameters locally as a substitute of via the cloud. It will likely be available in May with an initial price of $3,000.
