Nvidia’s tiny $3,000 computer wins the show at CES

Nvidia CEO Jensen Huang was welcomed as a rock star at this week’s CES in Las Vegas after AI boom this made the chip maker the second most valuable company in the world.

In his nearly two-hour keynote on Monday, which kicked off the annual conference, Huang packed the 12,000-seat arena, drawing comparisons to how Steve Jobs unveiled products at an apple events.

Huang finished with an Apple-like trick: a surprise product reveal. He brought out one of Nvidia’s server racks and, using stage magic, held up a much smaller version that looked like a tiny computer cube.

“It’s a supercomputer with artificial intelligence,” Huang said, donning an alligator leather jacket. “It works with Nvidia’s entire AI stack. All of Nvidia’s software runs on it.”

Huang said the computer is called Project Digits and runs on a relative of Grace Blackwell’s graphics processing units (GPUs), which currently power the most advanced AI server clusters. The GPU is paired with an ARMGrace CPU based. Nvidia partnered with a Chinese semiconductor company MediaTek to create a system-on-a-chip called GB10.

Formerly known as the Consumer Electronics Show, CES is usually the place to launch flashy and futuristic consumer gadgets. At this year’s trade show, which began Tuesday and ended Friday, several companies announced the integration of AI into home appliances, laptops and even grills. Other major announcements included a laptop from Lenovo that has a collapsible screen that can expand vertically. There were also new robots, including a Roomba competitor with a robotic arm.

Unlike traditional Nvidia gaming GPUs, Project Digits isn’t aimed at consumers. instead, it’s aimed at machine learning researchers, small companies and universities that want to develop advanced artificial intelligence but don’t have billions of dollars to build huge data centers or buy enough cloud credit.

“For data scientists and ML researchers who are actively working, actively building something,” Huang said. “Maybe you don’t need a giant cluster. You just develop early versions of the model and iterate constantly. You could do it in the cloud, but it costs a lot more money.”

The supercomputer will cost around $3,000 when it becomes available in May, Nvidia said, and will be available from the company itself as well as some of its manufacturing partners. Huang said Project Digits is a placeholder name, indicating that it could change by the time the computer goes on sale.

“If you have a good name for it, contact us,” Huang said.

Data center sales accounted for 88% of Nvidia’s $35 billion in revenue the last quarter.

Wall Street is focusing on Nvidia’s ability to diversify its business to become less dependent on the handful of customers who buy massive AI systems.

“It was a little scary to see Nvidia release something so good at such a low price,” Melius Research analyst Ben Reitz wrote in a note this week. He said Nvidia may have “stolen the show” with Project Digits, as well as other announcements including gaming graphics cards, new robot chips and a deal with Toyota.

Project Digits, which runs on Linux and the same Nvidia software used in the company’s GPU server clusters, represents a huge increase in opportunity for researchers and universities, said David Bader, director of the Institute for Data Science at the New Jersey Institute of Technology.

Bader, who has worked on research projects with Nvidia in the past, said the computer appears to be capable of processing enough data and information to train the largest and most advanced models. He told CNBC Anthropic, Google, Amazon and others “would pay $100 million to build a supercomputer for learning” to get a system with that capability.

For $3,000, users will soon be able to get a product that can be plugged into a standard electrical outlet at home or in the office, Bader said. This is especially interesting for scientists, who often go to private businesses to get access to larger and more powerful computers, he said.

“Any student who can have one of these systems that costs about the same as a high-end or gaming laptop, they’ll be able to do the same research and build the same models,” Bader said.

Wrights said the computer could be Nvidia’s first step into the $50 billion market for chips for PCs and laptops.

“It’s not too hard to imagine that it would be easy to just do it yourself and let the system run Windows someday,” Reitzes wrote. “But I guess they don’t want to step on too many toes.”

Huang did not rule out such a possibility when asked by Wall Street analysts on Tuesday.

He said MediaTek may sell the GB10 chip to other computer makers in the market. He made sure to leave some mystery in the air.

“Obviously we have plans,” Huang said.

WATCH: Nvidia pullback due to CES expectations