March 20, 2024: Nvidia, the US chip manufacturer with a market value of over $2 trillion, has expanded its leadership position in the sphere of artificial intelligence with the notice several groundbreaking technologies at its annual developer conference on Monday.
The company presented the “AI chip series “Blackwell”designed to power the expensive data centers that train cutting-edge AI models equivalent to the most recent generations of GPT, Claude and Gemini.
The Blackwell B200 chip represents a big improvement over Nvidia's previous H100 AI chip, reducing the variety of chips and power consumption required to coach a large AI model like GPT-4 by 75%.
The company also introduced the GB200 “Superchip” that mixes two B200 chips and Nvidia’s Grace CPU on a single board, offering “30x performance” for server farms running chatbots like Claude or ChatGPT, while reducing energy consumption by as much as 25x.
Nvidia CEO Jensen Huang presented even larger configurations, equivalent to the GB200 NVL72, a single server rack with 72 B200 chips connected by nearly three kilometers of cable, and the DGX Superpod, which mixes eight of those racks into an AI data center the dimensions of a shipping container in a box.
In addition to the brand new AI chips, Nvidia announced Project GR00T, a basic model for controlling humanoid robots.
This model, together with the Jetson Thor system-on-a-chip, goals to create autonomous machines that may be given instructions using natural language to perform common tasks, even those for which they’ve not been specifically trained.
“The GR00T project will help robots understand natural language and mimic movements by observing human actions, quickly learning coordination, dexterity and other skills to navigate, adapt and interact with the actual world,” Huang explained.
Nvidia has also entered the sphere of quantum computing with the launch of a cloud service that uses its AI chips to simulate quantum computers.
While the service won’t initially be connected to an actual quantum computer, it would allow researchers to check their ideas without having to resort to rare and expensive quantum hardware.
Nvidia plans to offer access to third-party quantum computers via the platform in the long run.
As Nvidia continues to push the boundaries of artificial intelligence and robotics, the corporate's latest innovations underscore its commitment to leading the industry and shaping the long run of those transformative technologies.
